IBM announced the PureSystems converged systems last week, mashing up servers, storage, networking, and systems software into a ball of self-managing cloudiness. What the launch did not talk a lot about is the underlying Flex System hardware which is at the heart of the PureFlex and PureApplication machines. So let's do that …

COMMENTS

House rules Send corrections

This topic is closed for new posts.

Monday 16th April 2012 16:01 GMT Anonymous Coward

yay copper porn pictures :))

Looks nice though still would have gripes with it once had a proper play I suspect, but nothing major I hope. Though that back picture does seem to indicate that they have plastic pull handels that are dispossed to block as much of the airflow comming out as possible, impacting the amount of air comming in the front, given its pulled thru.

"I also believe that IBM needs to get Power7+ processors into the field soon sporting LR-DIMM memory support and if not on-chip PCI-Express 3.0 support, then at least some way to hang PCI-Express 3.0 slots off the GX++ bus on the Power chips. IBM needs to draw even with Intel on the feeds and speeds and slots and watts."

Why do IBM need PCI-Express 3.0 slots on a Power7+ processor? Also even by my standards or lack of, that last sentence is something I would be proud of and this and that.

Strangly enough I find myself wanting to see a thermal picture of one of these that has been running at 80% load for 4 days solid. Hey if the milatary can put them in a hut on a raft and blow a few pounds of explosive underneaf them to get there jolly's then I'm sure I can be forgiven this foible.

1 1
1. Monday 16th April 2012 20:37 GMT Anonymous Coward
  
  Re: yay copper porn pictures :))
  
  I wouldn't trust IBM do deliver completely bug-free hardware or software, but I would trust them with basics such as cooling. You have to look at their intricate mainframe multi-chip modules on ceramic multilayer + spring-loaded copper bolts which hug to the backside of the chip. THAT is porn.
  
  PORN: http://www-03.ibm.com/ibm/history/exhibits/mainframe/mainframe_intro3.html
  
  EXTRA PORN: http://www-ti.informatik.uni-tuebingen.de/~spruth/ECvorles/Vorl1011/ec02.pdf
  
  So, IBM is the Ultimate MILF of cooling.
  
  1 0
Monday 16th April 2012 16:49 GMT tom 99

chipset...

"...and linked with an IBM Power7 SMP chipset. (It is not entirely clear where IBM hides this chipset, but it is possible that the Power7 architecture supports glueless connections across four processor sockets.)"

What chipset? Power7 systems support glueless configurations up to 32 sockets. The same applied to Power6 systems.

0 0
1. Monday 16th April 2012 21:22 GMT Allison Park
  
  Re: chipset...
  
  i believe they call it a distributed switch. They took the power5 chipset and put it on every power6 chip. Interestingly that allows them to not only avoid the latency of a chipset but the switch runs at half the speed of the speedy chip. More reason fast chips are good.
  
  I think power7 was a continuation of the switch technology
  
  0 0
Tuesday 17th April 2012 01:38 GMT zen1

Is it just me?

Or does that thing look like a Borg ship? your ass will be laminated by IBM

0 0
Tuesday 17th April 2012 06:16 GMT Anonymous Coward

New IBM Blade enclosure or not?

When HP introduced a new blade enclosure in 2006 IBM criticised them heavily for it:

http://www.youtube.com/watch?feature=endscreen&NR=1&v=t9vPL2f1iBw

Obviously now HP have totally domainted the blade market with >50% share in most markets, IBM realised that their own Blade enclosure just wasn't cutting it, but in a clever piece of marketing they are trying to avoid the same criticism by not calling this a blade enclosure when it quite clearly is (iti is much more similar ot HP's and Dell's blade enclosures than the old IBM blade enclosure was).

So we have a new blade enclosure, which doesn't as fasr as I can tell have anything in the way of new features over HP's 6 year old blade enclosure, which needs new management software and new "cloudy" software.

IBM does have a reasonable track record of coming into markets late and still managing to carve out a reasonable market share, but usually that's with a strong product differentiator - I'm just not seeing it here - this looks to me like IBM's take on HP BladeSystem and HP CloudSystem without the years of implementation experience

1 1
1. This post has been deleted by its author
2. Tuesday 17th April 2012 07:05 GMT Anonymous Coward
  
  Re: New IBM Blade enclosure or not?
  
  IBM is not ending support for BladeCenter H or the other existing chassis. They just released HS-23 and several new blades for the new Intel chip line on the existing chassis a few weeks ago. This is an additional product, not a replacement product.
  
  Most people are not going to buy PureSystems because they think the way IBM designed the fans or the back-plane is superior or inferior to HP's chassis (both are perfectly acceptable and not all that interesting if you are looking at it from a CIO level). They are going to buy PureSystems for the software automation "patterns" analytics and automated provisioning, build management, deployment management, etc. Flex chassis is part of an integrated system, not a stand alone hardware component like an HP chassis. Apples and oranges. IBM isn't late to this market, especially as compared to HP. The hardware level of this systems is what HP is trying to get to in a few years with Project Odyssey. HP has no plans to do anything similar to the application level work in PureSystems.
  
  0 1
  1. Tuesday 17th April 2012 08:45 GMT Anonymous Coward
    
    Re: New IBM Blade enclosure or not?
    
    >> The hardware level of this systems is what HP is trying to get to in a few years with Project Odyssey.
    
    In what way is it like project odyssey? Please provide some examples
    
    0 0
    1. Tuesday 17th April 2012 12:57 GMT Anonymous Coward
      
      Re: New IBM Blade enclosure or not?
      
      Well, as Odyssey is still in the planning stages, it is difficult to determine what it is going to look in a few years. Nevertheless, the high point of Odyssey are that they are going to unify the Unix and x86 architectures in the same enclosure around a common chassis called HydraLynx. Very similar to what IBM has done in combining POWER and x86 in the Flex chassis... actually very similar to what IBM has done for years with PS-Power blades and HS-Xeon blades in the BladeCenter H chasis.
      
      Flex Manager combines a bunch of HP's software features in Superdome as well as bunch of other features that are currently additional licenses from HP, such as automated provisioning and build software and management through the VM layer. Flex Manager has HP Insight Manager and HP's x86 Analysis Engine functionality built in which is going to be part of these Odyssey systems. Flex Manager also has Serviceguard type technology to automatically fail-over or motion workloads. I don't think there is any complete electrical isolation in the Flex chassis, similar to nPar, but everything is redundant so not a major issue.... Basically, the idea behind Odyssey is to bring "mission-critical" management and resiliency to x86 (as well as to give the current HP users something besides a dead end on Itanium). PureSystems does that and much more.
      
      0 0
      1. Tuesday 17th April 2012 15:37 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        HP's Project Oddity is even less than that.
        
        They are trying to get the SX3000 chipset to work with x86 so they can compete with the x5 from IBM. Then they want to be able to put x86 in the superdome blades so they can still get the gravy train of profit from us Tandem customers who can't afford to move YET.
        
        HP-UX is dead as they have said they will never port it to x86 because there is no interest.
        
        imma B imma B imma Matt
        
        0 1
        
        Wednesday 18th April 2012 03:23 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        Agree, HP-UX will go down with the good ship Itanic. According to Meg, HP put feelers out to their ISVs that support HP-UX on Itanium to see if they would be willing to support a port of HP-UX to x86 and the answer was a resounding no (probably in stronger language from Oracle). It is a fairly small market to start with and the number of people who would decide to migrate from PA-RISC/Itanium to x86 with HP-UX, instead of just going to Linux - x86, is really small and not worth a port.
        
        VMS and NonStop are probably going out really fast as they have all of the Itanium problems and a bunch of those customers, nearly all of the VMS users, have Oracle software as part of the stack (Rdb for VMS, Tuxedo for NonStop). It will be a cold day in Haiti when Oracle agrees to support Tuxedo for NS and they are just outright killing Rdb for all platforms. Out of curiosity, what sort of application is running on NonStop?
        
        0 1
        
        Wednesday 18th April 2012 09:57 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        Yes, HP-UX will go down. I hear that HP has financial troubles and will very soon announce the ending of HP-UX. You better migrate to IBM because Power servers will save your ass. On top, you will save lot of money and get higher performance. Migrate now, before it is too late!
        
        1 1
        
        Wednesday 18th April 2012 13:13 GMT Matt Bryant
        
        Re: New IBM Blade enclosure or not?
        
        "....HP-UX will go down with the good ship Itanic..." Well, I suppose it's a good thing there are two fully-detailed versions of Itanium still to come, then! Oh, by the way, how many versions of Power are there on the IBM roadmap after Power7? What, none with any details after Power7? Power8 is just a placemarker on the public roadmap. Instead we have another non-development called Power7+, running six months late, which has been added the IBM roadmap to hide the lack of progress. And what happens to AIX after Power8 seeing as it also has no x64 port? Trolls in glass houses should've learned by now not to throw stones.....
        
        "....Out of curiosity, what sort of application is running on NonStop?" Well, for starters there is Oracle 11g GoldenGate for NonStop, then there's IBM's WebSphere middleware, TIBCO, BEA Weblogic..... Oh, did I just poke another big hole in your FUD? Maybe you should leave it to Ali, she can at least get one good point every now and again, whereas you seem stuck on a duck every innings.
        
        0 1
        
        Wednesday 18th April 2012 23:35 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        "Well, I suppose it's a good thing there are two fully-detailed versions of Itanium still to come, then! "
        
        It doesn't matter how many versions of Itanic are released if none of the major ISVs, except IBM, will support them. The HP - Itanium brands are sinking at 27% per quarter. We have already covered the Power stuff. It is taking over the Unix market because people believe that IBM and the ISVs will support it.... Read HP's own e-mails about Itanic. There is no reason for you to think better of HP - Itanium than the HP's execs:
        
        http://www.oracle.com/us/corporate/press/1505405?rssid=rss_ocom_pr
        
        "....Out of curiosity, what sort of application is running on NonStop?" Well, for starters there is Oracle 11g GoldenGate for NonStop, then there's IBM's WebSphere middleware, TIBCO, BEA Weblogic..... Oh, did I just poke another big hole in your FUD?
        
        Not really. If that is true, that makes a stronger case for the death of NonStop than I thought. I assumed that people primarily ran NonStopSQL instead of Oracle. If they are running Oracle DB and stack, then they really need to migrate fast. If they were all HP NonStop stack, they would have time.
        
        0 0
        
        Thursday 19th April 2012 09:24 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        "....It doesn't matter how many versions of Itanic are released if none of the major ISVs, except IBM, will support them...." So, is SAP not a major ISV? LOL!
        
        "....It is taking over the Unix market ...." Hmmm, whilst that's debateable, it also ignores the fact that the UNIX market is shrinking, has been shrinking for years, and was predicted to decline long before Itanium arrived. And what is making it shrink? Well, that would be x64 eating it from below. And which company is the undisputed market leader in x64, either Windows or Linux, especially in the blades market the new IBM Flex is launching into? Well, that would be hp. So, good luck in the long term when IBM don't have a CPU roadmap, no options to port AIX, and are already losing the replacement battle!
        
        "....Not really...." LOL! So demonstrating that Oracle and a host of other software providers are available for NonStop isn't completely demolishing your attempt to imply NonStop had no software support? Wow, you're just so blinkered, did you used to "work" for Sun?
        
        0 0
        
        Saturday 21st April 2012 21:05 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        "So, is SAP not a major ISV? LOL!"
        
        SAP is a major ISV, but you need some sort of DB to run underneath SAP. DB2 is the only enterprise grade option which could fill the Oracle sized hole as MS SQL is already gone on future versions. HP-UX and the other Itanium platforms are dependent upon IBM going forward. If someone is going to run DB2, they will probably just put it on Power or x86. As Itanium products are largely running DB layer, meaning largely running Oracle, SAP's application layer isn't all that important to Itanium as it runs on x86, generally, pulling from the DB layer on Itanium.
        
        "And which company is the undisputed market leader in x64, either Windows or Linux, especially in the blades market the new IBM Flex is launching into? Well, that would be hp. So, good luck in the long term when IBM don't have a CPU roadmap, no options to port AIX, and are already losing the replacement battle!"
        
        Yes, Unix is shrinking. Those that are staying on Unix are overwhelmingly choosing Power - AIX for the same reason people chose IBM z over Wang or Amdahl when the mainframe market was shrinking, IBM looks like the only company that is going to be in the market in a few years. Yes, HP is the leader in x86 and many of the Itanium users will probably move to HP x86. Nevertheless, HP makes gobs of profit from each Itanium system (with their proprietary OSs, management software, stack software for NonStop and to some extent in VMS). They, like all other x86 server OEMs, make much lower margins from x86 gear. The replacement of Itanium with x86 blows a huge hole in HP's profitability. They are no longer going to be able to subsidize x86 system technologies, like the future SX socket from Itanium, if they are not selling the high margin systems.
        
        "So demonstrating that Oracle and a host of other software providers are available for NonStop"
        
        If NonStop users were running HP NonStop software, HP would have control over the future roadmaps. As it runs on Itanium, that means Oracle is gone in the next generation and there is a much higher level of disruption and migration required. If you are right, about NonStop users running Oracle and a range of other ISVs that have, or will soon, drop support for Itanium.
        
        0 0
        
        Sunday 22nd April 2012 20:29 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        "....DB2 is the only enterprise grade option...." I think you need to go tell SAP that, they have plans for a little something called Sybase. Non-SAP, there are other hp-ux options like PostgreSQL, which is a lot better and cheaper than DB2 even in the full-fat EnterpriseDB form. But I wouldn't expect an IBM troll to know that.
        
        ".....HP-UX and the other Itanium platforms are dependent upon IBM going forward...." Well, seeing as IBM Software sell more software licences on hp kit than IBM's own, I'd say IBM Software was the one more dependent on hp.
        
        "....The replacement of Itanium with x86 blows a huge hole in HP's profitability..." Not that big. Looking at the spread of profits, if hp were to lose the Itanium bits tomorrow it would be painful but not that bad, they are well-diversified. If IBM lost the Power/mainframe gig then it might even be fatal. After all, hp have already chased them out of desktops, just about chased them out of x64 servers, and IBM are now dropping the POS bizz as well. And IBM have nothing to compare to the success of hp's networking or printer businesses.
        
        Your predictions of Itanium's demise, as with all the IBM FUD depends on two falsehoods - Oracle have already won their courtcase; and that all hp-ux servers run nothing but Oracle instances. Seeing as one Oracle fraud claim has been rejected already, Oracle are not looking good for their courtcase. They also have the partnership agreement from way back that predates the Hurd agreement, all hp is arguing is that the Hurd settlement was a re-affirmation fo the previous agreement, i.e., that Oracle would release versions of their software for Slowaris and hp-ux concurrently. And as for not all hp-ux servers running Oracle, I have over a dozen hp-ux servers in my care running no Oracle software at all. From what I hear from other hp-ux users, the majority of hp-ux servers don't run Oracle either. Which kinda shows the limits of the desperate IBM FUD.
        
        0 0
        
        Sunday 22nd April 2012 22:24 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        "Looking at the spread of profits, if hp were to lose the Itanium bits tomorrow it would be painful but not that bad, they are well-diversified. If IBM lost the Power/mainframe gig then it might even be fatal."
        
        Losing Itanium BCS won't sink HP. It will just take away a highly profitable division and no longer allow them to subsidize x86 technologies with Itanium systems. IBM will still be able to pass down technology from z and Power to x86 (e.g. eX5) whereas HP is looking at a Dell scenario where they can't spend large amounts on server R&D because their x86 only server lines will be unprofitable or uncompetitive..... There is no indication that Power/mainframe are going anywhere. It would not be fatal as IBM still has services and software, but it is strawman argument as there is no reason to believe it will happen... like "if Oracle lost database or if MS lost Windows." Yes, either of those scenarios would not be good, but there is also no indication that they are happening.
        
        "After all, hp have already chased them out of desktops, just about chased them out of x64 servers... POS, print, etc"
        
        HP didn't chase IBM out of anything. IBM realized that doubling down on PCs (a low profit profit commodity) and print (a rapidly declining industry) was irrational. While HP was buying Compaq to become the top PC seller in the world and relying on ink for profits, IBM was buying PwC and software companies to become a software and services company and moving away from boxes (hardware without software). HP has since realized their mistake and has been desperately trying to catch up in services and software (EDS, Opsware, Mercury, Autonomy, etc). POS terminals are largely Windows based today which means they are a waste of resources from an OEM perspective (the same PC commodity problem). The volumes in PCs gave HP pricing advantages over IBM and others in x86 and components, but I would hardly say they chased them out of x86 servers. IBM is still number 2.
        
        "Your predictions of Itanium's demise, as with all the IBM FUD depends on two falsehoods - Oracle have already won their courtcase; and that all hp-ux servers run nothing but Oracle instances."
        
        First, IBM is trying to get Itanium users to migrate to DB2 off of Oracle. They would be pleased if everyone with Itanium decided to stay on Itanium and migrated away from Oracle. Hell, IBM is the only friend HP Itanium has at the moment.
        
        I have never heard of any "partnership agreement" which contractually compels Oracle to support Itanium or any other OEM. Port agreements have generally been understanding based upon mutual benefit. HP has never mentioned any binding development agreement in their press releases. If HP wins the court case, which is supremely unlikely, it will be years from now and Itanium, HP-UX will have long since been virtually gone. The best outcome for HP is that Oracle has to fork over a bunch of damages cash.
        
        HP-UX servers certainly run other applications outside of Oracle, but Oracle is far and away the most valuable and widely install application. I think that is almost undeniable. If you take away Oracle, Microsoft, and Red Hat, you are really only left with IBM as a major ISV, particularly a DB ISV, which runs on Itanium.
        
        0 0
        
        Monday 23rd April 2012 19:14 GMT Matt Bryant
        
        Re: New IBM Blade enclosure or not?
        
        "....IBM will still be able to pass down technology from z and Power to x86...." Was that a joke!?!?!? IBM has no technologies that would be either of interest or of value to Intel. The IBM mainframe tech was matched years ago with Xeon kit like the Unisys offerings, and exceeded by Itanium kit like the Superdome. One glaring example of that was how IBM reacted when PSI started loading their own BIOS onto Itanium Superdomes and running emulated mainframes at a fifth of the cost and twice the performance - IBM squealed for the legal eagles, tied PSI up in court and then bought and killed them. IBM can't afford anything that exposes the sham of their mainframe operation for the old-age tech it is.
        
        0 0
        
        Sunday 22nd April 2012 22:37 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        "I think you need to go tell SAP that, they have plans for a little something called Sybase. Non-SAP, there are other hp-ux options like PostgreSQL, which is a lot better and cheaper than DB2 even in the full-fat EnterpriseDB form. But I wouldn't expect an IBM troll to know that."
        
        You are really getting desperate. As you probably know, Sybase begged SAP to certify them for SAP applications for 15 years. SAP refused. Now that SAP bought ASE, incidentally as they were acquiring mobile from Sybase, they will certify it, but Sybase has about 1.5% of the DB market. ASE was thoroughly beaten by Oracle decades ago. EnterpriseDB was never interested in Itanium until Oracle left them without a DB partner. Now, after the Oracle situation, EnterpriseDB is supposed working on an Itanium port. As PostgreSQL is not supported by SAP, Oracle ERP, or any other major application players, it isn't going to be a major deal.
        
        "Well, seeing as IBM Software sell more software licences on hp kit than IBM's own, I'd say IBM Software was the one more dependent on hp."
        
        I am not sure if that is true. If it is, that is like saying that Microsoft is dependent on HP because a bunch of their Server and Windows PC licenses run on HP hardware. HP x86 is commodity. If they were to go away, people would just put the software on some other x86 gear. I don't think HP ProLiant is going to put the screws to WebSphere or Windows Server.
        
        1 0
        
        Monday 23rd April 2012 19:19 GMT Matt Bryant
        
        Re: New IBM Blade enclosure or not?
        
        "....You are really getting desperate..." Which is funny, seeing as I hadn't seen a bit of IBM FUD attacking Sybase ever, but within a month of the SAP purchase it started flowing at full volume! Looks like IBM are the ones getting desperate.
        
        "....EnterpriseDB is supposed working on an Itanium port...." EnterpriseDB ASE is already supported on hp-ux on Itanium, as you can see here: http://www.enterprisedb.com/products-services-training/products/postgres-plus-advanced-server/downloads . I note it is not an option for AIX yet. Once again, you're just showing that you know nothing outside your little IBM bubble.
        
        0 1
        
        Tuesday 24th April 2012 05:50 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        I haven't seen any FUD from IBM on Sybase. I think IBM, like Oracle and everyone else, had basically forgotten that Sybase ASE existed since the 90s. ASE used to be a player but it was hammered by Oracle back in the day and has been limping along with around 1% market share for the past decade. SAP didn't think it was worthwhile to incorporate into their ERP ecosystem until they acquired it as a kicker in the Sybase acquisition... which they were acquiring for mobility.
        
        Yes, EnterpriseDB created a port, but they only created a port to try to capitalize on the Oracle - HP fallout. It wasn't on the radar until they thought they could find a few HP-UX - Oracle migrations. It all happened as a short term sales tactic to capitalize on the HP - Oracle fallout. Regardless, as I mentioned, there are not many, really any, major ISVs (SAP, Oracle, Lawson, MA, Red Prairie, etc) that support Postgre, so unless you have a major custom developed Oracle application on HP-UX, it isn't an option.
        
        0 0
        
        Tuesday 24th April 2012 13:43 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        So all that is just a very meandering and evasive way of admitting you were completely wrong about Sybase and EnterpriseDB on hp-ux.
        
        0 0
        
        Thursday 19th April 2012 10:11 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        Oh, almost forgot your segway into the Oracle Land Of Fantasy! After reading the Oracle press release, where I'm sure they quote any hp or Intel emails in their full and not out of context, Shirley, you might want to swing over to the hp pressroom which has the more legally damaging (for Oracle) fact that Oracle's fraud calim got kicked out of court (http://www8.hp.com/us/en/hp-news/press-release.html?id=1166547#.T4_hzdXLsuI).
        
        0 0
        
        Saturday 21st April 2012 21:22 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        Oracle's PR has actual quotes taken from HP execs. How do you take these out of context?
        
        "HP-UX is on a death march due to inevitable Itanium trajectory."
        
        “The regions are unaware of the situation with Itanium and the impending end of life.”
        
        Unless they were followed by "just kidding", they seem to be unequivocal.
        
        The HP PR is the same stuff they have been pushing. That the Hurd settlement = a product support contract with Oracle. HP is clinging to the possibility that the courts will consider the Hurd settlement PR as a formal contract to support all of HP's products seemingly until the end of time. I generally don't fall in Oracle's corner, but this is too absurd on HP's part. How many other product support contracts have they negotiated through an employee non-compete settlement lawsuits? Dorian Daley, Oracle GC, explicitly told HP that the Hurd settlement was not a guarantee of product support or terms and forced HP to take any language relating to Itanium out of the PR.
        
        0 0
        
        Sunday 22nd April 2012 20:41 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        "Oracle's PR has actual quotes taken from HP execs...." Yes, and please do give details of exactly which hp personnel, from whom to whom the emails went, and the content of the emails? Oh, you can't - not a surprise. In fact. all you have is Oracle's sliced'n'diced version, which I'd take at as much face value as their claim that ZFS is stable.
        
        ".....HP is clinging to the possibility that the courts will consider the Hurd settlement PR as a formal contract to support all of HP's products seemingly until the end of time....." Oracle and hp had a development agreement from prior to the Sun purchase that committed Oracle to releasing versions of their software on Slowaris and hp-ux concurrently. Before that, Oracle had developed primarily on SPARC and released the Slowaris version of a product first, then ported it to PA-RISC and Itanium. As part of the agreement, Oracle started developing on Itanium and also took a load of hp servers and storage in joint hp-Oracle demo centers (IIRC, the one in the UK was the Oracle ETC up in Reading). You may also recall at the time that hp helped Oracle with their Collaboration Suite, providing the know-how to get Oracle's Worldwide email system off a large number of Wintel servers and onto a few Itanium boxes. All hp meant was that they asked if the Hurd agreement meant Oracle was back to business as usual and was re-affirming that agreement still held.
        
        0 0
        
        Wednesday 18th April 2012 23:43 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        When I asked "Out of curiosity, what sort of application is running on NonStop?", that wasn't an insult in NonStop's direction. NonStop is solid gear with great I/O bandwidth. It just had the misfortune of falling into the hands of HP and being put in their land of misfit toys with VMS - Alpha, Tru64, PA-RISC, etc.
        
        0 0
        
        Thursday 19th April 2012 09:32 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        "....that wasn't an insult in NonStop's direction..." No, it was an attempt at FUD.
        
        ".....NonStop is solid gear with great I/O bandwidth..." But it runs on the same hp Itanium servers as OpenVMS and hp-ux, which you seem intent on FUDding on one hand but then say is "solid gear with great I/O bandwidth", so I would have to presume you actually don't have a clue about the Itanium, hp-ux, OpenVMS or NonStop products.
        
        "....land of misfit toys..." More FUD. OpenVMS has a clear roadmap with a lot more details and range than any IBM AIX roadmap, either public or NDA'd. Tru64 was always for the chop simply because hp were committed to hp-ux which had a bigger marketshare, as good or better features (especially with the Veritas content), and superior performance on Itanium to Tru64 on Alpha. And PA-RISC was hp's own CPU technology, which Itanium was designed (by hp) to replace. You need to go do a lot more reading outside of IBM sales brochures.
        
        0 0
        
        Saturday 21st April 2012 21:13 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        NonStop and VMS (rich man's MS NT) were solid products, but HP never invested in them and they eventually became a small part of BCS without much direction other than to try to hold on to the lucrative install base for as long as possible.
        
        OpenVMS is truly and properly dead. The vast majority run Oracle's Rdb database (originally the database DEC designed for VMS, but sold to Oracle). As Oracle has announced they will not support Rdb on any platform (killing the product) and Oracle DB is gone for Itanium, there are no options but to migrate.
        
        Yes, PA-RISC was viable, but because HP did not want to continue with the costs of CPU design and manufacture. They cast their lot with Itanium, which was the end of these platforms.
        
        0 0
        
        Wednesday 18th April 2012 09:32 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        wow the FUD is flying today... comepete with x5? I think you'll find that Proliant competes pretty well with x5 if market share means anything? And as for getting the sx3000 chipset for x86, well yes that is a goal, but not to compete with x5, the DL980 already has pretty much all the sx3000 technology in it that can be used with the current generation of Xeon and is more than capable of competing with x5 based systems... suggest you actually do some reading rather than just listening to the FUD from IBM.
        
        0 0
        
        Wednesday 18th April 2012 12:56 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        Your arguments are highly detailed and backed up by a swathe of references. Oh, actually they're not, it's all just FUD. Honest, this is my surprised face.
        
        0 0
      2. This post has been deleted by its author
      3. Wednesday 18th April 2012 09:45 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        >> Well, as Odyssey is still in the planning stages, it is difficult to determine what it is going to look in a few years. Nevertheless, the high point of Odyssey are that they are going to unify the Unix and x86 architectures in the same enclosure around a common chassis called HydraLynx.
        
        It might just be in the planning stages, but there's enough publicly available material at their website to correct some of your errors...
        
        HP already offer Unix and x86 in a common enclosure- it's called BladeSystem c7000, and you can put blades in it running Windows/Linux on x86 and HP-UX/OpenVMS on IA64. That's been in the marketplace as an offering for a good 5 years, just as IBM have offered Power and x86 in their BladeCenter H enclosure. If you think that is what Odyssey is, you are wrong.
        
        The hardware side of Odyssey (ignoring the software/services and other components) is about:
        
        i) Producing a "scalebale x86 blade" similar to the BL860/BL870/BL890 IA64 blades where you can grow an existing blade from a 2-socket to a 4-socket to an 8-socket blade by adding additional blade modules and then combining them together using a blade link. This is similar to what IBM do with their p5x0 components, except in a blade enclosure and without all those nasty/messy interconnect cables. This is HyrdaLynx.
        
        ii) Producing a "x86 Superdome" - that isn't in the same BladeSystem c7000 chassis, but in a Superdome 2 blade enclosure which shares many components with the c7000, but is different in that it has a resilent compute fabric for interconnecting the blades and IO enclosures to create electrically isolated partitions, and deliver enhanced failure detection/correction on a level you see in Integrity and Power systems, but don't see in the x86 world right now (not in a Flex Chassis either). This is known as "DragonHawk" - if IBM were to do something similar, it would be more akin to a x86 version of the p795, not this Flex chassis.
        
        >> Flex Manager combines a bunch of HP's software features in Superdome as well as bunch of other features that are currently additional licenses from HP, such as automated provisioning and build software and management through the VM layer.
        
        A closer comparison is of course would be with BladeSystem/VirtualSystem/AppSystem/CloudSystem, which as I said previously have offered these sorts of capabilities for both Unix and x86 acrhitectures for a few years now. Any conversation about licenses is irrelevant without doing a full TCO compare, which I hope you will agree is outside the scope of a friendly discussion on a forum.
        
        >> Flex Manager has HP Insight Manager and HP's x86 Analysis Engine functionality built in which is going to be part of these Odyssey systems.
        
        I'd love to hear where you read that - if you talk to HP in any detail, they will tell you that the reason there isn't currently a x86 Analysis Engine similar to the one in the current Superdome 2 IA64 system, is because the x86 processors won't have the required features until the next iteration of the Xeon processor - so unless IBM have done a ton of firmware work here that they won't need in the next rev of their Xeon processors, I find that highly unlikely.
        
        1 0
        
        Thursday 19th April 2012 00:17 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        Meh, you could be right. I don't really care to get into the malaise of socket compatibility, what IBM x86 RAS analysis does vs. HP's proposed future x86 analysis engine, etc, especially as all the documentation is available FAQ pages and stuff HP sales execs said. It is not like there is any Redbook equivalent for Odyssey. Anyway, we are talking about the least interesting part of PureScale at the hardware level. It is all about the automation and software level integration.
        
        I think HP is going down the wrong path in trying to produce a massive scale-up x86 box with no application level value add. It just isn't a requirement for many people. You can achieve greater reliability and resiliency using a bunch of clustered 2U, two socket boxes, e.g. Hadoop style framework or global file system in general. Maybe there are people out there that are just begging for a large OS image on x86 and not to just VM slice and dice it across a cluster, but I don't know any. I think IBM understands, and Oracle understands, that any scale-up systems will need to do more than just RAS (as you can get all of those benefits by clustering) and big OS images. There are some interconnect latency benefits, but nothing to write home about. If you are going to do scale-up it needs to be integrated through the software level with a high degree of automation. If you are just proposing a bid box with no software value add, there isn't much point anymore. Just scale it out.
        
        0 0
        
        Thursday 19th April 2012 09:39 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        "....It is all about the automation and software level integration...." Yeah, go take a look at the hp Cloud Matrix offering, that might give you a clue as to how far ahead hp are of IBM. And as for application integration, hp do that with a range of partners (including IBM Software), whereas IBM have this fraught relationship where they can't talk honestly to a customer about a competitor's software product without pushing the IBM equivalent.
        
        ".....I think HP is going down the wrong path in trying to produce a massive scale-up x86 box...." But hp aren't just going down the massive x86 scale-up option, they offer everything from ultra-dense x86 HPC blades through the market-leading 2-socket rack server to scale-up and scale-out x86 solutions, all with integration offerings for a swathe of partner software. It's that range of offerings which is why they are the leading x86 vendor, not IBM. The only unique IBM have is the mainframe customer gouging operation, and that is shrinking too.
        
        0 0
        
        Saturday 21st April 2012 21:27 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        "Yeah, go take a look at the hp Cloud Matrix offering, that might give you a clue as to how far ahead hp are of IBM."
        
        HP does nothing above the OS layer. They are not automatically tuning databases and provisioning application servers or anything of the sort.
        
        "But hp aren't just going down the massive x86 scale-up option, they offer everything from ultra-dense x86 HPC blades through the market-leading 2-socket rack server to scale-up and scale-out x86 solutions."
        
        They should focus on those commodity two-sockets, because the market for x86 Superdome is small.
        
        0 0
        
        Sunday 22nd April 2012 20:48 GMT Matt Bryant
        
        Re: New IBM Blade enclosure or not?
        
        "....HP does nothing above the OS layer..." Actually, hp does a lot through partner programs with companies like Microsoft and VMware, integrating their products with hp's control software, meaning they do provision databases and application servers.
        
        "....They should focus on those commodity two-sockets, because the market for x86 Superdome is small." And your supporting statistics for that are .... non-existant! Just because you want it to be so, doesn't make it so. Indeed, I can remember when we first looked at Unisys E7000s back before the Y2K party, and IBM were desperately FUDding Unisys as they knew that customers would switch out mainframes for cheaper Unisys Xeon-based gear. I'm not surprised the IBM trolls are now throwing FUD at what would be a Xeon server with mainframe class RAS from the leading x64 vendor.
        
        0 0
        
        Sunday 22nd April 2012 22:42 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        "Actually, hp does a lot through partner programs with companies like Microsoft and VMware, integrating their products with hp's control software, meaning they do provision databases and application servers."
        
        Yes, in the same sense that Tivoli, BMC and the other systems management ISVs have partnerships with MS and VMware, but it isn't baked into the system. All of that software still has to be implemented, configured, some patch process has to be created for the individual environment, etc.
        
        "And your supporting statistics for that are .... non-existant! Just because you want it to be so, doesn't make it so"
        
        Yes, I don't have any adoption rate statistics for a non-existent product, but, in my opinion, big x86 doesn't seem to be solving any needs. Everyone is going the VMware, Hadoop, etc route of slicing and clustering. It is my opinion.
        
        "Unisys E7000s back before the Y2K party, and IBM were desperately FUDding Unisys as they knew that customers would switch out mainframes for cheaper Unisys Xeon-based gear."
        
        If that is the case, it seems to have worked.
        
        0 0
        
        Monday 23rd April 2012 13:40 GMT Jesper Frimann
        
        Re: New IBM Blade enclosure or not?
        
        Actually I think there is kind of an underlying movement in the market. A consolidation and movement towards that suppliers can do an (almost) vertical solution stack. This is kind of like back to the future, to a time where you got your whole IT from a single vendor. This is IMHO much more a trend rather than partnerships between different vendors in the solution stack.
        
        I don't think this is a good thing for us that have to procure the whole software and hardware stacks. I think we are going to see less open standards, less portability and more vendor lock-in.
        
        If this movement continues, and it is a big IF, there are going to be further consolidations, and to be quite frank then HP wouldn't be one of the companies that would be able to buy up other big companies quickly right now. I mean HP's long term debt is more than 50% of the current marked cap of the company. I know that the debt is in practice deducted from the maket cap, but it's still a huge chunk of depth. For comparison IBM (although having more long term debt than HP) it's still less than 15% of the total company cap, around the same % as Cisco and Oracle has.
        
        So again if, and there is a lot of if's here, this vertical trend continues, then personally I think HP needs to merge with someone.
        
        // Jesper
        
        1 0
      4. Wednesday 18th April 2012 12:54 GMT Matt Bryant
        
        Re: New IBM Blade enclosure or not?
        
        ".....the high point of Odyssey are that they are going to unify the Unix and x86 architectures in the same enclosure around a common chassis...." Yeah, only hp having been offering Itanium and Xeon (and AMD) blades in the same blade formats in the same balde chassis for years. All the software features you say will be in the IBM Flex offering are already available for hp. What Odyssey will bring is common servers (not just "in the same enclosure") for Itanium and Xeon, ie socket-level compatibility, which IBM simply can't offer (and probably never will) with Power, meaning hp will be able to maximise the savings of a common platform whilst IBM will still be gouging their customers for two seperate hardware streams.
        
        0 0
        
        Wednesday 18th April 2012 23:49 GMT Anonymous Coward
        
        Re: New IBM Blade enclosure or not?
        
        Yeah, alright, so in HP's plans to have a plan, they are going to used the same SX socket for x86 and Itanium. Maybe there is some benefit for HP in consolidating manufacturing costs, but it doesn't do much for customers.... especially as IBM will probably be on eX7 when these systems see the light of day which will be ahead of SX in SMP tech, I/O, etc. Hard to say as we are arguing about what might happen in the future.
        
        "All the software features you say will be in the IBM Flex offering are already available for hp."
        
        Yes, that was because the question was how is Flex like Odyssey, so I didn't include things which were not available in Odyssey. Look at the entire PureApplication system for things which HP doesn't have any answer for.
        
        0 0
      5. Thursday 19th April 2012 09:10 GMT Matt Bryant
        
        Re: Re: New IBM Blade enclosure or not?
        
        "Well, as Odyssey is still in the planning stages...." I love how you start by admitting that you actually don't know anything about Odyssey, and then go on for two paragraphs insisting it will be this, that and the other!
        
        0 0
  2. Wednesday 18th April 2012 12:48 GMT Matt Bryant
    
    Re: Re: New IBM Blade enclosure or not?
    
    "....Flex chassis is part of an integrated system, not a stand alone hardware component like an HP chassis...." So you've never heard of the hp Matrix offerings then?
    
    0 0
3. Wednesday 18th April 2012 13:28 GMT Matt Bryant
  
  Re: New IBM Blade enclosure or not?
  
  "......(iti is much more similar ot HP's and Dell's blade enclosures...." Don't be silly! In the same 10U the hp c7000 blades chassis has more blades slots, more switch module slots, a superior converged solution with Virtual Connect and Virtual Flex, even the option of a DVD drive. Oh, I see now - this is a new IBM blades chassis that is only slightly more of a copy of the hp one than the existing IBM blade chassis, but still lightyears behind hp's! Can we expect the IBM sales FUD to drop the bit about hp's "small and noisy fans" now that they have copied them too? Now, if only IBM could produce something to match the hp Matrix offering, which has only been out there gaining a headstart for two-and-a-half years.....
  
  0 0
Tuesday 17th April 2012 07:16 GMT Bob Boblowski

Thanks!

That was actually a informative article. Makes it much easier to understand what these new systems are about and what we can expect from them.

0 0
1. Wednesday 18th April 2012 08:28 GMT Jesper Frimann
  
  Re: Thanks!
  
  I think he got most of it from www.redbooks.ibm.com, not that this is not an excellent boildown.
  
  IMHO this looks like 'just another scale out platform, kind of looks like a modified iDataplex solution.
  
  // Jesper
  
  0 0
Wednesday 18th April 2012 13:47 GMT Matt Bryant

More constrained Power blades!

Well, it's actually an advance over the preceding IBM Power blades, which require you to use expensive low-profile memory, in that you can actually use cheaper (that's relative - no IBM memory is ever cheap!) full-height memory now! As long as you don't want onboard disks, that is, in which case you're back to the extra-expensive modules. And then the disks aren't hotswap. Again. But you still have to use the crippled versions of the Power7 CPUs, not the full-power (4.25GHz) ones. Again. I'm also wondering, where are the built-in NICs? Seriously, you have to buy mezz cards just to get LAN connectivity? Why are the Power blades such retarded designs, is it really so difficult to get full-powered and featured Power blades working inside the power and coolling capabilities of IBM's blade chassis designs?

0 1
1. Friday 20th April 2012 03:53 GMT Allison Park
  
  Re: More constrained Power blades!
  
  "But you still have to use the crippled versions of the Power7 CPUs, not the full-power (4.25GHz) ones."
  
  Are you the HP Bigot who likes Tukwila which is sadly still dual core, sadly 65nm, 1.75GHz?
  
  How's that glass house you are throwing rocks out of? And I say glass house cause Oracle outed HP's planned demise of Itanium so now everyone knows the real roadmap.
  
  0 0
  1. Friday 20th April 2012 14:38 GMT Matt Bryant
    
    Re: Re: More constrained Power blades!
    
    "....Tukwila which is sadly still dual core...." Oh dear, nowadays Ali is slipping almost as badly as Wunderburp! Tukwila is quad-core, and the bench figures Oracle are sitting on seem to show it ouperforming even the Power7 chips too powerful for the IBM blades to handle, let alone the crippled ones. Back to troll school for you!
    
    0 0
    1. Saturday 21st April 2012 00:24 GMT Allison Park
      
      Re: More constrained Power blades!
      
      sorry yes Tukwila is quad core while everyone else is octo-core ..... my bad still not something to be proud of.
      
      and you refer to benchmarks that oracle wont release as "proof" that Itanium is competitive? Sorry HP gave us the SAP numbers for tukwila and they were embarrasing compared to the DL980 let alone compared to the Power 780 we ended up buying.
      
      Nice try....thanks for not addressing the two generation old fab tech, the incredibly slow chip speed, or the fact that kittson is the last chip for Itanium. Will be interesting to see if the initial Kittson will be crippled to give the perception that there is a kittson+ later like HP was planning.
      
      kisses
      
      0 0
      1. Monday 23rd April 2012 19:21 GMT Matt Bryant
        
        Re: More constrained Power blades!
        
        "....or the fact that kittson is the last chip for Itanium...." Neatly ignoring that Poulsen is the next generation of Itanium due, already taped out, and Kittson will follow that. Details of both have been released by Intel. Now, how many future generations of Power are there on their public roadmap with any details? Oh, none!
        
        0 0
        
        Friday 27th April 2012 02:18 GMT Allison Park
        
        Re: More constrained Power blades!
        
        Power7+ and Power8 with full details.
        
        I saw it on the website called the register http://www.theregister.co.uk/2011/08/31/ibm_power_chip_roadmap_update/
        
        All I have seen of Poulson is 32nm and more cores.
        
        Kittson is only a name with promised socket compatibility which now I hear will not be the case.
        
        0 0
        
        Friday 27th April 2012 11:11 GMT Matt Bryant
        
        Re: Re: More constrained Power blades!
        
        LMAO! Let's just look at the detail on that slide for PowerL8 (sic), shall we? Does it have a date for PowerL8? No, we have to guess it may be in 2013, going by the "3 years revolution, 18 month + evolution", but by that Power7+ is already late, pushing PowerL8 out to 2014 or later. If Power7+ is just the easy evolution (read "speedbump") and that's running late, what does that say about IBM's chances of making a real architecture change and delivering it on time?
        
        Anyhooooo, back to the IBM roadmap slide. Does the IBM it say how many cores PowerL8 will have? No, it just says "many". Any details on cache size? No. Any figures at all? There is only one figure giving any details and that is where it does say it will be 22nm. In all, a big, fat nothing there.
        
        Compare this IBM leak to the Intel public roadmap, which has been up on their web for ages (http://www.intel.com/content/www/us/en/processors/public-roadmap-article.html), and you'll see on slide 19 the details for Poulsen with a lot more detail.
        
        0 0
    2. Saturday 21st April 2012 11:39 GMT Jesper Frimann
      
      Re: More constrained Power blades!
      
      Well, why talk about benchmarks that supposedly exist in HP labs. Why not simply look at the ones that are out there.
      
      From what I've been able to see then the p460 will outperform the BL870c i2 by a factor of.. well four at aprox the same price. It'll outperform the BL890c i2 with a factor of two at aprox half the price.
      
      Sorry but Tukwila is capable of competing with POWER6(+) but against POWER7 and current Xeons' it not even close. I mean it even get creamed by SPARC.
      
      And last.. then people buy hardware to run software on, and the club of software that runs ontop of Itanium is shrinking fast.
      
      I take no pleasure in stating that IMHO Itanium is a dead man walking.
      
      // Jesper
      
      0 0
Monday 23rd April 2012 19:22 GMT Matt Bryant

IBM troll tactcis.

Interesting to note that the IBM trolls don't like trying to defend the new IBM designs (not surprising really), but instead go on the FUDpath.

0 0