back to article Oracle fires Itanium countersuit at HP

Late Friday, Oracle filed a countersuit against HP, which sued Oracle back in June because Oracle said in March that it would not be developing future releases of its database, middleware, and application software on future Itanium processors. It's hard to tell who is stretching the truth more it in the ongoing lawsuit, and …

COMMENTS

This topic is closed for new posts.
  1. Anonymous Coward
    Anonymous Coward

    Go on HP

    Move your Oracle customers to Postgres on Itanium. You know you want to...

    1. Anonymous Coward
      Anonymous Coward

      PostgreSQL

      While Pg is a damn fine DB in its own right, it is no way a replacement for Oracle. It's slowly getting there, but it has a long way to go.

      1. Matt Bryant Silver badge
        Boffin

        RE: PostgreSQL

        "....it is no way a replacement for Oracle....." Yes and no. Whilst PostgreSQL can't offer all the features of the full Oracle suite, it doesn't need to for it to match 90+% of the installed Oracle instances out there. You'd be surprised the number of companies I hear of that have site-wide, full Entreprise licences, yet are only using the basic database for the majority of work. We had a review a few years back and managed to cut our Oracle bill in half by getting rid of unneccessary Enterprise licences and replacing some stand-alone Oracle instances with other, cheaper, DB products (including PostgreSQL). Don't get me wrong - we have Oracle at the core of our business and relie on it - but we're not fooled when the Oracle reps start doing their "do I have a deal for you" schpiel.

      2. Anonymous Coward
        Anonymous Coward

        Probably not....

        If Postgre gets close to Oracle DB in functionality/performance, I think Oracle will decide to acquire EnterpriseDB... pull a MySQL. DB2 is the only database that has a chance to cut into Oracle's market share.... maybe Sybase ASE if SAP really puts some cash into it.

        1. Matt Bryant Silver badge
          Facepalm

          RE: Probably not....

          Oracle DB's biggest threat has been eating into the Oracle isntalled base for years already - Microsoft SQL Server.

          1. Anonymous Coward
            Facepalm

            Probably not....

            MS SQL taking out Oracle? No way, they are totally in different leagues. If someone is replacing Oracle with MS SQL, they should have never purchased Oracle in the first place. Per Gartner, MS SQL has held steady at 10-15% market share for years (no major growth). Oracle is the clear number one, DB2 is the clear number two.

            There are a host of reasons why MS SQL is not in Oracle/DB2's league, but here are a few.

            - In MS SQL, the DBA has no control over memory utilization or caching. Memory allocation is decided globally within the server, not for caching, sorting, table space, etc.... Oracle and DB Deuce are the only granular resource DBs.

            - MS SQL blocks are always 8k. There is no way to handle large data objects.

            - MS SQL has no conception of partitioning for large tables and indexes. You can not break a table into smaller chunks for performance, maintenance, application purposes.

            - There are just a slew of features not included in MS SQL, such as star queries, bitmap indexes, reverse keys, etc, etc.

            - Last, but not least, MS SQL only runs on x86 hardware which doesn't provide anywhere near the cost, resiliency advantages or scale of Linux (cost) or Unix/mainframe (RAS, scale).... I am sure you are going to say that you can run MS SQL on Itanium, but not for long. It is EOL like the rest of Itanium.

            Siemens, SAP, Welchs, Pepsico, Coca-Cola, etc have all replaced Oracle with DB2 in recent years. I am not aware of a single replacement of Oracle with MS SQL for a large transactional workload.

            1. Matt Bryant Silver badge
              Happy

              RE: Probably not...

              Ah, nothing like an ostrich with his head firmly in the IBM sandpit! I would suggest you go take a look at just about any business out there, they are all implementing departmental servers with MS Windows and SQL Server (sadly, not much Linux), not AIX and DB2, nor Oracle and anything. That departmental database market used to be a key Oracle market. Why else do you think Oracle are desperately trying to shove appliances down their customers' throats? And it's not just there - tasks that used to be on UNIX and used Oracle databases are being displaced by MS SQL Server, often on VMware and on cheaper x64 servers which outperfrom many of the old UNIX systems they are replacing. And then there is the SQL Server underlying MS's Azure cloud offering, which is scaring the pants off Oracle and makes IBM's cloud offering look rather lacking.

              I also love the feature-sale - haven't seen such a funny drivelling of minor points since Sun died! All ignoring that many of the points mentioned are of zero interest to many database users (especailly those appliance users that just want a database out0of-the-box), which are usually just using the database a the back-end to a central application, and it is the latter where the tuning is usually most intense. I know DBAs like to think they make the whoel World run faster but the truth is there is a lot more important performance tuning outside the database.

              I would suggest you go spend some time reading at http://www.microsoft.com/sqlserver/en/us/product-info/why-sql-server.aspx, consider it as good for your career.

              1. Anonymous Coward
                Happy

                Firstly, you do not need to run DB2 on AIX. You can run it on anything you want to run it on, Linux on x86 included. Even MS Server if you really like applying patches and watching servers reboot.

                Secondly, it is time to buy a new calendar. It is not 1998 anymore. How many large organizations are in the process of implementing little workgroup database servers? None. They are ripping those out and centralizing on large, browser based applications for common functions with a common back-end DB. Centralizing small databases into large master databases is a ubiquitous initiative at every organization. No one wants to have 242 DBs with a "customer" table that all need to be synchronized to have one source of the truth. The Thick Clients and Little Server architecture is over. It is too administratively intensive and just generally a poor architecture.

                Thirdly, those are not minor points. The inability to performance tune, carve up data blocks, and scale muy importanto. Not to mention, no one wants to double down on Microsoft right now. Their era is coming to an end.

                Fourthly, database "out of the box", much to Oracle marketing's dismay, doesn't exist.

                I don't think Oracle or IBM is afraid of Microsoft anymore than usual. If anything, they were much more concerned five-ten years ago when it looked like Microsoft was going after enterprise IT. Now Microsoft is happily throwing Windows-Office monopoly money down the drain on consumer businesses while their MS Server, MS SQL, Dynamics businesses have stalled out.

                1. Matt Bryant Silver badge
                  FAIL

                  RE: Wunderbar1

                  "Firstly, you do not need to run DB2 on AIX...." Yeah, and the market figures for DB2 on anything but AIX are dismal to non-existant! Even PostgreSQL has a larger market share. DB2 is a minor player in the x64 database market even compared to homeless efforts like MySQL. Talking of MySQL, you may also want to check out the chart here http://www.mysql.com/why-mysql/marketshare/ which shows M$ SQL Server with almost twice the number of deployments as DB2. Epic fail!

                  "....How many large organizations are in the process of implementing little workgroup database servers? None....." Ah, did you stop to consider anything than the slice of the enterprise market IBM still manages to play in? How about SMBs? Still lots of them implementing small database servers. Same goes even for medium-sized businesses. But then I'm not surprised you wouldn't know that as that's two database markets IBM has been failing in for years.

                  "....Centralizing small databases into large master databases is a ubiquitous initiative at every organization...." LOL! I remember that schpiel, only the last time I heard it it was all about email, and using Lotus Notes and Domino (which is basically just a big DB2 instance with some very poor interfaces) to replace all those M$ Exchange Servers many business had back in the day. Yeah, like that worked for IBM - NOT! The complete failure of Notes/Domino to end the dominance of the M$ Outlook/Exchange combo just goes to show IBM's rediculous fascination with trying to sell Pee-series servers as centralised systems just doesn't chime with the customers. Trying to repeat the mainframe model with AIX and Pee-series was IBM's attempt at re-living the '80s, so it looks like you're the one needing a calendar.

                  ".....Thirdly, those are not minor points...." Sorry, I've sat through far too many Sun feature-sells not to laugh at your insistance that I have to consider points that have no impact on my projects just because you say they are vital. I suggest you try updating your sales schpiel.

                  "....Fourthly, database "out of the box", much to Oracle marketing's dismay, doesn't exist....." As long as you ignore the embedded database market, which is what IBM have been doing for years. There are tons of apps out there with embeded Oracle and M$ SQL instances bundled in, which simply work out-of-the-box and do not require DBAs to administer or tune them. Bl;ackberry BES is a very simple and common example, with many smaller businesses. But again, you won't know that as it's an area DB2 is simply absent from.

                  "......I don't think Oracle or IBM is afraid of Microsoft anymore...." Looking at the figures of the MySQL webby, it would probably be fair to say Oracle are a lot more worried about M$ SQL Server than DB2.

                  /SP&L

                  1. Anonymous Coward
                    FAIL

                    "Yeah, and the market figures for DB2 on anything but AIX are dismal to non-existant! Even PostgreSQL has a larger market share. DB2 is a minor player in the x64 database market even compared to homeless efforts like MySQL. Talking of MySQL, you may also want to check out the chart here http://www.mysql.com/why-mysql/marketshare/ which shows M$ SQL Server with almost twice the number of deployments as DB2. Epic fail!"

                    - I don't have any data on DB2 by platform. I am thinking you don't either. It wouldn't surprise me if AIX is the number one platform, but there are certainly large implementations on other platforms. SAP runs their internal SAP implementation on DB2 on HP-UX.... DB2 has 100% of the System i install base. The lion share of the mainframe install base, which are some of the largest RDBMS implementations on the planet.

                    - That MySQL link is laughable. Yes, MySQL and Postgre have more *deployments* than DB2 or Oracle (duh). Actually, I have one on this laptop. When software is free, it is easy to get millions of downloads. Look at the revenue numbers, look at the size of each deployment. Likewise, MS SQL has more deployments than DB2 or Oracle, but there are tons of Mom and Pops running MS SQL.

                    "Ah, did you stop to consider anything than the slice of the enterprise market IBM still manages to play in? How about SMBs? Still lots of them implementing small database servers. Same goes even for medium-sized businesses. But then I'm not surprised you wouldn't know that as that's two database markets IBM has been failing in for years."

                    - Again, these SMB companies are not implementing little workgroup DBs (which you said was the reason for MS SQLs impending dominance) those are centralized DBs. Yes, MS SQL's primary market is SMB. It can't scale and it is has SMB functionality, so that is expected. Again, MS SQL, holding steady in the Gartner numbers, no growth-no losses.

                    "the last time I heard it it was all about email, and using Lotus Notes and Domino (which is basically just a big DB2 instance with some very poor interfaces) to replace all those M$ Exchange Servers many business had back in the day. Yeah, like that worked for IBM - NOT! The complete failure of Notes/Domino to end the dominance of the M$ Outlook/Exchange combo just goes to show IBM's rediculous fascination with trying to sell Pee-series servers as centralised systems just doesn't chime with the customers. Trying to repeat the mainframe model with AIX and Pee-series was IBM's attempt at re-living the '80s, so it looks like you're the one needing a calendar."

                    - Nice diversion from the obvious fact that I am right about thick client - workgroup DBs being replaced by thin-client and centralized DBs.

                    - Since you want to discuss e-mail instead of databases, alright. Again, you see to be equating IBM software (platform independent) with the System p. It is not the case. Second, the obvious fact about the success of Microsoft Exchange is that it succeeded because it had Microsoft's Windows/Office look and feel and integration, not because Exchange is a technically superior architecture to Domino... or because distributed is superior to centralized. Actually, Lotus is still a solid second in the e-mail market. It is primarily because large companies, such as US Bank for instance with 100,000 seats, cannot play around with MS Exchange DAGS and DIY, flimsy mailbox server builds.

                    "Trying to repeat the mainframe model with AIX and Pee-series was IBM's attempt at re-living the '80s, so it looks like you're the one needing a calendar."

                    - Haha, if by the "mainframe" model you mean centralized, yes, everyone is trying to re-live the mainframe model. It is now called "cloud." How many companies have local copy installs on Microsoft Windows for new applications vs. running them on a centralized server via browser. Would that be all of them? SAP, Oracle... both entirely centralized and browser based. Facebook, Google, your precious Azure... all centralized, zero little workgroup thick clients. Back to the original point, no one is implementing a bunch of small workgroup MS SQL database. Everything is centralizing and coming through a browser.

                    'Looking at the figures of the MySQL webby, it would probably be fair to say Oracle are a lot more worried about M$ SQL Server than DB2."

                    - Yes, if Oracle and IBM decide to charge one price per implementation (the person who downloads DB2-C on their desktop pays the same as Exxon Mobil), you would have a point. However, deployments are meaningless to revenues, which is why the MySQL site is of no value unless you are trying to count free downloads.... By your ludicrous deployment measure, it is not MS SQL that Oracle should be concerned about, it is MS Access. After all, that is far and away the most deployed database in the world.

                    1. Matt Bryant Silver badge
                      FAIL

                      ".....I don't have any data on DB2 by platform...." I know, isn't it just so shockign that IBM try so hard to hide the fact that DB2 outside of AIX/mainframe is tiny? Let's put it this way - I have worked in educational, financial and government

                      "....That MySQL link is laughable...." So, roadmaps that prove you are wrong are "meaningless", and websites that display the error in your arguments are "laughable". I'm guessing that means you don't want any info in the discussion that doesn't have an IBM logo on it, yes? So, that makes you either an IBM employee, reseller, or just another big sucker of an IBM mainframe customer. Your denial just underlines your inability to see beyond your blinkered viwepoint.

                      "....MS SQL's primary market is SMB. It can't scale...." This is a very outdated view, probably due to your career being based on dinosaur mainframe tech. Even the standard Datacenter edition can scale to 256 logicial processors. I also suggest you go look at the MS SQL that underlies the Azure cloud service, just so you can see how out-of-touch you are. Major fail again! I suggest you just admit that you don't have a clue of anything outside of the mainframe ecosystem and leave it at that.

                      I love the bit where you rant about MS Exchange/Outlook, right after having to admit that it is massively more popular than Domino/Notes. Your "solid second" is second by a massive amount, even when you restrict it to business users. It's fair to say more people use Hotmail than Domino/Notes. BTW, here's another link showing just how unpopular Notes is, lagging waaaay behind Hotmail! http://litmus.com/resources/email-client-stats. And you really won't like this one (http://www.campaignmonitor.com/stats/email-clients/), where the Notes share is too small to make the pie-chart! Is that one of those "great IBM software successes" you were whitterring on about? LOL!

                      "....everyone is trying to re-live the mainframe model. It is now called "cloud." ...." Which completely avoids the point that today's clouds are being built on anything but IBM mainframes. The usual choice seems to be Xeon. If anything should show you that mainframe is toast, it is the cloud push.

    2. TeeCee Gold badge
      Facepalm

      Ok, that's the database sorted[1]. Now there's just the little matter of all the *other* Oracle software........

      [1] For a given value of "sorted" that includes; "That was a shitload of effort and it doesn't work very well..."

      1. Allison Park
        FAIL

        another Sun mistake

        go down that road and get jonathan to give you a purchase price....he likes round numbers with B's behind them

  2. Anonymous Coward
    Anonymous Coward

    the fighting purses are out

    Prada? Vuitton? D&G?

  3. Allison Park

    Itanium is dead

    Interesting how HP was so scared of their ex-CEO working for Oracle they tried to secure support for Itanium. And the "Itanium Collaboration Agreement" obviously has an end date to Intel support which was blacked out but their obviously is one. If we still bought Itanium systems these are the questions I would be asking my HP rep.

    1) How much does HP pay in the "Itanium Collaboration Agreement"

    2) What is the end of support date Intel is committed to in the "Itanium Collaboration Agreement"

    3) what is the minimum performance required of future chips

    4) Is the only "contract" with HP to keep supporting Itanium a reaffirmation agreement which was created because HP fired and Oracle hired Hurd?

    5) Does any contract talk to future products or just support

    6) How do you justify tukwila pricing of 2X montvale when the license performance is not higher

    7) Is the Odessey announcement related to the reality of Itanium?

    8) HP Itanium revenue is in a free fall according to Gartner and IDC, can HP afford to invest?

    9) Itanium Poulson looks like it got moved out to 2013 on the latest roadmap (chip end of 2012, systems 2013)

    10) Does Meg care about Itanium and does Matt B feel betrayed?

    1. Matt Bryant Silver badge
      FAIL

      RE: Itanium is dead

      The only value in Alli's posts are they warn you in advance as to exactly what the schpiel from next IBM rep visit will be! In this case, Alli is actually behind the curve (maybe this new FUDpack came out whilst she was on holiday?) - we've already had almost exactly the same points bleated at us over a fortnight ago. Alli, you need to keep your FUD up-to-date, girl!

      "....1) How much does HP pay in the "Itanium Collaboration Agreement"...." Why? What difference does it make to the customer if they get the product they want? How much does IBM pay to any of their suppliers for any of the compnents in their products? Does Alli suggest you should ask IBM how much they pay for their disk drives, or how much they pay for their RAM, or any other component? No, becuase it's a silly non-argument. If hp contributes to the development of Itanium so what? If us customers get the solution we want at a better price than IBM, then surely that's the "payment" we should worry about. It obviously has IBM very worried.

      "....2) What is the end of support date Intel is committed to in the "Itanium Collaboration Agreement"...." Seeing as it covers two future generation of Itanium minimum, it's a lot longer than IBM's public commitment to Power. People in glass houses.....

      The rest of the post is a bit garbled (Alli, our IBM rep had a different wording which made a lot more grammatical sense - no more technical sense, though - than your version). But the simple answer to all the rambling FUD is the same - the proof is in the products coming to market. HP has two future generations going into socket-compatible servers available now, whereas IBM has only one generation more of Power, with no release date, and with no Pee-servers available to take the new Pee8 (should it ever arrive) without a massive rip'n'replace gouging of their customers.

      ".....does Matt B feel betrayed?..." No. I have now servers in-place that I can upgrade with two future generations of Itanium, which is a lot better than what IBM or Snoreacle can offfer. Alli, if you did go to Barbasos, was it to a Betty Ford clinic there? I think you may need some additional treatment, tbh.

      1. Allison Park

        wow you could only answer 2 questions

        I was in Jamaica for Thanksgiving...but I don't think you have those in England

        1. Matt Bryant Silver badge
          Happy

          RE: wow you could only answer 2 questions

          No, I could only be bothered to answer two rediculous bits of FUD. The rest was simply too comic to waste time on.

          And we're (apparently) a "cosmopolitan" nation, which basically means we nick anyone else's good holidays as an excuse to have a piss-up! As an example, we had our own mini-Oktoberfest not too long ago, much more fun than Thanksgiving. Thanks to our South American employees, in November we celebrated Brazil's Proclamation Day (salsa night!), and already in December we've had a Dominican Republic Discovery Day (hey, it was educational, honest!). Whilst Halloween is fun, until you guys can come up with a more interesting holiday we'll quite happilly ignore Thanksgiving.

      2. Anonymous Coward
        FAIL

        Wow....

        IBM does not publicly announce chip set roadmaps, but, if you do the history, Power's roadmap has been like clockwork (a new chip every 3-4 years) whereas Itanium is synonymous with missed roadmaps and over-promising/under-delivering.

        Per Gartner, IBM Power - Unix now has over 50% of the Unix market. Sun is about 18%, HP-Itanium is about 17% and falling fast. As IBM has been gaining several percentage points of market share every quarter, I think Power 8 will arrive.

        Again, if you do the history, IBM supports their platforms forever (you can still get VMS support). HP has a laundry list of dead-ended platforms. Meg Whitman straight admitted in their last earnings call that Integrity is a "business in decline" and they need to "manage" the business as it ends (i.e. migrate as many customers as possible to x86-Linux).

        The writing is on the wall. IBM will own Unix in five years (albeit a smaller Unix market). As with mainframe and mini-computers (AS/400), IBM is the only company with the market share and cost base to continue.

        1. Matt Bryant Silver badge
          FAIL

          RE: Wow....

          "IBM does not publicly announce chip set roadmaps...." Erm, yes they do! You may want to read here http://www.theregister.co.uk/2011/08/31/ibm_power_chip_roadmap_update/ for an example.

          "....Power's roadmap has been like clockwork...." Laughable! So, what happened with the Pee7, due in 2008, didn't arrive until 2010....? Or the Pee6+, due 2007, never actually seemed to show. And the Pee6 was six-months late on the ADJUSTED IBM release date! Please do carry on spouting such easily disproven "facts", they only make you look all the less informed.

          ".....IBM Power - Unix now has over 50% of the Unix market...." I think you're getting your reports mixed up. No-one disputes that the current largest number of installed UNIX systems goes to Slowaris, due to the large number of customers that bought Sun SPARC servers and are now stuck with them.

          "....IBM supports their platforms forever...." LOL! Please tell that to Lenovo, which is selling all the IBM x86 lines that Big Blue couldn't make any money on. You may also want to go read up on IBM's efforts with Monteray, their half-hearted attempt to get AIX running on Itanium, and then the IBM customers that bought IBM x450 Itanium servers (yes, IBM tried to sell Itanium servers, they actually sold 10,000 units!) before IBM left them adrift.

          "....Meg Whitman straight admitted in their last earnings call that Integrity is a "business in decline"...." Surprise, the whole UNIX segment is in decline. Each year the UNIX vendors are fighting over a smaller and smaller pie. The reason is x64 servers eating the market from below. Now, guess who is the leading x64 server vendor - yes, hp!

          "....The writing is on the wall...." You only seem capable of reading a tiny bit of the writing. Fact - hp and Intel have a longer public roadmap for Itanium than IBM does for Power. Fact - IBM's mainframe market and revenue is also shrinking, meaning less money to prop up the Power-AIX franchise. Would you like me to write it in extra large letters on your wall, to help you read it?

          /SP&L

          1. Anonymous Coward
            Stop

            Wow - Part 2

            "IBM does not publicly announce chip set roadmaps...." Erm, yes they do! You may want to read here http://www.theregister.co.uk/2011/08/31/ibm_power_chip_roadmap_update/ for an example."

            - That site looks like an El Reg URL, not an IBM URL. IBM does not announce GA dates for their chips until they are ready to GA. Are you honestly comparing the delivery of Power with the delivery of Itanium? I am not aware of a single user that has been wringing their hands waiting for Power. Itanium delays are measured in years. When they come out, a la Tukawila with doubling the capacity and doubling the size of the die (i.e. no performance improvement), they are laughable. IBM Power has consistently released chips with meaning performance enhancements and I am not aware of a single IT manager that would say they cannot receive Power upgrades at a schedule which meets their performance requirements.

            From the NYT, "Intel and Hewlett-Packard promised the first version of Itanium way back in 1999, only to delay that chip and ship it a couple of years later, kicking off a string of delays and performance issues that turned the product into a joke in the chip industry. The chip even earned a nickname of the “Itanic.”

            http://bits.blogs.nytimes.com/2009/02/09/ten-years-after-first-delay-intels-itanium-is-still-late/

            "I think you're getting your reports mixed up. No-one disputes that the current largest number of installed UNIX systems goes to Slowaris, due to the large number of customers that bought Sun SPARC servers and are now stuck with them."

            - I think you are getting your reports mixed up. No one denies that in the 1990s Sun was the primary Unix supplier. The question is, what are people buying in 2009, 2010, 2011, 2012? Not what do they have on the floor that they are trying to get rid of at the moment. It has overwhelmingly been Power in the past few years with a level of Unix dominance that mirrors Oracle's lead in DBs. The only difference is that Power is picking up more market share every single quarter. Also, you cannot compare a Power server with a Sun server in a one-to-one ratio. Sun primarily ships small network servers. IBM ships giant DB servers.

            "Please tell that to Lenovo, which is selling all the IBM x86 lines that Big Blue couldn't make any money on. You may also want to go read up on IBM's efforts with Monteray, their half-hearted attempt to get AIX running on Itanium, and then the IBM customers that bought IBM x450 Itanium servers (yes, IBM tried to sell Itanium servers, they actually sold 10,000 units!) before IBM left them adrift."

            - Platforms, kimosabe, platforms. A platform is an operating system. IBM, like the rest of the industry, bought into the Itanium hype. Like the rest of the industry except HP, they moved away years ago when it became clear that Itanium was going down. Those AIX users on Itanic were ported to a viable chip set, Power, and largely still use AIX today.... What does Lenovo have to do with anything? That isn't a platform at all. You can buy Windows on x86 today.

            "Surprise, the whole UNIX segment is in decline. Each year the UNIX vendors are fighting over a smaller and smaller pie. "

            - While it is true that Unix is declining, this is what Meg Whitman said on the earnings call:

            "The BCS business is a declining business. It is a slow decline, but I don't think you're going to see an accelerating growth rate in that business," she said. "And so we just have to manage that as best we can and invest in R&D so we get to a new platform as fast as we possibly can that allows us to service the clients that need this kind of power."

            BCS, not Unix (actually IBM's Unix business grew at 27% last quarter while HP's was down 23%). Also, note that she didn't say anything about Itanium replacement for HP-UX. They are getting users to a "new platform" (i.e. Linux) and HP-UX goes away.

            On x86, yes, it is true that HP is the leader. It is also true that they lead because they ship a ton of little x86 Windows servers. IBM leads in four-socket plus servers. HP's x86 RAS is way behind IBM's RAS. I would never put an Oracle DB on HP x86.

            http://www.theregister.co.uk/2011/11/22/hp_to_make_x86_superdomes/

            "Fact - hp and Intel have a longer public roadmap for Itanium than IBM does for Power."

            Meaningless when HP and Intel have never delivered a single chip on time in the history of Itanic. Also, HP needs to reassure customers that think Itanium is going away. No one has any concerns about Power, so IBM doesn't have to get the marketing folks working on an elaborate public roadmap.

            "Fact - IBM's mainframe market and revenue is also shrinking, meaning less money to prop up the Power-AIX franchise."

            That is just not true. Wall Street Journal in July, "revenue from the high-end machines known as mainframes surged 61% in the second quarter, capping the best four quarters of growth for the segment in five years." Two, AIX is taking over the Unix market (albeit a smaller Unix market), I don't think it needs to be propped up by mainframe, Power also has the large and profitable System i install base and their OEM chip business (Wii, PS3, XBOX, medical equipment all run on Power).

            http://blogs.wsj.com/digits/2011/07/28/behind-the-youthful-sales-surge-for-ibm-mainframes/

            1. Matt Bryant Silver badge
              FAIL

              RE: Wow - Part 2

              ".....That site looks like an El Reg URL...." Wow, so you can actually read stuff not in an IBM brochure! You just don't seem able to take it in unless it's coming from Big Blue. Which bit of the roadmap in the article do you contest? Oh, you can't hence the evasion by saying "it's Reg, therefore it doesn't count". Try again!

              "....IBM does not announce GA dates for their chips until they are ready to GA...." Really? Because, as an IBM customer, I can call that complete male bovine manure! We saw THREE dates for Pee6 before it finally got announced, and then didn't see actual systems for a while after that.

              ".....I think you are getting your reports mixed up...." Really? Go read back through your own previous post, you said: ".....Per Gartner, IBM Power - Unix now has over 50% of the Unix market....", implying IBM had 50% of the UNIX base out there. Now you're doing more wriggling evasion, spinning like a top and backtracking to sales figures instead. Make your mind up.

              "......bought into the Itanium hype...." LOL! That was 10,000+ instances where IBM tried to sell first a mainframe, then a Pee-series AIX server, then had to sell an Itanium box because that was what the customer asked for. The figures I've seen say 95+% of those IBM Itanium servers didn't have AIX on them, and the majority of those servers were swapped out for other vendors' servers running non-IBM OSs. Not surprising seeing as the Itanium version of AIX 5 was completely incompatible with the Pee-series version. But then you usually have to upgrade to a new version of AIX with each Power generation to actually gain any advantage over the previous Power generation, so I'm not surprised IBM would have trouble migrating between completely different chip families.

              I also love all the US-centric figure you spout, completely ignoring the fact that the Asia market is currently the only market with enough growth to drag the under-performing US and European sales into a Worldwide positive figure. And 98% of the total server shipment last year across the Asia market were x64, not Power at all (http://www.theregister.co.uk/2011/09/07/idc_gartner_server_spending_analysis/). In fact, IBM was lucky that the recent round of cyclic mainframe refreshes bumped their figures up, but it will be interesting to see where the growth comes from next year now that mainframe refresh is just about over.

              I also like how you take a Meg comment on the general decline in the enterprise market and make it into something completely different! Where does she mention Linux? Please at least try and keep one foot in reality, it might help.

              Strange how you also declare the Intel/hp roadmaps "meaningless" because they have had slippage in the past, but deny IBM has ever slipped! I used to joke about the Sunshiners and their Sunshine blinkers, but you must have been in a whole alternate reality for the last ten years if you really want to pretend that IBM has never slipped! What a tool! Us customers want roadmaps because it gives us something to plan around and confidence in a vendor. You may recall (if it penetrated into your alternate dimension) that one of the criticisms of Oracle was they were slow to bring out a roadmap for the server bizz they inherited from Sun, because all that delay led to old Sun customers getting worried and investing in other vendors' servers instead.

              And can we please ditch the pretence the chips in the Wii, PS3 or any other console have even the slightest link with the Power CPUs. IBM could kill Power tomorrow and still go on making OEM chips, so they do not guarantee Power's survival in the slightest. And before you try claiming they bring in mountains of cash, please go look at the price of a games console and then compare it to a single-socket server. Even a 1U x64 box is massively more expensive, and offers support options (does IBM provide support and gain revenue from PS3? - I don't think so). In short, IBM's OEM bizz is not going to prop up Power.

              /SP&L

              1. Anonymous Coward
                Stop

                "it's Reg, therefore it doesn't count" and "....IBM does not announce GA dates for their chips until they are ready to GA...." Really? Because, as an IBM customer, I can call that complete male bovine manure! We saw THREE dates for Pee6 before it finally got announced, and then didn't see actual systems for a while after that."

                - Yes, when I am looking for an IBM roadmap, I would like it to be from IBM not an unaffiliated internet site. If it is as public as you suggest, there should be no trouble finding it on their website. It isn't public. If an IBM rep told you three dates and Power was not out by those three dates (unlikely story), they must have been making it up because it surely did not come from development. Lets go through it again: Power 4 - 2001, Power 5 - 2004, Power 6 - 2007, Power 7 - 2010. See the pattern? Every three years a meaningful release.

                ".....I think you are getting your reports mixed up...." Really? Go read back through your own previous post, you said: ".....Per Gartner, IBM Power - Unix now has over 50% of the Unix market....", implying IBM had 50% of the UNIX base out there.

                - A market would imply a market. Are those legacy Sun server for sale? No. The Unix market (i.e. the sale of new servers) is owned by IBM.

                "That was 10,000+ instances where IBM tried to sell first a mainframe, then a Pee-series AIX server, then had to sell an Itanium box because that was what the customer asked for. "

                - I don't know where you are getting those numbers, but it is possible. HP and Intel really whipped up quite a furry over Itanium ten years ago. It was going to be the chip in everyone's PC (I guess they are still working on that). It wouldn't surprise me if customers asked for Itanium because they thought HP and Intel knew what they were talking about (fail).... An interesting exercise, as you are apparently familiar with the buying process at each of these 10,000 plus users, would be to ask those users if they wish they would have chosen System p if they knew then what they know today.

                "And 98% of the total server shipment last year across the Asia market were x64, not Power at all (http://www.theregister.co.uk/2011/09/07/idc_gartner_server_spending_analysis/). In fact, IBM was lucky that the recent round of cyclic mainframe refreshes bumped their figures up, but it will be interesting to see where the growth comes from next year now that mainframe refresh is just about over."

                - Ahh, the standard "there are many more x86 servers shipped than Unix/mainframe, therefore x86 is the only path." Well, 98% of the vehicles in the world are cars, but I think they will still make planes. Apples and potatoes. People have been predicting the end of mainframe forever, but it is still the only platform in the world for applications that require 100% uptime.

                "I also like how you take a Meg comment on the general decline in the enterprise market and make it into something completely different! Where does she mention Linux? Please at least try and keep one foot in reality, it might help."

                - I agree that it would definitely have been advisable for her to say "Unix", but she said BCS. That probably has something to do with the fact that IBM's Power business is booming (at the end of a chip cycle, so little refresh revenue). She doesn't say Linux, but it is obvious that she means Linux and Microsoft.... Good point, maybe by get them to a "new platform" she means they are bringing back Alpha boxes and VMS. It is HP, so it wouldn't be the strangest announcement they made this year. Maybe a "new platform" is the secret HP plan for webOS. Put all of those Superdome users on TouchPad, a little performance boost?

                "Strange how you also declare the Intel/hp roadmaps "meaningless" because they have had slippage in the past, but deny IBM has ever slipped!"

                - See above three year cadence.

                "What a tool! Us customers want roadmaps because it gives us something to plan around and confidence in a vendor."

                - You know what is better than a meaningless roadmap that a vendor never, ever lives up to, a provider that you don't have investigate to make sure they are doing what they should be doing. There is no need for customers to interrogate IBM because they have a history of reliability and trustworthiness. You, my friend, are an HP tool... not the first, but maybe the last.

                "You may recall (if it penetrated into your alternate dimension) that one of the criticisms of Oracle was they were slow to bring out a roadmap for the server bizz they inherited from Sun, because all that delay led to old Sun customers getting worried and investing in other vendors' servers instead."

                - Again, Apples and potatoes. Sun was a company that was purchased in the red by Oracle, a company with no hardware background. There was legitimate reason for concern. Not so with IBM.

                "In short, IBM's OEM bizz is not going to prop up Power."

                - Not independently, but 1) IBM has a billion plus year business selling IP to other companies. 2) It is not just gaming counsels (although that is a sizable piece of revenue), it is MRI machines, Sonogram, etc. 3) IBM sells this stuff which HP recently found out about called software around their servers. That attach rate is very high for AIX servers. 4) IBM will have the Unix market sewn up in a few years. True, many will go to x86, but those who don't will go to Power as it will be the last Unix system standing. More than enough to keep Power viable.

                1. Matt Bryant Silver badge
                  FAIL

                  RE: Wunderburp1

                  "....when I am looking for an IBM roadmap, I would like it to be from IBM not an unaffiliated internet site...." More evasion! Please explain how it is not an IBM roadmap, or is not valid? It looks just like the one I've been shown by IBM reps. And given the amount of airtime TPM gives IBM, I wouldn't be bashing the Reg if I were you!

                  "....If an IBM rep told you three dates and Power was not out by those three dates (unlikely story)..." Jokes on you! I'd even send you the IBM emails and slides with the different dates on, but I'm afraid I'd be in breach of the IBM NDA we signed before they gave us each version. I still note that you can't dispute that IBM has had product slips, you just try to deny they exist.

                  "....Well, 98% of the vehicles in the world are cars, but I think they will still make planes..." More evasion. IBM don't make cars or planes, so you're the one trying to compare apples and oranges. Trying to ignore the market trends because they simply didn't go with the company strategy is what killed Sun. Offering systems that can mix x64 and a UNIX-capable CPU (like Itanium) in the same chassis is going to be a lot better play in the x64-dominated market than IBM's insistance on trying to ignore x64. Face it, hp have outmanouvered IBM in the x64 market for years, and adding Xeon into Superdome2 will just make it easier for hp to sell into IBM accounts where IBM will not be able to match the offering.

                  "....but it is still the only platform in the world for applications that require 100% uptime...." Yeah, because IBM mainframes never fail? Complete male bovine manure! And hp NonStop also offers 100% availability anyway (at a lower price point), as well as comparable mainframe products from non-IBM vendors. Heck, even Unisys have been offering "100% availability" systems using Xeons for years! Your viewpoint is so cripplingly IBM-centric it's laughable.

                  "....She doesn't say Linux, but it is obvious that she means Linux and Microsoft...." No, you assume again. Unless you have the ability to read minds you're just barking up the tree you want it to be. And, no, voices in your head don't count.

                  "....You know what is better than a meaningless roadmap that a vendor never, ever lives up to...." So, which roadmap would that be? The IBM one promising the non-existant Pee6+, maybe? Strange, if roadmaps mean so little, why do all the vendors' reps (including IBM's) seem to have them, and why do all us customers ask to see them? Oh, I see, they only count when they back up IBM trolls' arguments.... You really need to get out from under that bridge a lot more often and actually get to talk to some of us customers, it would help a lot with your lack of perspective.

                  "....Sun was a company that was purchased in the red by Oracle, a company with no hardware background. There was legitimate reason for concern. Not so with IBM...." Sun was killed by a lack of customer confidence, not by a lack of product. Strange, you can cast aspersions based on wild assumptions about anyone else, but just not IBM? What otherworldly, godly power do you assume IBM have that makes them different from other vendors? If Oracle was unable to tell me when their next CPU was going to be available I'd be worried, but if they couldn't give me any details on what was coming after then I'd be seriously worried, and you'd probably be trolling long and loud about it. But that's exactly what the position is with IBM's Power line. Please try and deny it. Meamhwile, I can buy Tukzilla servers from hp now, and hp has the Intel roadmap showing that I can plug in a further two generations of Itaniums and have the future option to mix Xeons and Itaniums in the same frame, a flexibnility not even mentioned int he most ambitious IBM roadmap. Now, applying your low level of proof required to make wild assumptions, I could quite comfortable say Power is dead. Well, maybe if I had those voices in my head that you seem to have.

                  1. Anonymous Coward
                    WTF?

                    "More evasion! Please explain how it is not an IBM roadmap, or is not valid? It looks just like the one I've been shown by IBM reps. And given the amount of airtime TPM gives IBM, I wouldn't be bashing the Reg if I were you!"

                    - 1) Not bashing El Reg. I just pointing out that IBM and Reg are separate entities that do not speak for each. 2) This is a nuanced point so I don't except you to understand it.

                    "Jokes on you! I'd even send you the IBM emails and slides with the different dates on, but I'm afraid I'd be in breach of the IBM NDA we signed before they gave us each version. I still note that you can't dispute that IBM has had product slips, you just try to deny they exist."

                    - Yes, the joke is indeed on me. You have documents, which we can't see, but if we could see them, all would become clear. Man, you do sound like HP. Power 4 - 2001, Power 5 - 2004, Power 6 - 2007, Power 7 - 2010.... No slippage, undeniable for anyone able to look at it objectively.

                    "Yeah, because IBM mainframes never fail? Complete male bovine manure! And hp NonStop also offers 100% availability anyway (at a lower price point), as well as comparable mainframe products from non-IBM vendors. Heck, even Unisys have been offering "100% availability" systems using Xeons for years! Your viewpoint is so cripplingly IBM-centric it's laughable."

                    - Wow, I am not talking about the marketing. Nearly every bank, insurer, government, etc. in the world runs System z for mission critical. Not Turtledome, not Unisys (if they still make hardware), Tandem/Compaq/HP NonStop (if they still make them). IBM System z has the highest meantime between failure of any platform, per Gartner, IDC, Forrester. IBM System z is the only EAL5 level secure platform in the world.

                    ".She doesn't say Linux, but it is obvious that she means Linux and Microsoft...." No, you assume again. Unless you have the ability to read minds you're just barking up the tree you want it to be. And, no, voices in your head don't count."

                    - Alright then, she doesn't mean Linux and Microsoft. When she says that they need to get users to a "new platform" as quickly as possible, which platform is she talking about? The point is moot because it will definitely be a migration away from Itanic and HP-UX/old OSs.

                    "Strange, if roadmaps mean so little, why do all the vendors' reps (including IBM's) seem to have them, and why do all us customers ask to see them?"

                    - I seriously doubt you are a customer. Customers don't care this much about their Unix platforms. You know thought those DB issues with MS SQL were minor, so probably not daily hands on in IT. I am guessing you are in sales/pre-sales..... Already, asked and answered the roadmap question.

                    "not by a lack of product. Strange, you can cast aspersions based on wild assumptions about anyone else, but just not IBM? What otherworldly, godly power do you assume IBM have that makes them different from other vendors? If Oracle was unable to tell me when their next CPU was going to be available I'd be worried, but if they couldn't give me any details on what was coming after then I'd be seriously worried, and you'd probably be trolling long and loud about it. But that's exactly what the position is with IBM's Power line. Please try and deny it. Meamhwile, I can buy Tukzilla servers from hp now, and hp has the Intel roadmap showing that I can plug in a further two generations of Itaniums and have the future option to mix Xeons and Itaniums"

                    - Noise. IBM is the market leader by a wide margin and they have the best financial situation. HP changes their business model on a weekly basis, is $24 billion in the hole with long term debt, Itanium sales are crashing. Sun Sparc server sales were/are crashing, poor financials. People are not concerned about Power for the same reason they are not concerned about the next version of Cisco's switches. They are the market leader.

                    1. Matt Bryant Silver badge
                      FAIL

                      ".....1) Not bashing El Reg. I just pointing out that IBM and Reg are separate entities that do not speak for each. 2) This is a nuanced point so I don't except you to understand it...." So, just more evasion of the question why is it not a valid roadmap? Could it be because it pokes a big hole in your drivel? Don't tell me, it's "meaningless" or "laughable", right? There's a word for that - denial!

                      In fact, your pointless denial of everything that demonstrates your ignorance is getting very boring. Howabaout we just concentrate on one argument at a time to give it a chance to penetrate the mainframe world - where is the Power6+? If IBM never has slippage, and they bring out chips like "clockwork", what happened to the Pee6+? I await your evasive, denial-packed dribblings with unseasonal glee.

          2. Allison Park
            Angel

            wow on you

            yes IBM updated their roadmap. while IBM did not want to release too much data about future chips so intel copies them with x86 they also wanted to show that poulson is not as good as p7, kittson not as good as p7+ and since there will not be an Itanium chip after Kittson it is obvious p8 continues the leadership.

            Clockworks....new generations have been every 3 years and a + between 18-24 months. Itanium has been years late and will never be competitive before it is discontinued

            Yes IBM saw the writing on the wall that Itanium was dead when Merced became just a development chip.

            Unix actually grew 2% so far this year and if you look at the US market IBM is now 55% vs. Itaniums 15%. Revenue is what matter because that is who vendors invest. Oracle is getting out of the "shipment" business asap.

            Look at IBM's recent earnings reports websphere and Power have been the stars.

            1. Matt Bryant Silver badge
              FAIL

              RE: wow on you

              "yes IBM updated their roadmap...." You IBM Elmers need to talk more, Wunderburp was insisting that there was no IBM roadmap! Did IBM employ all those laughable Sun marketting bods? That would explain a lot.

              ".....IBM did not want to release too much data about future chips so intel copies them...." Que? So IBM ditching the in-order execution of Pee6 to go back to out-of-order execution for Pee7 (as Xeon has) was Intel copying IBM, and not IBM copying Intel....? Yeah, right! It's also widely accepted that IBM rushed the announcement of Pee7 to try and defuse the news of Intel's Nehalem Xeons and Tukwila Itaniums. That was obvious from the long period between the Pee7 "launch" and the actual availability of systems.

              ".....Clockworks....new generations have been every 3 years and a + between 18-24 months...." Again, you IBM Elmers need to talk, you're both insisting on different periods between upgrades. I'm also curious as to which bit of "clockwork" includes the completely missing Pee6+, which never arrived! Strange how you IBMers love to gloss over that gaping hole in your argument.

              "......Yes IBM saw the writing on the wall that Itanium was dead when Merced became just a development chip....." Actually, what IBM (and Sun) saw was that staying on Itanium meant competing on a level playing field with Microsoft and hp, and neither fancied that. In IBM's case, it also meant less commonality with their mainframe bizz. Sun gambled on UltraSPARC and eventually died. IBM used their mainframe muscle to prop up the aging RISC design and extend Power. The problem for Power is the design is hitting the buffers, hence the switch back to out-of-order execution, die shrinks, packing more cores onto the die, and clockspeed ramping. There is little room left in the design for any form of real innovation. Which also explains the lack of anything after the vague Pee8 on the IBM public roadmap - IBM are running out of options.

    2. Anonymous Coward
      Anonymous Coward

      So true. Itanium is dead and we all know it. Oracle is revealing the dirty play behind the scenes and that is good for us customers so we can make long term decisions. Itanium is getting artificial respiration from HP, while HP customers are migrating to IBM POWER7 servers, which do have a future. Last I heard, IBM migrated 425 new HP customers to POWER7 servers. The reason? HP-UX was too unstable and lacked performance! Too expensive when you consider the measly performance you get.

      1. Ramazan
        Facepalm

        @AC 13:04 GMT

        HP-UX too unstable and lacked performance? It depends on what you mean by "unstable" and "performance", much like Bill's rant about meaning of word "is" was. :) Certainly, HP-UX is not even close to Linux or Solaris feature-wise, but it's rock solid stable. And not more ugly than AIX is (was, and will ever be. The single most bastardized version of UNIX is AIX without any doubts).

  4. Adam 61
    Facepalm

    So true my backside last AC

    If theres one thing HP-UX isn't its unstable, and performance is pretty good. P does have an advantage in pure performance terms but its not as big as they imply on there marketing slides.

    Itanium probably is on its last legs but not because the reasons you state - purely down to Oracles play pulling the rug out from under it.

    I say this as someone who works for a company who sells both IBM and HP, pre- the Oracle announcement the new Tukwila systems were building nicely especially for SAP based solutions. SAP and HP-UX has long been a trusted platform, and you don't expect people to run a core business app like that on something thats unstable, anything but is reality.

    1. Ramazan
      Holmes

      Itanic sinks

      "purely down to Oracles play pulling the rug out from under it". Wrong. Intel is happy selling lots of Xeons while having marginal sales of Itanics, why should it pay salaries and buy equipment for the 2nd unprofitable R&D team then?

      1. Matt Bryant Silver badge
        FAIL

        RE: Itanic sinks

        Same old FUD as ever!

        "....unprofitable R&D team...." Your FUD is easily debunked in two simple steps. Firstly, please provide a public document from Intel that shows Itanium is not profitable for them. You can't, beacuse Itanium makes Intel money. Then consider that if hp is also contributing to Itanium development, Intel probably has plenty of money to spend on R&D, especially considering the massive profits off the Xeon line.

        And, since you won't want it mentioned, maybe you'd like me to remind the readers that developments from the Itanium line are fed into the Xeon line, so Itanium R&D also benefits Xeon, which is Intel's real moneymaker. Consider your weak FUD completely debunked.

        1. Anonymous Coward
          Anonymous Coward

          No one can give you the Itanium financials because Intel (purposely) does not report them. It would be an embarrassment for Intel because it would demonstrate a large short fall in operations subsidized by pay-to-play cash from HP. A high-end Xeon (Nehalem EX) processor has a higher list price and a higher contributing margin than an Itanium chip. Look at the price sheets.

          Yes, Intel brought all of the RAS from Itanium to Xeon in Nehalem EX, but that was not an indication that Intel is going to pour R&D into the Itanium business. It was an indication that Itanium is no longer viable as a "high-end" alternative to Xeon and Intel is bringing any reason a person would purchase Itanium to their real business, Xeon. Even Intel admits that Xeon will outperform Itanium and is equally as reliable. Kirk Skaugen has said that the only reason to purchase Itanium is if you want HP-UX/legacy HP OSs.

          1. Matt Bryant Silver badge
            FAIL

            RE: Wunderbar1

            "No one can give you the Itanium financials...." So, what you're finally admitting is that all you IBMers that claim "Intel are losing money on Itanium" actually are talking out there backsides. Being economic with the truth. Sprwading FUD. That they are lying. Thanks for clearing up that point. Strangely, Intel does give out details on earning at the investor meetings, so if there was any "itanium-makes-a-loss" truth then it would have come out already.

            "....A high-end Xeon (Nehalem EX) processor has a higher list price and a higher contributing margin than an Itanium chip. Look at the price sheets....." Again, please do provide the facts to back up that claim that a Nehalem contributes more margin, otherwise I'll just have to conclude that you are just lying again. Truth is you have no insight into the financials of the different processes used for the two CPU families and no idea which costs more. For all you know, it could be twice as expensive to produce the 32nm Westmere-EX compared to the 65nm Tukwila. In short, you are once again making wild assumptions and trying to pass them off as facts. Which is more lying.

            "....but that was not an indication that Intel is going to pour R&D into the Itanium business...." But you have failed in showing that Intel won't. For all you know, they could be putting twice as much into Itanium development as Xeon. In short, once again, you have no facts to back up your assumptions, but that doesn't stop you passing them off as facts. More lying fail.

            Oh, and quoting Kirk Skaugen from his role as head of Xeon is a bit stupid. Just ask Larry Ellison what happens when you make business plans based on wild assumptions about what Kirk said. It's landed Larry in court.

            /SP&L

            1. Anonymous Coward
              Anonymous Coward

              If you are arguing that Intel makes it as difficult as possible to understand the financial and development activities around Itanium, I agree.

              "Strangely, Intel does give out details on earning at the investor meetings, so if there was any "itanium-makes-a-loss" truth then it would have come out already."

              1) If they give detailed financial on Itanium, I have never seen them. I assume you haven't either or they would be posted. 2) Intel investors care more about the pastry selection at the investor meeting than Itanium. It is such a miniscule part of Intel's business that it doesn't even hit their radar. I am sure they assume it is going poorly, but who cares because Xeon is going well. It is a rounding error for Intel. Not so much for HP.

              "...but that was not an indication that Intel is going to pour R&D into the Itanium business...." But you have failed in showing that Intel won't."

              - Here you go:

              http://www.xbitlabs.com/news/cpu/display/20110417112218_Intel_Relocates_Itanium_Engineers_to_Xeon_Projects_Sources.html

              "Oh, and quoting Kirk Skaugen from his role as head of Xeon is a bit stupid. Just ask Larry Ellison what happens when you make business plans based on wild assumptions about what Kirk said. It's landed Larry in court."

              - It is understandable that you think Skaugen just has Xeon, but he is actually the GM of their data center products including Itanium..... Larry loves being in court. They are actually countersuing HP for fraud at the moment on the Itanium issue. HP vs. Oracle in a fight.... I like Oracle to win that one.

              1. Matt Bryant Silver badge
                FAIL

                RE: Wunderbar1

                "If you are arguing that Intel makes it as difficult as possible to understand the financial and development activities around Itanium, I agree...." Nope, you said straight out that Intel loses money on Itanium, but when challenged on it you cannot back it up with any proof. In short, you lied, and you got called out on it. Your evasions are as weak as the rest of your schpiel.

                "....I am sure they assume it is going poorly..." Again, more assumption dressed up as fact. Or did those voices in your head give you the idea? So, once again we find that you have no grounds for the argument you are pushing, that it is all just based on what you want to believe, what you want to happen. Yeah, I know some people that used to think like that, they worked for a company called Sun....

                "....Here you go...." So, all you do is point to a source that says Itanium engineers that have finished on Tukwila were moved to Xeon projects to help bring in next gen Nehalem in on time. Simply sounds like good project management to me, moving a resource from one project to another to keep them busy. This also pokes another hole in your non-argument - you said Intel wasn't developing Itanium further, but here we have proof of not just a Poulsen team but a Kittson one as well! Strange, didn't you tell us Intel wasn't going to deliver any future Itaniums, that their roadmap was "meaningless"? Golly, if CPUs appear on roadmaps and actually have development teams, what does that say for CPUs that don't appear on roadmaps, it must mean they don't have development teams! Doesn't look good for Pee9 then. More epic fail from your non-arguments. Please try producing sources that at least vaguely back up your schpiel. Unless you're going for comedy, that is.

                "....but he is actually the GM of their data center products including Itanium...." Not at the time he made the comments you are selectively quoting, which were at an Inetl Xeon event. Oops, did that little bit of context just poke another big hole in your waffle? Shame, try again, little troll.

                ".....I like Oracle to win that one." Understandable that you would want Oracle to damage a competitor that IBM has been unable to beat in the market. Of course, it completely ignores the fact (and there seems to be a lot of both ignoring and general ignorance in your posts) that there is nothing stopping Larry doing the same to Power tomorrow. And, going by those database market figures, Power-Oracle customers would be more likely to switch to another vendor than suffer DB2.

                /SP&L

                1. Anonymous Coward
                  FAIL

                  "Again, more assumption dressed up as fact. Or did those voices in your head give you the idea?"

                  If you are an investor in Intel, where Itanium is 1-2% of the company's revenues, I think it is fair to say that it is not your primary concern. Also, if you are an Intel investor, there is no way to call Intel out Itanium because there are no financials.... It is like saying that if Microsoft is losing money on Songsmith, surely an investor would have pointed it out. Why would a Microsoft investor care about Songsmith? It is not large enough to make any meaningful difference whatsoever in the valuation of the company. Like Itanium, it is a failed venture, but so what, it is not an important business for Microsoft as Itanium is not an important business for Intel.

                  "So, all you do is point to a source that says Itanium engineers that have finished on Tukwila were moved to Xeon projects to help bring in next gen Nehalem in on time."

                  Yes, you keep believing that Intel is just reassigning resources only to have the entire company working on that critical Kittson release (they could literally sell hundreds of those chips). Intel is permanently reassigning resources to Xeon because Itanium is dead. Intel doesn't care if Itanium dies or not. It is a meaninglessly small piece of revenue.

                  "you said Intel wasn't developing Itanium further, but here we have proof of not just a Poulsen team but a Kittson one as well! Strange, didn't you tell us Intel wasn't going to deliver any future Itaniums, that their roadmap was "meaningless"?"

                  - No, that is not what I wrote at all. Their roadmaps are meaningless because Intel has never, ever delivered an Itanium chip anywhere near on-time in the history of Itanium. As long as HP pays for it, I am sure they will push out some half baked Poulson and Kittson chips with hand me down R&D from Xeon.

                  "Not at the time he made the comments you are selectively quoting, which were at an Inetl Xeon event. Oops, did that little bit of context just poke another big hole in your waffle? Shame, try again, little troll."

                  - Skaugen was the GM for Data Center Products when he made that comment.... Haha, lets assume you are correct, that Skaugen (Xeon bigot) made this outrageous statement that Xeon outperforms Itanium and the only reason to buy Itanium is if you are stuck with legacy HP OSs. In your scenario, Intel was soooo upset about their Xeon chief belittling Itanium that they subsequently put him in charge of all Itanium development. Epic fail!

                  "Understandable that you would want Oracle to damage a competitor that IBM has been unable to beat in the market. Of course, it completely ignores the fact (and there seems to be a lot of both ignoring and general ignorance in your posts) that there is nothing stopping Larry doing the same to Power tomorrow. And, going by those database market figures, Power-Oracle customers would be more likely to switch to another vendor than suffer DB2."

                  - Unable to beat HP in the market? IBM has beat HP in every market of importance. Storage - IBM, Unix servers - IBM by a mile. Mainframe - IBM by a mile. Software - IBM by a mile. Services - IBM by a mile. x86 Severs - HP in total revenue, but IBM in the high-end x86 servers that produce profit. Congratulations, HP is beating IBM in the zero technology value add low-end x86 server market. Lets not forget PCs, if HP sells PCs this week.

                  - Yes, Oracle could drop Power, but they won't. 1) Because Power is taking over the Unix market. 2) Because IBM is the largest reseller of Oracle software in the world through Global Services. IBM makes Oracle more money than they take out, unlike Highly Precarious (HP). 3) IBM can compete with Oracle in software with the second largest DB business in the world and the number one application server in the world. HP has "no software value add", quoting Ellison. 4) IBM is their largest Java customer and their most important Java roadmap partner by a huge margin. 5) Because Larry generally respects IBM. Quoting Larry, "Oracle is a cheetah, IBM is a stallion, HP is a turtle."

                  1. Matt Bryant Silver badge
                    FAIL

                    ".....where Itanium is 1-2% of the company's revenues....." Hold on a sec - didn't you admit a few posts ago that you don't have any Intel financial figures for Itanium, yet here you are pushing "1-2%"! OK, seeing as it's probably not PC to accuse someone as deep in denial of being a liar, I'll just chalk that one up to those voices in your head. Naughty voices!

                    "....Also, if you are an Intel investor, there is no way to call Intel out Itanium because there are no financials......" oops! Looks liek I'll have to apologise to those voices, you just admitted you're talking male bovine manure! You deny anything that discredits your arguments and then make up figures to try and inflate your claims. Basically, you have zero credibility, you just sound like an over-eager saleman pushing mainframe tin.

                    I particularly enjoyed the "IBM, rah, rah rah" session at the end of the post, all that "we are the biggest, we're too big to fail" blather really reminded me of Sun. Please do explain how any of that would stop Oracle shafting IBM with no future support for Power? Oh, it doesn't! What a suprise that another of your "counters" turns out to be nothing more than bluster and denial.

                    /SP&L

  5. Allison Park
    FAIL

    Itanic sunk

    Itanium would not be profitable for Intel unless HP paid X.X Billion to keep it going and that is really just base mfg cost.

    Intel would rather develop mission critical for xeon than itanium then wait to put it in xeon.

    April 13, 2011 - "We used to position Itanium up here: highest performance, highest reliability .. It is really now a choice of operating system, Xeon's reliability and performance is now equal and in some cases better than Itanium…. So, if you like HP-UX, OpenVMS, HP NonStop, other mainframe operating systems, we are fully going to support you on Itanium,” –

    Kirk Skaugen, vice president of the Intel architecture group and general manager of Intel's data center group, during his keynote at the Intel Developer Forum

    1. Matt Bryant Silver badge
      FAIL

      RE: Itanic sunk

      Alli, did you do the editing fo rthe Wikileaks video? That was the most pointless out-take I've seen for ages, you virtually destroyed your own non-argument with the final line:

      "....So, if you like HP-UX, OpenVMS, HP NonStop, other mainframe operating systems, we are fully going to support you on Itanium...."

      Says it all.

  6. Anonymous Coward
    Anonymous Coward

    If Intel really want a 64 bit future sans the dinosaur of x86...

    ... they could always define a clean easily shrinkable architecture that could be tweaked for good throughput per watt and sold by the bucketful. (They could even add various configurations, say minimal for phones, midi for embedded and maximal for server farms.)

    But what would they call it? How about the Digital Enhanced CPU? Let's see, they could give them a fancy branding, how about Greek letters, that'd make the first one the Alpha?

    I wonder if they ever thought of (or acquired IP to) a concept like that...

    1. Matt Bryant Silver badge
      Facepalm

      RE: If Intel really want a 64 bit future sans the dinosaur of x86...

      So, you're sugegsting they replace one "dinosaur" tech, x86, with an already fossilised tech, RISC. Good thing you're probably not in any position where you could make a mess with a decision like that.

      1. Anonymous Coward
        FAIL

        RISC is not a dinosaur

        Right, those RISC based ARM chips are really struggling these days (read: sarcasm). x86 is based upon CISC. It is an older architecture. Very good for batch, but not so good for power utilization or OLTP.... Why do you think Intel created Itanium in the first place? To get away from a legacy instruction set. AMD started eating their lunch when they went 64 bit with x86, so Intel had to ditch Itanium and EPIC (failure) and scramble to protect its core business.

        1. Matt Bryant Silver badge
          FAIL

          RE: RISC is not a dinosaur

          "....those RISC based ARM chips...." HIlarious, even beyond apples and oranges! Please do supply details of the ARM-based server capable of running enterprise apps to the same level as a an Itanium Superdome2. Oh, you can't. I'm not surprised, given your complete failure to back up any of your other nonsensical posts. I'll go easy on you and just restrict it to one case - please supply details of the ARM-based server that can run Oracle database products. Again, more fail from Wunderburp1.

          ".....Intel had to ditch Itanium...." Intel didn't ditch Itanium at all, in fact they took on the hp Itanium team and the ex-Compaq Alpha team and carried on developing Itanium. Please do try and remember that, just because you fantasise about something happening, it didn't make it a reality.

          1. Anonymous Coward
            Facepalm

            You argued that the RISC *instruction set architecture* is a dinosaur, nothing about servers (even though your company has just announced an ARM based server). You start off discussing one concept and then jump to another when you get stuck... CISC vs. RISC instruction sets... woah, you forgot RISC is actually the fastest growing instruction set on the planet... so lets talk about enterprise servers instead of instruction sets (which RISC also leads via Power, btw).... MS SQL is going to take over the world with a bunch of little workgroup servers... woah, every company and .com is implementing the opposite approach... lets talk about Lotus of all things instead.

            Again, if CISC is the wave of the future (a 40 plus year old instruction set), why was there a need for Itanium in the first place?

            ".....Intel had to ditch Itanium...."

            - They certainly did ditch it from a strategic perspective. If you recall, x86 was going to be completely replaced by Itanium. Every PCs, server, anything that computes was going to run on Itanium when it was announced. Intel ditched that plan and went back to their legacy CISC technology because they needed to defend against AMD 64 bit. Where is my Itanium based PC? Round and round we go.

            1. Matt Bryant Silver badge
              FAIL

              I see we have yet another "counter" where you fail to answer any of the queations raised. All my replies and new points were in answer to points you raised, so don't act all surprised when your evasions get called out. And RISC is only the fastest growing instruction set if you include phones, which ins't going to help your dinosaur mainframes. It also isn't the most common instruction set in computers even including phones - thanks to Intel, that's x86 CISC.

              "....They certainly did ditch it from a strategic perspective...." Seeing as I can bet with certainty that you have no more acces to Intel strategy docs than you do Intel financials, we'll just chalk that up to your overactive imagination. Please don't repeat it again and again like you did with the "Itanium makes a loss" fallacy as I wouldn't want to have to prove you a liar a second time.

    2. Anonymous Coward
      Anonymous Coward

      Shrinkable?

      I seem to remember that the Alphas had much larger dies than the competition.

  7. apleszko

    x86-based Superdome = HP recognizing Itanium is DEAD

    Nobody remembered that HP recently announced that will build a x86 Intel based Superdome besides their Itanium offering!!

    It seems a recognition that x86 is the target platform for Mission Critical workloads. They can't admit it right now just because porting HP-UX AGAIN (remeber PA-RISC to Itanium migration...) would hurt even more their Unix business, and surely wouldn't be an easy and fast task to accomplish. It could even be a sign that HP-UX would be abandoned, given that UNIX market share is dropping and doesn't have great prospects for the future...

    Oracle on its side also doesn't seem to care that much to Solaris and SPARC, focusing on extremely high margin Exadata appliances, based on cheap servers with x86 processors...

    1. Anonymous Coward
      Anonymous Coward

      POWER7

      The sooner Itanic dies the better. Die! Die! Die!

    2. Matt Bryant Silver badge
      FAIL

      RE: x86-based Superdome = HP recognizing Itanium is DEAD

      What you missed is the bit that this is exactly what Intel has been planning and announcing for a very long time - socket-compatibility between Itanium and Xeon, which means Itanium gains even greater cost savings by maximising the investment in Xeon tach, whilst Xeon gains many of the high-availability goodies that have traditionally only been in RISC/EPIC servers. And it will be an industry unique tech.

      What you should be asking is where are the Snoreacle or IBM equivalent 32-socket Xeon servers? If IBM even managed to scale their tech that high, would they be able to offer hardware and software partitioning tech inside that same Xeon frame? Could they do it and offer the same mainframe-class resilience as Superdome2 already has? Will IBM ever make a box that can mix Xeon and Power partitions in the same frame? Don't make me laugh at the idea Snoreacle could do it, they can't even grow Niagara past four sockets without crippling the chip! Why else do you think all those Excretadata appliances are using x64.

      Getting Xeon into Superdome2 is not detracting from the Itanium-based Integrity line, it reinforces it, and benefits hp's Prolaint range by giving it a solution no other top tier x64 vendor can match.

      /SP&L

      1. Allison Park

        RE: x86-based Superdome = HP recognizing Itanium is DEAD →

        What you missed is the bit that this is exactly what Intel has been planning and announcing for a very long time - socket-compatibility between Itanium and Xeon, which means Itanium gains even greater cost savings by maximising the investment in Xeon tach, whilst Xeon gains many of the high-availability goodies that have traditionally only been in RISC/EPIC servers. And it will be an industry unique tech.

        Where is the IBM equivalent 32-socket Xeon servers?

        HP and IBM both have an 8 socket box....and Odessey is not until 2015...first to announce a strategy because Oracle outed the dead of Itanium does not make you a leader.

        Would they be able to offer hardware and software partitioning tech inside that same Xeon frame? Both IBM and HP have blade enclosures with x86 and itanium/power whats your point

        Could they do it and offer the same mainframe-class resilience as Superdome2 already has?

        Superdome2 does not have "mainframe-class" resilience

        Getting Xeon into Superdome2 is not detracting from the Itanium-based Integrity line,....no distraction at all...everyone knows Itanium is dead and HP is quickly moving their whole "Integrity" team to x86.....why the date is so far from no is a mystery...I thought the common socket, same QPI links would make it sooner than a target date of 4 years out

        1. Anonymous Coward
          Anonymous Coward

          True, only mainframe has mainframe class reliability (hence the name) with parallel sysplex technology, EAL5, electrically isolated partitions, redundant everything.

          There is a reason that HP is saying Superdome has mainframe reliability and mainframe is not saying they have Superdome class reliability. If Superdome was up to par with mainframe, they would not have to reference mainframe.

          1. Ramazan
            Coffee/keyboard

            superdome vs mainframe

            HP superdome certainly doesn't posess mainframe-class reliability - 2.5 ago we maxed RAM on PA-RISC superdome, but one memory module was faulty. Guess what? Superdome crashed.

            1. Anonymous Coward
              Happy

              True

              Yes, I think the real up and comer is System z - Linux (IFLs). If you price out a System z with IFL engines, it is not only less expensive than every Unix system on the market, it is less expensive than x86 - Linux. I wouldn't say that System z - Linux is at the same RAS level as z/OS, it doesn't use parallel sysplex in the OS, but it is a far sight more reliable than x86.

              1. Matt Bryant Silver badge
                FAIL

                RE: True

                Would that be before or after you factor in the never-ending Global Screwups charges for the consulting required to implement the Linux on mainframe solution, or the massive IBM support costs of just mainframe regardless of OS? Not to mention the hours of meetings you'd have to waste before IBM actually realise you are serious about Linux. Try again, troll.

                1. Anonymous Coward
                  Stop

                  A troll for writing something positive about a product.... That's a new definition. All of those claims are just ludicrous on the surface. I have implemented Linux on z. It is much less complex than x86 because it is an integrated stack where firmware, hypervisor, OS and tools are all pre-configured and ready to roll. The support costs are not at all massive if you do not use z/OS. They are less than a comparable number of small x86 servers. You can run 100 or even thousands of hosts from one mainframe with 1-2 staffers. You can't even begin to think about running a comparable x86 environment without a team of admins.

                  1. Matt Bryant Silver badge
                    FAIL

                    RE: Wunderbar1

                    ".....I have implemented Linux on z. It is much less complex than x86...." Really? OK, let's compare. I have a ProLiant and a bogstandard RHEL or SuSE DVD. To install them, all I do is put the DVD in the ProLiant drive and off I go. An hour later and I'm in production with the apps I want already pre-tested and downloaded off the Web. If I'm using VMware it's even easier as I can pre-build the image. To do the same on an IBM mainframe? Well, first I have to confirm with IBM what Linux versions are supported with my mainframe, what patches and drivers I need to add, and then "adjust" (which means get charged more money) my mainframe support agremeent to have it amended to include my new Linux partition. At this point I usually find that the Linux app I want to use is actually not available for mainframe (because the majority of Linux devs are working with x64) or that IBM will not support that app on the mainframe unless I go through a support assessment exercise (yes, pay more money again) to have it added. Then I have to roll my own stack, test it, etc. With all the IBM faffing around, this is usually six months after the x64 option has gone into production.

                    ".....You can't even begin to think about running a comparable x86 environment without a team of admins." It is glaringly obvious that you have zero experience outside the IBM mainframe bubble. Please go read up on a little productset called VMware for a start, then hp's iLO and SIM (and IRS) technologies, then realise that today it is very easy for a single admin to manage thousands of Windows and Linux instances on a VMware farm, at a much reduced cost and just as good levels of availability (if not better when using VMwares Vmotion and remote failover capabilities) than IBM's mainframes can manage, and with a much wider range of applications as well.

                    /SP&L

          2. Matt Bryant Silver badge
            FAIL

            RE: Wunderbar1

            ".....There is a reason that HP is saying Superdome has mainframe reliability...." Yes, it's because they are targetting IBM's mainframe customers, you great twonk.

        2. Matt Bryant Silver badge
          FAIL

          RE: RE: x86-based Superdome = HP recognizing Itanium is DEAD

          "....first to announce a strategy because Oracle outed the dead of Itanium does not make you a leader....." Intel have been talking about socket-compatible Xeons and Itaniums for years before Larry went off the deep end. It was an idea that was designed into the QPI tech first built into Xeons in 2009. Once again, Alli, you need to do a bit of background reading. Or is it that you just don't want to admit that IBM does not have a strategy for making a 32-way Xeon server?

          ".....Both IBM and HP have blade enclosures with x86 and itanium/power...." Completely different. The Superdome2 tech includes crossbar switches that allow you to link the individual blades into larger hardware instances - IBM cannot do this with their blades. Their largest x64 blade has only two sockets! The largest Power instance they can offer in their blades is sixteen cores, less than an hp BL890c i2 Itanium blade (32 cores), and much less than the current Itanium Superdome2 (up to 256 cores, which can be split into several hardware configurations). Trying to offer IBM's PS blades (with their crippled version of the Pee7 CPUs) as an alternative to the hardware partitioning flexibility of Superdome2 is just too funny for words! In short, IBM don't have an answer to either the SD2 now or a 32-socket, Xeon-based Superdome. Heck, IBM have been unable to match the Giant Proliant DL980, why should anyone expect them to be able to match a Xeon Superdome2?

          1. Anonymous Coward
            Mushroom

            "Intel have been talking about socket-compatible Xeons and Itaniums for years before Larry went off the deep end. It was an idea that was designed into the QPI tech first built into Xeons in 2009."

            - Hmmm, the timing is interesting. They have had this in place since 2009, but they announced it is going into development it right after their key ISV pulled all support effectively killing Itanium.... Almost as though they had this plan waiting for the "inevitable" end of Itanium so they could milk those large Unix support contracts for as long as possible before pulling the rug out and telling their customers it is time to change chip sets, OSs, etc again.... It seems to me that a large database company came to a similar conclusion.

            - They have been working on this since at least 2009. The socket compatibility is in place. Pretty straight forward. When will these urgently needed systems be on the market HP? Two years. HP is perpetually two years away from everything. They're good at making roadmaps though.

            "Heck, IBM have been unable to match the Giant Proliant DL980, why should anyone expect them to be able to match a Xeon Superdome2?"

            - See IBM System x3950.

            - Match Superdome?... Read any Superdome ad. What does it say? "Mainframe level xxxx!" The TurtleDome is a mainframe knock-off.

            1. Matt Bryant Silver badge
              FAIL

              "....They have had this in place since 2009, but they announced it is going into development it right after their key ISV pulled all support effectively killing Itanium...." Wow, it's amazing how you manage to produce one nonsense post, and then completely out-do the level of nonsense with your very next post! The QPI in Xeon was announced as a product in 2009, it was announced as a technology even earlier! A quite tragic case of major fail, completely undermining your whole post. It would probably be best if you saved what little face you have left and just left the conversation to the grown ups.

              "....The socket compatibility is in place. Pretty straight forward. When will these urgently needed systems be on the market HP? Two years...." Ah, but you neglected to say that QPI is not the backbone of the Superdome2 technology, it is the cross-bar switch backplane. You also forgot to mention that hp, unlike some other vendors, actually does testing before they release a product. Seeing as hp like to offer a 5-nines guarantee with their products, that's a lot of testing, which takes time. You'd know that if you actually knew anything about the industry.

              "....See IBM System x3950...." LOL! So you're trying to match an 8-way Xeon server with two 4-ways cobbled together! You even have to bolt on a memory expander to get above 2TB of RAM, for a max of 3TB, and still don't match the 4TB capability fo the DL980. Would you like to compare benchmark records (it will keep Kebbabfart happy to include some benchmarks, so for him go look at http://h18004.www1.hp.com/products/servers/benchmarks/products.html) between the DL980 and the cobbled-together X3950? I don't think so, for two reasons; firstly, the DL980 canes the IBM x3950; and secondly because IBM are very cautious about showing what even their own 8-way Xeon box can do as it makes their Pee-series look slow and expensive.

              ".....The TurtleDome is a mainframe knock-off." Exactly! Hp are targetting all those custoemrs Superdome2 offers mainframe capabilites with greater performance at a lower price-point, which is why IBM trolls like yourself FUD it at every opportunity. You only have to go read up on Platform Solutions (http://www.channelregister.co.uk/2009/10/20/ibm_mainframe_i_monopoly/) and how IBM eventually bought them out to realise just what a threat to the IBM mainframe bizz the Superdome has always been.

              /SP&L

      2. Anonymous Coward
        Anonymous Coward

        Not really, "Project Odyssey" (apparently no one at HP has read The Odyssey or understands it is the story of an arduous, multi-year journey into the unknown) is HP's plan to make way for the inevitable end of Itanium and hold on to their high-end "Integrity" line of servers by turning them into x86 servers... at some point in 2015, which means 2018 in HP years.

        The real news is that there is no port for HP-UX, OVMS, NonStop off of Itanium to x86, which every HP-UX user in the world (except you, apparently) was banking on. Those OSs will go down with Itanium... at a very rapid rate. As you cannot run MS, Oracle, Red Hat, etc on Itanium, that basically leaves IBM DB2 and WebSphere. HP either has to get their users to migrate to IBM software or get them to x86.

        1. Matt Bryant Silver badge
          FAIL

          RE: Wunderbar1

          Oh dear, I'm beginning to suspect that Wunderbar1 is just another itteration of Kebbabfart, with the same complete lack of knowledge about anything hp

          ".....HP's plan to make way for the inevitable end of Itanium and hold on to their high-end "Integrity" line of servers by turning them into x86 servers...." Clueless! Hp has had a widely reported plan to share as many components between the Itanium and Proliant ranges for at least ten years! The older PA-RISC servers shared very few components with the old hp Xeon Netservers, but hp saw (long before IBM) that they could reduce costs if they shared components between the ranges. For example, if you go look at the old rx26x0 series you'll soon realise it many components with the DL380 servers. Compare this to the completely different hp D-class and old Netservers. This startegy allowed hp to offer their racked servers at much lower prices than either IBM or Sun could and still make margin. Sun's desperate price-matching (AKA dropping their trousers) killed their margin and doomed Sun. The ultimate expression of this phillosophy is the move is putting Xeons into the Superdome2. Selling lots of Xeon Superdome2s will allow hp to reduce the costs of the Itanium versions, making them even more price-competitive against the IBM and Oracle offerings. IBM cannot match this as they cannot share as many components between their remaining x86 lines and their Pee-series. Snoreacle is even worse off than IBM as they have a relatively tiny x86 server bizz to share components with.

          ".....The real news is that there is no port for HP-UX, OVMS, NonStop off of Itanium to x86...." Why is this news? There has never been a plan to port any of those OSs to x86. Can I suggest that since IBM has never announced a plan to port AIX or their mainframe OSs to Oracle's Niagara that they are all doomed? There are, however, long and detailed roadmaps for hp-ux and OpenVMS and for the two coming generations of Itanium. Try keeping at least one foot in reality next time.

          ".....As you cannot run MS, Oracle, Red Hat, etc on Itanium....." Which just shows your complete lack of knowledge mentuioned above. I can install Windows Server 2008 or RHEL 5 on all the current Integrity servers and run them with an hp support contract. If I want, I can install hp-ux and Integrity Virtual Machines and run hp-ux, OpenVMS, Linux and Windows VMs all at the same time. Please try and do a little learning before you post anything so stupid again.

          1. Anonymous Coward
            IT Angle

            Do you deny that Microsoft, Oracle and Red Hat have all dropped support for future versions of their software on Itanium?.... Sure, you can run Server 2008 R2. I hope you enjoy it, because it is the last version of Windows that will run on Itanium. Wait a second, I am running RHEL 6 on System z. Why haven't you upgraded to the latest version of RHEL on Itanium?

            "Why is this news? There has never been a plan to port any of those OSs to x86.".... "Try keeping at least one foot in reality next time."

            - I am honestly more amazed than upset. I didn't know there was a person in the world that cared this much about Itanium. This should be obvious, but the reason a customer would want HP to port OSs off of Itanic to x86 is because their primary ISVs, Oracle and Microsoft, have both stated that they will not support Itanium after the current versions. If they want to run their applications (assuming that they are not using IBM stack), they will need to get off of Itanium.

            1. Matt Bryant Silver badge
              FAIL

              RE: Wunderburp1

              "Do you deny that Microsoft, Oracle and Red Hat have all dropped support for future versions of their software on Itanium?...." Strange, I don't ever recall a version of Windows Server for IBM mainframe or for Pee-series. As for RHEL, would you like to divulge how much IBM paid RedHat for that to remain viable? And, seeing as there is also NonStop and OpenVMS available (with ongoign roadmaps) as well as hp-ux, there is still more operating system choice on hp Integrity than on IBM Pee-series.

              ".....I didn't know there was a person in the world that cared this much about Itanium....." I'm not surprised that would be news to someone living in the IBM mainframe timewarp. You might want to talk to some of those IBM mainframe customers that wanted to try PlatformSolutions offerings on Itanium, only they couldn't because IBM was so scared of it that they lawyered up and eventually bought and killed the company. IBM has a habit of doing that to any company (like TurboHercules) that threatens to free their customers from the IBM mainframe cash-gouging process.

              Oh, and btw, Integrity's "primary ISV" is hp itself. The vast majority of Itanium servers that have been sold have been for hp-ux and OpenVMS. RedHat and Microsoft gave up because they couldn't convince enough customers to ditch hp-ux in favour of RHEL or Win Server. That's because hp-ux and OpenVMS both offer customers smooth and simple upgrade paths, mission critical levels of reliability and resilience, and a wide choice of applications.

              Let's do a simple comparison. Hp has Integrity with two future generations of Itanium coming, with Poulsen taped out and Kittson in development (you yourself provided a link to a source talking about the Kittson development team, thanks), and that will plug into the current servers, and with the choice in two years to use existing technology to also plug Xeons into the same chassis. Hp can offer two fully proven OSs (hp-ux and OpenVMS) with large market share in the enterprise space that Integrity is aimed at. Hp also has numerous partnerships with many non-alligned ISVs such as SAP, which means hp is seen as neutral partner in many instances, and can also supply pre-tested and integrated stacks using software from competitors such as IBM and Oracle.

              Compare this to IBM and Pee-series. IBM has a current range of servers which it cannot confirm will be upgradeable to Pee-8 without major hardware changes. That's even with IBM unable to give a firm date for Pee8, which isn't even close to being taped out, and with zero public information on what comes after it. IBM can only offer one future OS on Pee-series with certainty, that being AIX, because who knows what the agreement is with RedHat and even if RHEL will be available for Pee8. And when we get to considering software it gets even funnier, with IBM reps under pressure to sell the unpopular DB2 (less popular than MySQL!) and other IBM software, and being anything but impartial (guess why SAP likes partnering more with hp than IBM). We also have zero assurance that Oracle won't pull the same lack of future support with Pee-series, and they'd have even more grounds considering the lack of IBM roadmap. IBM would be worse placed to defend against it than hp as hp has development agreements with Oracle (the basis of the hp court case) which IBM does not. And IBM does not even have a plan for a 32-way Xeon offering to put up against a Xeon Superdome2, let alone one that would allow mixing of Xeons and Power in the same frame. IBM does not have and does not plan to have any technology that would allow you to plug a Pee8 into a Xeon socket or vice versa.

              /SP&L

              1. Anonymous Coward
                Thumb Down

                IBM isn't in the business of paying ISVs and chip providers for support. They leave that to HP, per this article you are commenting on. Why is it again that you haven't upgraded to RHEL 6?

                "Strange, I don't ever recall a version of Windows Server for IBM mainframe or for Pee-series"

                True enough, Microsoft was duped like the rest of the industry and HP's customers by Intel and HP into thinking that Itanium was going to be THE Intel chip of the future. When it became painfully obvious that this is not the case, they dropped out like everyone else. No more MS on Itanic.

                I assume Microsoft doesn't support IBM's platforms because IBM, unlike HP, is a technology company that *creates* their own products and is not willing to be a Microsoft lackey/distributor with zero IP. You can't run Windows on Apple computers either for the same reason. IBM or Apple could buy HP five times over for the same reason.

                "Oh, and btw, Integrity's "primary ISV" is hp itself. The vast majority of Itanium servers that have been sold have been for hp-ux and OpenVMS."

                Hahahaha, yes, I am sure CFOs, HR VPs, Sales VPs, other line of business executives buy HP servers because running an operating system (in and of itself) creates business value. Not! They buy servers to run applications which perform a business function (e.g. SAP, Oracle, WebSphere, Cognos, Filenet, Lotus, Exchange-Outlook, etc).

                "NonStop, OVMS, HP-UX" etc, etc

                - Dead, dead, dying.

                "DB2, less popular than MySQL"

                - Already addressed this point. You are talking about raw *install* count, not revenue, not enterprise implementations. I have MySQL on my laptop, so what. You might as well take out MySQL and use Excel. It makes as much sense.

                "guess why SAP likes partnering more with hp than IBM"

                - Do a little research? - Which company is the largest SAP implementation provider in the world? IBM. Which company is the "preferred" DB provider for SAP? IBM. Which company has the largest Unix-SAP install base? IBM. SAP started as a spin off of which company? IBM.... HP isn't a "partner" in any true sense of the word. They are an undifferentiated x86 pusher. Like Dell with ink.

                "IBM would be worse placed to defend against it than hp as hp has development agreements with Oracle (the basis of the hp court case) which IBM does not."

                - Haha, not in Oracle's opinion. HP is trying to turn the Hurd press release into a contractual obligation to support their legacy product forever. Honestly, a press release as a contractual agreement. HP has zero chance to win this case.

                "Blah, blah.... HP is going to make a big x86 server"

                - If IBM wanted to get in a race to the bottom in the low-margin x86 server business, they could certainly compete with HP. IBM, like Oracle, is a technology company. They do not want to make boxes for other companies technology to sell at distributor margins. This is comparable to when HP was buying Compaq so they could beat IBM and Dell in PCs.... IBM had already move on to businesses where they could add technology value.

                1. Matt Bryant Silver badge
                  FAIL

                  "IBM isn't in the business of paying ISVs and chip providers for support....." = evasion again.

                  "....Why is it again that you haven't upgraded to RHEL 6?...." Because us customers preferred hp-ux and OpenVMS to Linux, which meant RedHat couldn't make the money they wanted to from RHEL on Itanium. Just to put that in the simplest of terms, so even you can understand it, us customers chose hp-ux and OpenVMS over RHEL as it give us better performance, reliability and resilience, than RHEL, and at a pricepoint where we happilly paid up.

                  Compare this to Linux on the IBM mainframe, where every case mentions the customer's overwhelming need to reduce the horrendous costs of being stuck with a mainframe. Indeed, the whole Linux on mainframe push from IBM has always been about stopping customers ditching their mainframes by giving them an option to reduce costs. It is never the first option offered by IBM mainframe salesmen trying to shift a new mainframe, but as a defensive measure to try and stem the flood of customers trying to get off mainframes.

                  1. Anonymous Coward
                    Stop

                    "IBM isn't in the business of paying ISVs and chip providers for support....." = evasion again.

                    - Has anyone, other than you, accused IBM of pay-to-play? You have no evidence or even a reliable source whatsoever. Oracle Corporation is accusing HP. HP does not deny it.

                    "....Why is it again that you haven't upgraded to RHEL 6?...." Because us customers preferred hp-ux and OpenVMS to Linux, which meant RedHat couldn't make the money they wanted to from RHEL on Itanium."

                    - You said that you had an instance of RHEL 5 running on an Itanium server next to HP-UX and OVMS. The answer is: You cannot upgrade to RHEL 6 as it is not supported. Red Hat, and the rest of the major software providers except IBM, have declared Itanium end of life.

                    "Compare this to Linux on the IBM mainframe, where every case mentions the customer's overwhelming need to reduce the horrendous costs of being stuck with a mainframe."

                    - What are the horrendous costs of the "inevitable" (from Oracle) migration effort off of Itanium and HP's legacy OSs? The largest banks, insurers, retailers, governements in the world all use mainframe as their mission critical standard. They certainly have the budgets and the internal capability to migrate off of mainframe, but they do not. Why? Several reasons, but these are the three major reasons. First, it is the *only* EAL5 level secure platform in the world. It is quite literally in a category of its own from a security perspective. Secondly, it has, far and away, the highest uptime levels in the world per Gartner, Forrester, IDC... everyone who monitors average mean time between failure. Third, it is incredibly efficenciet from an HR perspective. Many companies run huuuge transactional systems (e.g. the global financial backbone) with 4-5 admins. It would take dozens in a distributed environment..... As a bonus I will throw in that customers have been able to upgrade their mainframe environments for decades and decades. HP-UXs, VMSs, Wangs, Amdahls, Suns, Tandems, etc come and go with each generation... mainframe is still there.

                    - On your "point" that mainframe users are migrating from z/OS to Linux to save cost, z/OS and Linux have zero in common from an OS perspective. While there are undoubtedly users that run z/OS and Linux, some of the largest case studies on IBM's website are x86 to mainframe running Linux. Look at Blue Cross Blue Shield, migrated a 100 HP x86 DL servers to System z - Linux, saved $2 million per year in their SAP environment. Look at Transzap, migrated their Oracle environment from Unix/x86 to System z - Linux, saved millions.... More if you care to read.

                    1. Matt Bryant Silver badge
                      FAIL

                      "....Has anyone, other than you, accused IBM of pay-to-play?...." How much does IBM pay into the "joint marketting fund" compared to RedHat, that "marketting fund" being what pays for the RHEL on IBM development? Strangely, no-one at IBM seems willing to say, even when directly asked (I asked our IBM rep to find out). I'm not surprised IBM don't want to advertise how desperate they are to maintain mainframe support revenues even when it means losing the OS to another company.

                      "....You said that you had an instance of RHEL 5 running on an Itanium server...." Yes, I do, quite happilly, with full support from hp. It's on Superdomes and has another year to go on the support agreement. Seeing as the stack on that servers was designed to be easily portable (it's all commercial apps with multi-OS support), and we already have it tested on several OSs, if we do decide to move off the old Superdomes it will be a simple task. I could even fire the stack up on AIX if I was really desperate, but the costs make other platforms much more attractive. As it is, hp are quite happy to extend the support, and as the solution is meeting the performance requirements quite nicely, I predict the decision will be to take the path of least risk/effort/cost and keep it as it is.

                      "....You cannot upgrade to RHEL 6 as it is not supported...." Yes, but I don't need to upgrade, so what is the issue? I can get support on the current stack for as long as I'm likely to need it.

                      "....the rest of the major software providers except IBM, have declared Itanium end of life..." What's the betting you only class "major software providers" as ones that agree with your point of view? As just one example, I'm guessing you must think SAP is not a "major software provider" then? You make it so easy to poke holes in your silly arguments! I suggest you go read here http://www.sdn.sap.com/irj/sdn/hpux. In the meantime, I'll just be laughing at you some more.

                      "....What are the horrendous costs of the "inevitable" (from Oracle) migration effort off of Itanium and HP's legacy OSs? ...." Dunno, because I don't need to do it. Should I need to do so, we already have our stacks pre-tested with other database software, including DB2. We're currently doing dev work with PostgreSQL and our current app stack to give us a third option (we like to have two options for every part of our solution, so if Oracle does drop out then we can still choose between two database suppliers). As it is, hp have managed to provide plenty of help on PostgreSQL, and so far it has been pretty easy sailing, with no great cost at all. Sorry, you'll have to go find some new FUD as that point is minor in the extreme.

                      "....The largest banks, insurers, retailers, governements in the world all use mainframe as their mission critical standard...." So what? Hp will tell you they have the FTSE 100 all using hp Integrity. The majority of what you mentioned is also legacy mainframes, and the costs are driving those customers to find other options like x64. That makes a Superdome2 with Xeons a real risk to IBM's mainframe business, so I'm not surprised you're FUDding it as hard as you can. The gradual shrinking of the mainframe financial-house base has been going on for years, as is shown in this article http://www.onwindows.com/Articles/Mainframe-migration-in-insurance/2296/Default.aspx.

                      "....More if you care to read." No thanks, I have enough case studies coming with the FUD direct from IBM ,thanks. But, seeing as you love those case studies so much, please do read the case studies in the following Alinean whitepaper, which includes a SAP migration off mainframe that cut operating costs by 71% and saved the customer $7.68m, much more than your example: www.alinean.com/PDFs/Intel-Mainframe_Migration-TCOStudy.pdf

                      /SP&L

    3. Anonymous Coward
      Anonymous Coward

      Yes, Meg Whitman explained it on their last earnings call. HP apparently went around to the ISVs and asked them if they would support HP-UX on x86. They said no because the market would not be large enough to justify a port, so it has to stay on Itanic.

  8. TeeCee Gold badge
    WTF?

    "...HP is paying Intel to keep the Itanium processor alive..."

    So a customer gets a supplier to manufacture a product line specifically for them and they have to *pay* for it?

    Sorry, but was there actually anyone out there who *hadn't* already worked this out?

  9. Jesper Frimann
    Holmes

    Well...

    Oracle is currently doing all it can to destroy UNIX, as we know it. Shame on them.

    But one thing is sure, there are quite a few migrating Oracle on HPUX -> Oracle on something else projects being spawned.

    // Jesper

    1. Anonymous Coward
      Anonymous Coward

      Very true, for all the griping people are doing about Oracle at the end of the day they will probably take a look at the cost to migrate a platform vs. the cost to migrate a DB and stay with Oracle.... and Oracle knows it.

      I did read a survey of CIOs with Oracle DB on HP-UX on one of the IT websites in which something like 25-30% said they planned to migrate to DB2 and another 10-15% said they planned to migrate to MS SQL. The number that said they planned to migrate to Oracle DB on Sun was like 2%. I think it will cost Oracle some market share in the DB space, especially in the SAP user base, but not as substantial as expected.... Although, with DB2 licensing EnterpriseDB's PL/SQL compatibility tools, it is possible.

      1. Matt Bryant Silver badge
        Happy

        RE: Wunderburp1

        All neatly ignoring the scenario IBM fears most - if customers do look at migrating off hp-ux because of Oracle support, the first choice is going to be to migrate to x64, which hp owns. Given that hp already has many app stacks validated on x64 (such as SAP), and Oracle support on x64 is unquestioned, it is much more likely that customers would choose the much simpler and cheaper path of a migration to Linux or Windows on Proliant than IBM Pee-series. And, of course, customers that do that for hp-ux and Integrity might also start thinking about getting off Power aswell, especially considering the short and detail-free Power roadmap, and as there is nothing to stop Oracle doing the same to IBM.

  10. Adam 61

    Well.... part 2

    Agreed Jesper, and not just HP-UX - seen legacy Power (less of) and SPARC (lots of) migrations to something else as well being looked at - especially to Linux.

    Some people very wary of Oracle and want to void lock in

  11. Anonymous Coward
    Anonymous Coward

    They are both correct

    Oracle is correct that Intel is completely focused on x86 and Itanium is nearing end of life. HP would have preferred to keep the high support costs of Integrity, especially NonStop, in place for several more years while gradually migrating users to x86-Linux. Oracle basically screwed up their financial analysts' NPV models.

    HP is correct that Oracle is only dropping support for Itanium at this stage to create a market for their Exa-xxx machines. Oracle is aware that Sun Sparc is too far gone and they are hoping that the combined migration of the large Oracle - HP-UX user base and legacy Sun user base will provide a viable hardware business.

    1. Matt Bryant Silver badge
      FAIL

      RE: They are both correct

      "Oracle is correct that Intel is completely focused on x86 and Itanium is nearing end of life...." Again, more presumption dressed up as fact. Please supply even the tiniest bit of evidence to back any of that up. Honest, I'll try and act surprised when you can't. And you may want to consider that Oracle are currently off to court because hp and Intel know Oracle are talking out their anti-competitive backsides. As trolls go, you're both uninformed and wildly presumptive.

      1. Anonymous Coward
        Facepalm

        It has been stated publicly by the CEO of HP. I will copy it again for those that need a little more time to understand.

        Meg Whitman, CEO of HP, Q4 Earnings Call:

        "The BCS business is a declining business. It is a slow decline, but I don't think you're going to see an accelerating growth rate in that business," she said. "And so we just have to manage that as best we can and invest in R&D so we get to a new platform as fast as we possibly can that allows us to service the clients that need this kind of power."

        Again, "so we get to a new platform as fast as we possibly can"

        1. Matt Bryant Silver badge

          RE: Wunderbar1

          Again, you have proven nothing. All you have done is quote Meg and then make a massive jump to a pat conclusion. Who's to say that new platform is not Odessey, the mix of Xeon and Itanium? Who's to say it's not something planned for ten years from now, such as a development of Itanium. She was talking BCS, which covers storage as well as servers at hp, so she might even have meant a new storage platform or even a combined storage-server "appliance". Where does it say in Meg's comment "we are dropping Itanium for x64" - it doesn't. Just more wishful thinking from the Land of the IBM Mainframe Timewarp, where denial is more than just a hobby.

          1. Anonymous Coward
            Thumb Down

            "Where does it say in Meg's comment "we are dropping Itanium for x64"

            - Haha, yes, she is going to come out and explicitly tell Itanium customers not to buy Itanium when they are years away from the replacement. No, you need to read between the lines because it is not in HP's interest to be truthful.

            Again, we need to discuss what a platform means. It is an operating system. HP is migrating away from HP-UX and Itanium to x86 and Linux. As everyone in the world, except you, has already figured out. I don't see anyone on this board (or planet for that matter) arguing that Itanium is going to be around other than you. Who is in denial?

            1. Matt Bryant Silver badge
              FAIL

              "....Haha, yes, she is going to come out and explicitly tell...." So, when you said "....It has been stated publicly by the CEO of HP..." (Friday 16th December 2011 08:24 GMT), you actually lied. Thanks for clearing that up. I think you'll find "....you need to read between the lines...." is a long way from "stated publicly". If that's the IBM equivalent of a certain fact, publicly stated, then I wouldn't take any of their assurances about the future of Power with anything mroe than a pinch of salt!

              1. Anonymous Coward
                Thumb Down

                I did not lie. Per the post that you are responding to, Meg Whitman said that they are going to get customers to a "new platform" as "quickly as they possibly can." Being that you think the definition of a platform is not a platform, this doesn't register.

                You want the CEO of HP to state, "Please do not buy our servers" (i.e. we are dropping Itanium and Odyssey will not be out for years). The writing is on the wall for anyone willing to look at it objectively.

                1. Matt Bryant Silver badge
                  Stop

                  RE: Wunderbar1

                  "I did not lie...." Yes you did! The rest of your post is just more evasion and procrastination. You were caught out lying and got called on it, man up and admit it. What, you want me to put it in IBM terms you might find acceptable? OK, howabout we say you "extended the truth into an area where it was performance deficient to the point of inoperativeness"? There, does that make you feel better? BTW, it still means you lied.

This topic is closed for new posts.

Other stories you might like