back to article IBM: Power7 to rollout throughout 2010

IBM's chief financial officer, Mark Loughridge, explained in a conference call with Wall Street analysts yesterday that the company thinks its high-end server business is on the mend, and will start growing profits in the fourth quarter. Loughridge also let the cat out of the bag a little about future Power7 RISC/Unix and z11 …

COMMENTS

This topic is closed for new posts.
  1. -tim
    Flame

    Just another PPC?

    I would love to see a pair of these in a hackintosh. I wonder how that would rank in the benchmarks.

  2. Matt Bryant Silver badge
    Happy

    Interesting times ahead!

    Can't wait to play with Power7! Darn, the IBM salesgrunts might read that, I must act nonchalant in the hope of a bigger discount. Then again, the hp salesgrunt could also read that and offer a discount.... This could be tricky!

    "....Ironically, as Loughridge was talking up the current product line, his comments about the future Power7 and z11 products could actually undercut sales in the coming quarters...." Well, yes and no. Yes, many customers will think about waiting. But the death of Sun did make roadmaps a lot more important to many in the industry, so Loughridge was probably trying to reassure IBM customers that they're on track and have a future.

  3. Anonymous Coward
    Terminator

    Too bad the same cannot be said for Tukwila

    I am very disappointed with Itanium. HP is now admitting Tukwila will be late 1Q and the systems will be late 2Q. On top of the 18 month delay there is no upgradeability. There is no way I will buy and current machines which were supposed to be replaced in 2008.

    The worst is I hear there is no core performance improvement and Oracle is set to increase the core factor for Tukwila.

    I need to find a better database platform and dump Oracle.

    Lounge Lizard Larry Lies

  4. Joe Cooper

    IBM chips

    "New products always have better bang for the buck, so customers who can wait often do"

    How come I only _ever_ hear this when people are talking about IBM chips?

  5. Kebabbert

    Hopefully

    Hopefully the 5 GHz Power7 will be several times faster than the Power6+ so it can match the 1.4GHz Niagara T2+. For instance, to match one 1.4GHz Niagara on Siebel v8 benches, you need six 4.7GHz Power6 and one 1.9GHz Power6.

    See IBM white paper on this fact:

    http://www.oracle.com/apps_benchmark/doc/IBM_Siebel8_7000_PSPP_On_AIX_POWER6%20Final.pdf

    And SUNs result:

    http://www.oracle.com/apps_benchmark/doc/sun-siebel-8-14000-pspp-on-solaris-benchmark-white-paper.pdf

    .

    So the 5GHz Power7 must be more than six times faster than the 4.7GHz Power6+ to match one 1.4 GHz Niagara. But that might be difficult to achieve as it seems that the Power7 core is weaker than the Power6 core, according to comments here:

    http://www.c0t0d0s0.org/archives/5941-The-new-TPC-H-benchmark-from-IBM.html

    It seems that the octo core Power7 is only 2-3 times faster than the dual core Power6+. Then it will not be fast enough to catch up the 1.4 GHz Niagara on Siebel.

  6. Nik Simpson

    Will Power7 fit in existing systems?

    Do you know if this will be a forklift upgrade for customers, or will the Power7 be a drop-in replacement in existing systems?

  7. Adam 61
    WTF?

    Kebbabert please shut up

    Stop raving on like an automaton about bloody Niagara, I think everyone has had enough of this to last them a lifetime (please back me up here guys!)

    Just change the record man its getting extremely boring!!! Not every debate has to turn into a Niagara benchmark debate, have you got any other points of view???

  8. David Halko
    Go

    It is nice to see IBM...

    It is nice to see IBM coming to the Octal-Core show!

    That means there are 2 major vendors in this area:

    Sun - first to the octal core, engineering everything in a single piece of silicon

    IBM - second to the octal core, cobbling their cores on multi-chip modules

    With the runners up, still building their projects:

    Sun - first to the hex core, engineering a single piece of silicon

    Intel - second to the hex core, cobbling their cores via 3 chips onto a multi-chip modules

    AMD - third to the hex core, engineering a single piece of silicon

    On the horizon...

    AMD - possibly be the first 12 core in a socket, using 2 chips in a multi-chip module in 2010???

    Sun - possibly be the first 16 core in a socket, using a single piece of silicon in 2010???

    I look forward to the benchmarks of the Power7 and the system lineup from IBM!

    It is GREAT to see the competition and innovation!

  9. Captain Thyratron

    Re: Kebabbert

    Reading threads in which you post is like watching a lecture on semiconductor physics when, suddenly, a malfunctioning robotic Billy Mays bursts through the door and tries to convince the audience that indium arsenide is more suitable for nasal insertion than gallium antimonide. I would call you the worst shill ever, but Sun's marketing staff, despite their illustrious record of brilliant ideas, would refuse to hire you out of fear that you would make them look bad (whereafter they would resume snorting Bolivian marching powder).

  10. Anonymous Coward
    IT Angle

    @Adam 61 - constant sun & ibm squabbling

    --- Matt Bryant - the death of Sun did make roadmaps a lot more important to many in the industry, so Loughridge was probably trying to reassure IBM customers

    --- Kebabbert - Hopefully the 5 GHz Power7 will be several times faster than the Power6+ so it can match the 1.4GHz Niagara T2+

    --- Adam 61 - Stop raving on like an automaton about bloody Niagara, I think everyone has had enough of this to last them a lifetime (please back me up here guys!)

    Matt Bryant through the first punch here, with a half-truth jab at Sun (they are being acquired, not going out of business.)

    I tire at constant squabbling - Adam not reprimanding both comments was self serving.

  11. Anonymous Coward
    FAIL

    but it runs AIX...

    ... so therefore pointless. (yeahyeah, linux too, but why since its incompatible with few commercial linux apps? If expensive TCO gives you a rod on IBM will even let you run Linux on zzzzSeries)

    Much as I despise the poor true performance of Power chips (which as far as I'm concerned is more about who they make money for) I really don't give a rats as its the OS you see everyday and interact with. Sadly AIX tends to play up the weaknesses in the kit by being a total bag of spanners.

    Whats that? DIMMS failed again? Oh that'll be the microcode...

  12. Anonymous Coward
    Anonymous Coward

    It might come close

    http://www.oracle.com/features/sunoraclefaster.html

  13. Anonymous Coward
    Go

    @Adam 61

    "(please back me up here guys!)"

    Uh, sorry - I happen to think these comparisons are interesting, and Kebbabert certainly appears to have several times the technical nous of the average commenter.

  14. Anonymous Coward
    Anonymous Coward

    To David Halko

    David, Power7 will be engineered on single piece of silicon as well. Moreover, it is the first chip with eDRAM L3 memory on the same chip as 8 processor cores.

    Remember Cell processor? It's 9 core chip: 1 powerpc and 8 spe cores.

    It's worth to mention that IBM was the first who implemented multicore technology. Power4 was first dual-core chip in history.

  15. Matt Bryant Silver badge
    Happy

    RE: Kebabfart and Novatose

    "Hopefully the 5 GHz Power7 will be several times faster than the Power6+ so it can match the 1.4GHz Niagara T2+...." Kebabfart, your level of naivete is only matched by your astonishing willingness to display your ignorance. No-one is ever going to believe that a crippled, mini-SPARC design is ever going to compare with a proper, general-purpose enterprise CPU like Power for the kind of applications run in the real World.

    "It is nice to see IBM coming to the Octal-Core show!...." Novatose neatly ignores the fact that Intel's and IBM's efforts have real cores, capable of good individual performance, backed up by large cache, whereas Sun's Niagara is forced to use a crippled SPARC core and limited cache, with lamentable individual performance, especially with the typical heavy thread apps found in use today. It's a bit like Cesna standing up and congratulating Boeing on the launch of the Dreamliner, saying; "Now we both make airliners!"

  16. kit
    Pint

    Power8

    Shall we see IBM consolidating all of its P, I, and Z platforms on Power8 chips. Certainly, its a matter of time. If not the Power8, perhaps, it may be P9 or P.... whatever.

  17. Kebabbert

    Adam 61

    You want me to shut up talking about how slow the legacy Power6 and Power7 are, and how fast the Niagara is? Why? Yes, the Power6/7 are fast. But when you compare them to a REAL cpu, their weakness is revealed. Do you feel mad each time I remind everyone how slow the Power6/7 are, when pitted against a real CPU? You really want the Power6 to be the fastest, and I shake your world each time I talk about the Niagara?

    I feel sorry for you, but I am not going to stop. Why? Because I am fed up of Trolls and FUDers here down talking the Niagara. There are so many of them. And they often post things like "I work at a large bank and we love the niagara but now we are migrating to power6". It is always the same, they work at a large bank. Always a bank. And they always love the Niagara, but now they are migrating to Power. Always. It is the same theme. It is highly unlikely there are different persons writing and using the same expressions, on the same theme. I am fed up of this FUDers and liars. Therefore, I will not stop pointing to hard benchmarks of Niagara and Power so people sees that it is FUDs and lies. I will stop when there are no more FUD spreading here. Then I stop. Until then I will show with hard facts and hard numbers exactly how slow the legacy Power6 is. When the 5GHz octo core Power7 comes, I will point to benches compared with 1.4GHz Niagara and show how slow Power7 is. I will not post things that I can not backup, I will not post FUD ie things without evidence. I will post hard links and hard facts. Let me do that now, below.

    The difference between me and those FUDers, are that I post links to hard benches, and they FUDers can not prove their claim. For instance "the Niagara suffers from a small cache" - but the benches show the opposite! The best part is when some moron says "in my OPINION the Niagara suffers from a small cache". It is like "in my opinion this 2m guy is shorter than that 1.75m guy". Hard benches show a totally different picture.

    And there ARE morons here, claiming that "SUN is using carefully crafted benches that benefits the Niagara and punishes other CPUs". But all niagara benches are official, like TPC-C or Siebel V8.0. SUN has not created these benches. Only a moron claims that SUN has created the Siebel v8 benchmark, or the TPC-C benchmark.

    .

    Anonynmous Coward

    About that sucky CELL cpu from IBM you talk about. I will show with hard facts how slow it can be. I will not do like the FUDers do "In my opinion/onion the Cell is slow", "I work at a large bank and we love the CELL but now we are migrating to C64", "the cell is the slowest fhing out there and it is dead, no one should buy CELL", etc. Instead, I will post hard benches. That is the method prefered by mathematicians. To prove things.

    The 1.4GHz octo core Niagara is thirteen times as fast as the 3.2 GHz IBM CELL with 9 cores in pattern matching. In "IEE Computer" IBM implemented a pattern matching algorithm with lots of heavy optimizations for the Cell. And SUN just implemented the algorithm directly from book without optimizations and is 13 times as fast as the optimised Cell version. Pattern matching is used in finance, bio informatics and other fields. But of course, soon some moron will say "SUN invented pattern matching because it benefits the Niagara and punishes other CPUs", yes? Soon all existing benches will benefit the Niagara and punish all other, does that mean that Niagara is faster on everything then?

    Ive told you before, and I will tell you again: IBM does not have that good tech that IBM marketing says. And it costs cruelly much more than SUN, with better and faster tech, and for one fraction of the price, and also open tech. When Oracle increases the price on SUN hardware many times, then you will maybe see how cheap the SUN hardware actually was.

  18. Jesper Frimann
    Big Brother

    RE:Hopefully By Kebabbert

    Kebabbert.

    You should check the facts,

    1) The two benchmarked solutions actually run somewhat different mix of transactions, the p570 runs more Partner Relationship Management transactions which, judging by the response time are the heaviest.

    2) You have to adjust for utilization of the machines as stated. The power 570'es run at low utilization on DB and Web, where as the T5440 runs at 80+ utilization.

    3) The T5440 consumes quite a lot of memory 94 versus 46 GB

    4) Opposite the recent TPC-C benchmark, then the response time for the p570 are much faster than the T5440

    5) The T5440 runs more or less in a 2-tier mode, where as the p570 uses physical network.

    if you try to adjust a little due to 1) then I get to the fact that the T5440 is aprox 2.4 times faster per chip on this particular benchmark. And this is what the T5440 does best, webserving, light application serving and a small database that can run on a few cores in this case 3 cores/12 threads.

    // Jesper

  19. Anonymous Coward
    Anonymous Coward

    whats all the fuss!

    whats all the fuss about, its not like Power7 will have a better fater or improved Altivec SIMD.

    its not like you can go down your local PC shop and buy a Power whatever motherboard, and chip , take it home and build the new cluster of PPC PC and slap linux on there.

    hell you cant even buy a resonable PPC SOC (system On a Chip) with all the generic bits you might want onboard and off the shelf and available TODAY at a good speed, quad core 2 gig,quad Altivec SIMD,64M L2, 512M L3, wirelessN, HDMI, openCL SOC HW assisted high speed Encoding/decoding for your mass use for any No of products etc.... and/or lower power (less than 20W) at GOOD freqs 2gig+

    its simple, you dont make it, supply it, or sell it, we cant buy it on mass and make your profits... make a PPC SOC at 2Gig with the above and OSS SW that uses the SIMD in its updated core linux librarys (PPC linux dosnt use ANY altivec optimisations EVER for instance, PPC linux GlibC) put the SOC package on a generic socket and produce cheap generic super small motherboards of varing sizes down to mobile phone sizes and we might be taking for once.....

    i want a lower power mobile phone sized PCB with all the video and related goods onboard that can be run off a cheap PV pannel and small battery pack on my window sil, one day i may get it, but it loks like the far east PMP vendors and innovators will be that OEM mass supplyer to the world, not some US mega corp be it IBM or SUN, as your all to slow giving the world what they want to buy on mass....

  20. Matt Bryant Silver badge
    FAIL

    Kebabfart and Cell - a warning for us all.

    And Kebabfart has provided yet another stunning example of the typical Sunshiner blindness to the realities of the business. He is slagging off Cell, which has probably outsold the whole SPARC range many times over just from Playstation alone, because he thinks he preaches from a position of technical superiority. Yet he can't see the fact that the Cell is an enourmous commercial success, that it made big profits for Sony and IBM, whilst Niagara failed to make enough profit to keep Sun afloat and independent. Like many Sunshiners, he has yet to realise that being clever is just simply not the point, that making profits is, beacuse profits mean you stay in business. It's this kind of technical snobbery that makes Sunshiners snear at the hp printer bizz, though that printer bizz regulalry makes a profit larger than the whole Sun hardware range. Remember how they used to slag off Intel, right up to the point when they bought an x64 company to try and prop up their server business?

    How do people end up so blind? It's called a sucker sell, and it builds on the old feature sell. The salesman convinces you that his tech is really smart by hyping one or more features unique to their product - the old feature sell - but adds that the sucker is also really smart for seeing how smart it is, so smart that only an idiot would not see how smart it is. Once converted, the sucker will then do the salesman's job for him, telling all how product X is the bee's knees. It works because it is human nature to like to think you are clever than others. To face reality and see that product X actually isn't that good becomes harder and harder, as to do so would be to admit he's actually not as smart as he thought, that he got suckered, and nobody like to think they've been suckered. When evidence to the contrary is presented, the sucker ignores it or stridently denies the reality.

    Is the sucker sell new? No, it's been around for decades, an extreme example being Hans Christian Andersen's "The Emperor's New Clothes". I'm not saying Sun's Niagara are a con or just imaginary products, just that Kebabfart is just like the emperor's courtiers, repeating the fallacy that the Emperor's clothes are just beautiful becasue they don't want to admit they can't see them (though, Rock fits the invisible clothes analogy perfectly, and I'm sure some readers will remember how Sunshiners like Kebabfart and Novatose were happy to tell us how wonderful Rock was, right up to the point it got canned). Kebabfart's case is especially amusing given that he has previously admitted his comapny don't even use Niagara but do use IBM's servers, so it is impossible for him to pretend his opinions are from personal experience!

    So, how do you defeat the sucker sell? Easy - shootouts! If the salesguy says product X will do better than a competitor's product, make him prove it (at his expense) in your environment with your test data, in an head-to-head with the competition (also at their expense). Don't let him dictate how it should be tested, insist it is as close to the expected production environment as possible, and then see if his hype holds any water. If it does, winners all round. If it doesn't, then you can pat yourself on the back for avoiding what would have probably been an expensive mistake.

    Do I think Power7 will be released? Yes, because IBM have done well with previous Power generations, and there don't seem to be any massive technical hitches left to overcome before it's release. Do I think it will be a success? Yes, because IBM have a good marketting machine with plenty of budget, lots of services to offer, are trusted, already have a large installed base, and have previously done well even wehn they didn't have the best offering. Do I think it will be more of a commercial success than Niagara? Yes, because I think it will be better suited as a general purpose CPU to the kind of enterprise applications in the market, whilst Niagara is too niche (and if it wasn't then they wouldn't need SPARC64), and becasue Power will still have the relatively secure mainframe market to help fund the push in the UNIX market. Which CPUs do I see as Power7's competitiors? Tukzilla and SPARC64, not Niagara. Niagara's real competition are Xeon and Opteron.

    /Point, laugh, repeat.

  21. Adam 61
    Coffee/keyboard

    RE: AC Sick of Squabbling & Kebbabert

    Me too hence the post and it was hardly self serving, I'm neither a SUN or IBM fan really, pretty agnostic in the scheme of things.

    True MB does throw the first one here but at least its with some semblance of a sense of humour (unfortunately Kebbabert hasn't got one of these from the evidence I've seen on various debates). He says he is fed up of trolls and fudders, I think he is turning into what he claims to loathe.

    Why would any sane person claim that people who have opinions are morons, everyone is entitled to have one, just because it doesn't agree with your veiwpoint doesn't mean they are a moron. What it does make is the poster a very small minded individual with no concept of actually holding a debate. Propoganda is not a debate!

    And as for feeling sorry for me then that had me in stitches, I have a life thanks. It looks like Kebbaberts consists of reading Niagara benchmarks and Solaris user manuals.

    The reality of the CPU benchmark is that it really doesn't matter. IT is increasingly controlled by the bean counters and they only really care about reducing risk, increasing profits and reducing cost of ownership.

    Unless you can translate these features into true business benefits then good luck selling it.

    Cue various unsubstantiated accusations of being a liar/fudder etc etc ad infinitum

    Truly guys its getting boring!

    Adios for now

  22. Anonymous Coward
    Dead Vulture

    RE: AC Sick of Squabbling & Kebbabert & Matt Bryant

    Adam 61 --- Me too hence the post ...True MB [Matt Bryant] does throw the first one here but at least its with some semblance of a sense of humour (unfortunately Kebbabert hasn't got one of these from the evidence I've seen on various debates). He says he is fed up of trolls and fudders, I think he is turning into what he claims to loathe.

    Dealing with people like Matt making unsubstantiated claims and predictions, which have little claim in reality, about the future of companies (he obviously carries a grudge against) decreases the credibility of his occasional more accurate assessment.

    MB's troll-like behavior incites more people than just Kebbabert. It really destroys what could otherwise be a very pleasant community and drives people away who could really bring insight into the industry.

    In the end - it leaves The Register with a fewer and poorer quality readership and participants. The Register is merely cutting their profits by allowing it to continue - a good quality staff must be paid for.

  23. Kebabbert

    Cell sucks on certain workloads:

    I forgot to post the links that show that 1.4GHz Niagara is more than 10 times faster than a 3.2GHz Cell. Here it is:

    http://blogs.sun.com/bmseer/entry/extremely_fast_pattern_matching_on

    .

    Mattie Pattie Laddie, it doesnt matter how much the CELL sells. You see, CELL is a commodity product for the masses and the Niagara is for servers. The market is smaller for servers. But again, you fail to notice that you can not compare them. Out of ignorance, or out of sheer dumbness. It doesnt matter what you say about the CELL, the true facts are that the CELL is not that almighty fast as IBM tries to say.

    I agree with one thing, that IBM has marketing muscles so IBM can fool anyone.

    And I am still astonished by your statement somethinge like "In my OPINION the Niagara is slow"!!! And you still say Slowlaris, despite Solaris and Niagara smokes Power6 and AIX which benches show. But of course, it is your OPINION that SUNs products are slower. And also, that 2m guy is shorter than that 1.5m guy - in your opinion. Nutcase. How the heck can your opinion be that a 2m guy is shorter than a 1.5m guy? Do you have a large machine helping you to breath?

    .

    Adam 61

    If you are fed up with me talking about the Niagara vs Power6 and CELL, do you think the SUN supporters are fed up with the FUDers here? So it is ok they morons continue down talk SUN here, but no one else is allowed to point to benches? Are you serious???

    And I dont buy your claim that you dont support any OS. Why would you react when someone tries to prove that Niagara is not slow, contrary to what some FUDers say? Am I not allowed to defend SUN and dispel the FUD? No, I am not? Why do you care, if you dont support any OS? Ergo, I dont buy your claims.

    I dont see why I should stop point to benches Niagara vs Power6, when Mattie Pattie and the others continues. You allow them to continue, but try to stop me. And you claim that you dont support any OS, yes? And you also claim the moon is made of cheese, right?

  24. Jesper Frimann
    Big Brother

    RE:Kebabbert

    "I forgot to post the links that show that 1.4GHz Niagara is more than 10 times faster than a 3.2GHz Cell. Here it is:

    http://blogs.sun.com/bmseer/entry/extremely_fast_pattern_matching_on"

    Sorry no points for quoting BMSEER, this is SUN FUD central, And you get no points for quoting HP the real story or 'IBM's pseudo business technology statements' either.

    You write that it might not be right to compare Cell to Niagara. And i agree the first is a HPC/gameconsole processor, the later is a Webserver/light application server product.

    Again this is a cherry picked benchmark.

    What our good friend BMSEER has done is to point to the place where Cell doesn't perform well on pattern matching. Actually for small dictonaries a single cell, using it's SPE's will do around 41 Gbit, which is 1.7 times what the Niagara processor will do.

    But this isn't really mentioned. Here is a presentation connected to the paper that BMSEER referes to:

    http://sti.cc.gatech.edu/Slides/Petrini-070619.pdf

    So again this is a case of not telling the whole story, which is that For small dictonaries Cell is faster for large ones Niagara is faster. And that there is quite a good reason for this, the two processors are targeted at different markeds/workloads.

    But cause you are picking a slam dunk x10 benchmark from a SUN FUD site then you get burned.

    // Jesper

  25. Adam 61
    Stop

    Re:Kebabbert

    I am allowed to be agnostic if I like, its about having an open mind and not being beholden to any particular IT company.

    It means I can be impartial when advising people what best fits their requirements rather than shoehorning in a solution that will leave them with a sour taste because its the only thing I can talk about even when its not appropriate. Its what makes me successful in my job and people value my opinion. You see I can appreciate that for some use cases Niagara is appropriate, the same for Power, the same for Xeon/AMD/Itanium, it all depends what they want to do.

    The reason I would like you to shut up and the reason I reacted is because you have one arguement, thats the sum of it, you have made your point (repeatedly) and its becoming tedious. Imagine listening to the same song on the radio ad infinitum and thats what you are like. You could have the best arguement in the world but I for one have tuned out to it.

    As far as I'm concerned anyone is entitled to their opinion (its not a facist state after all) providing they can back it up with reasoned debate. In fairness to you, you do attempt to do this where not all others do, but you can't go around calling everyone liars and morons where they have a conflicting view based on their personal experiences.

    Who are you to call them a liar, have you worked in their environment? No you haven't.

    Have you any non benchmark related real IT experience? its impossible to tell as you only talk about benches.

    Yes you are entitled to your opinion, all i am suggesting is you bring something new to the table

  26. Anonymous Coward
    Unhappy

    @Adam 61 - at least Kebabbert & Jesper Frimann post citations!

    @Adam 61 - Who are you to call them a liar...? Yes you are entitled to your opinion, all i am suggesting is you bring something new to the table

    Kebabbert & Jesper Frimann post citations when they levy their opinions.

    I have yet to see you take Matt Bryant to task when he does not cite verifiable references.

    Try being more consistent.

  27. Kebabbert

    Mattie Pattie Laddie

    How does it matter how much a consumer CPU as Cell has sold, compared to a server CPU as Niagara? Does this make Cell faster than Niagara? I dont get it. We are talking about performance, and if the Niagara is so slow as you claim. I try to show that you are wrong on this by posting benchmarks. But still you continue to spread FUD about how slow the Niagara is, despite it has several world records, including TPC-C and SIEBEL.

    If you can prove that a consumer CPU as Cell has sold more than server CPUs, does that prove that the Niagara is slow? Does that prove you are right? No. Maybe I should change tactic, and do exactly as you do and spread FUD, instead of posting links to benches? I dont get it, why do you continue to spread FUD regarding the Niagara, which forces me to dispel your FUD? Why are you allowed to spread FUD, but it is not allowed when someone tries to prove that you are not correct? At least I can apologize and accept corrections if I am wrong - whereas you are not.

    Regarding the Cell and Niagara, I must post links to my claim that Niagara is more than 10 times faster. Or I could be lying. Here is the proof.

    http://blogs.sun.com/bmseer/entry/extremely_fast_pattern_matching_on

    .

    "....Kebabfart's case is especially amusing given that he has previously admitted his comapny don't even use Niagara but do use IBM's servers, so it is impossible for him to pretend his opinions are from personal experience!..."

    Actually, I dont get this one. I have never claimed that my company uses IBM servers. Can you prove your false claim, then post that link. But my fortune 500 company in finance, uses Solaris servers. And are migrating to Linux. We dont use IBM servers, nor Power6 cpus. Why do you claim that?

    I dont get it. Why are you allowed to write false things in many posts, but I am not allowed to correct you so the readers get a correct view? If you would write that Niagara costs 1 million USD, and I correct you - then that is not allowed? But you are allowed to write whatever you want?

    .

    And I still dont understand how you can write that "in your opinion the niagara suffers from a small cache", when hard benches and numbers show the opposite. Do you also write "in my opinion that 2m guy is shorter than the 1.5m guy"?

  28. James Billingham
    Flame

    Seems Jesper has caught Kebabbert out

    Quoting biased benchmarks will never convince anyone - reading both the Sun reference from Kebabbert and the paper it originated from that Jesper quotes I could write my own blog post at IBM central stating one CELL BE with 1 core beats Niagra with 8 cores in this very important area of pattern matching ... blah blah blah

    Not quite true as CELL is only better in small dictionaries, it has SPEs which although not cores like the PPE should obviously be mentioned. Sun ref. states CELL has 8 cores - which is as untrue as my statement, one PPE which is a cut down Power core and 8 SPEs which are by no means general purpose CPU cores and as such should not be compared to the Niagras 8. (Sun thinks the 1 PPE and 8 SPEs adds up to 8 - I assume its because only the SPEs were 'doing' the work in the academic article, in fact the PPE orchestrates everything so maybe should be included to make the CELL look worse)

    Having read both I'd have to conclude the Sun article is a peice of marketing FUD, they specifically chose the particular area of this benchmark that the CELL is bad at and ignored the bit where it is a lot better. They exaggerated the number of cores in the CELL which indicates a lack of understanding of the CELL, especially as they compared an SPE to a Niagra core. Worst thing is using the academic article referred to by Jesper and not having a reference so deluded Sun fanbois like Kebab never see the full story....

    All fairly irrelevant any way - I work in the real world of performance (For IBM so I must be biased, actually a lot of IBMs customers use CPUs other than Power!) tuning customer systems and the CPU technology is rarely of interest (I don't get sales commission ;-)). Real applications perform badly because of bad code - order of magnitude performance improvements are not uncommon by fixing the poor code in the app. Architectural decisions, library decisions (I choose JSF for my high volume website!) etc have far far more effect on performance. Except for that customer who decided to run everything on a 486... But thats a different story.

    The rare optimised workloads that run on specialised processors like CELL, or the highly optimised cluster apps are rare and they probably do care more about the CPU tech specs - I'd guess they don't take BM Seer as a credible source. Interestingly IBM has far more workloads in the top 500 supercomputers than anyone else - so maybe they are good at that, I don't know cos I don't work in that area.

  29. Adam 61
    Thumb Up

    Re last AC

    I agree and I have credited them for that if you read my post, however due to the nature of peoples jobs its not always something thats referenceable due to confidentiality. Unfortunately people do use this to their advantage to promote misinformation but thats the nature of free speech.

    Citing benchmarks is fine - but how realistic are they in the context of what people do on a day to day basis? If all we can point to is super tuned benchmarks that don't use real configs that people can deploy in the real commercial world then how are they more relevant than someones opinion based on experience? Of course one is subjective and the other objective and thats the fundamental difference.

    One point Mr Bryant consistently makes is that the best thing to do is bench real configurations with your app stack and make your own mind up. To me thats a sensible point of view as it means you aren't getting false expectations based on a benchmark for a server that is far removed from what you'll actually deploy.

    However I also tire of his constant references to "Slowaris" which is just as tedious as Kebbaberts Niagara monologue.

  30. Jesper Frimann
    Big Brother

    Adam 61.

    "Citing benchmarks is fine - but how realistic are they in the context of what people do on a day to day basis?"

    Well you are absolutely right, and well.. the problem is that much sizing information is actually derived from benchmark results. If you for example take the SAP sizing data you can get from different hardware vendors then it is usually based upon SAP SD benchmarks. So in my day to day I indirectly use benchmark information, all the time.

    And we don't have the time to actually go and try things out, which certainly would be the best, we do a qualified guess and then add a little more MEM and CPU and see how it goes, we can then always deduct a little afterward.

    "However I also tire of his constant references to "Slowaris", "

    Well that is a part of the game. I always use the therms SlOwARIS, HPsUX, AIntuniX and Toonix. And not to forget Wintendo. Then you are sure you get to insult everyone, which is a very danish way of doing things, you start by insulting people and then you insult yourself to put you on the same level as them. Strange yes yes.. I know.. :)=

    And well a bit of teasing is ok, I mean it's not like people are seriously saying that Solaris is not a Unix, or not a fairly good one. Ever heard the amount of sh*t that AIX admins have to take about AIX, specially from Solaris admins ?

    If Slowaris should be an insult then it should be to the SPARC platform.

    // Jesper

  31. Kebabbert

    Jesper Frimann

    Even if I dont agree with you, I enjoy discussing with you. Because you are well educated and no troll. You back your claims up with links, and we can discuss around your links, and your explanations. That is academic. Matt Bryant, OTOH, just posts his opinions, which he has explained. He explained that "in his opinion, the Niagara suffers from a small cache". Well, it is a difference between facts and opinion. Never should you try to disguise your opinions as facts, as Matt Bryant does. You distinguish between your opinions and facts. I wish more people discussed like you do.

    .

    "...Sorry no points for quoting BMSEER, this is SUN FUD central..." You mean that because BMSEER posted this benchmark, the benchmark is not valid? Why is it not valid? Is is a fake? Has BMSEER made up the numbers? You mean the true numbers are maybe half of what is reported? Or? I dont understand. Even if IBM reported a benchmark, I would accept it. I dont expect IBM to lie about hard numbers?

    So, what do you think the true Niagara numbers are for string pattern matching? BMSEER reports that one 1.4GHz Niagara achieves 24.6Gbit/sec and two 3.2GHz Cell achieves 3.8 GBit/sec. Do you mean in reality one Niagara reaches maybe... 1GBit/sec and the two Cell reaches 50Gbit/sec? Because BMSEER is from SUN, he is lying about the true numbers? Could you elaborate on his lies? Exactly which number is a fake?

    .

    "....Again this is a cherry picked benchmark...." So what? Every solution will have it's strength and weaknesses. If IBM picks a benchmark that shows Power6 strengths, I would be dumb if I dismissed it as fake? I mean, the numbers are real. The Power6 got that performance. Why would I dismiss the numbers? I am trying to show that Niagara is not as slow as FUDers say. There are things where the Niagara is fastest on the planet. Of course, the Niagara is not fastest on everything, it has it's weaknesses. But is also has it's strengths. But when people try to show the Niagara's strong sides, you just dismiss them. About Power6, I know it has it's strong sides, I dont dispute about that. But people FUDs about Niagara has no strong sides at all, "it is just a weak CPU in general". I say "Wrong! It has strong sides". But they are just dismissed. I dont dismiss the Power6 strong sides, but you dismiss the Niagara's strong sides? Fair eh?

    .

    "...What our good friend BMSEER has done is to point to the place where Cell doesn't perform well on pattern matching. Actually for small dictonaries a single cell, using it's SPE's will do around 41 Gbit, which is 1.7 times what the Niagara processor will do..." Ok, it looks like one 3.2GHz CELL achieves 70% better score on small dictionaries. If I were you, I would have said "no, it can not be true, because IBM published that paper". But I do not. I accept that number. Instead I think that small dictionaries shows that CELL doesnt scale to larger realistic loads when the performance drops to 1.9 GBit/sec, that is 96% drop of performance, which in my eyes is terrible. And besides, who is interested in small loads? I mean, all realistic true life server workloads are large loads. Who is interested in small things that no one can use? Even the Power6 would perform well, if the load was sufficiently small enough to fit into it's cache. But CELL performs good only on small loads, not interesting for real life. Whereas Niagara has again proven that it does not need a large cache to perform superior on large heavy server loads. Mattie Pattie Laddie, here you see again that despite having a small cache, the Niagara does not suffer. Instead it smokes all compettion on large realistic loads.

    .

    "...So again this is a case of not telling the whole story, which is that For small dictonaries Cell is faster for large ones Niagara is faster. And that there is quite a good reason for this, the two processors are targeted at different markeds/workloads..."

    But who is interested in small loads? The difficult part is large loads, real life, realistic loads. And which solution should you choose then? For real life, realistic loads, you should choose Niagara. Crystal clear. Niagara wins again. If you have small loads, that no one runs in real life, then choose Cell. Or choose Niagara and you can use it for small loads AND large loads.

    .

    James Billingham

    "...Sun ref. states CELL has 8 cores - which is as untrue as my statement, one PPE which is a cut down Power core and 8 SPEs which are by no means general purpose CPU cores and as such should not be compared to the Niagras 8..."

    "...Having read both I'd have to conclude the Sun article is a peice of marketing FUD..."

    So you mean that CELL does not rely on it's 8 cores to do work? And therefore you should only count CELL as one core, the powerpc core? Cool. So you mean that the powerpc core alone, by itself, achieves 41GBit/sec on small loads? Or is it so that the CELL uses all 9 cores to achieve 41GBit/sec? So when you talk about the CELL you describe it as "multi core CPU having 8 cores" or as "single core CPU having 1 core"? What I am trying to say that you can describe the CELL as single core if you wish. Me, will describe it as 8 core + 1 core = 9 core CPU, just as SUN does. And that is a more truer description of CELL. And that is not FUD, neither from me or from SUN. Even IBM describes the CELL as multi core CPU. I must say that you IBMers have a strange view of what FUD is. "SUN stating CELL is multi core CPU is FUD" - yeah right.

    So you mean that when IBM claims one Mainframe can consolidate 1500 x86 servers are facts? It turns out that IBM assumes the x86 servers to all idle and the Mainframe loads at 100%!!! So I can claim my laptop can consolidate lots of servers, if they all do nothing - just as IBM does. This is just lies.

    IBM also claims that Power6 has 250GB/sec bandwidth. Then IBM adds all bandwidth in the CPU! That is a lie. If there is a bottleneck on 10GB/sec, the chip will never be faster than 10GB/sec. You can not add all bandwidth, that will never be achieved.

    Or when IBM claims "one Niagara core is slower than the Power6 core, ergo the Niagara CPU is slower than the Power6 CPU". That is also lies. Assume the worlds fastest CPU has 10.000 slow cores. Can you then say "one core is slower than Power6 core, ergo, the cpu is slower"??? No. That is just lies.

    I must again say that IBM has a strange view on what FUD is.

  32. Anonymous Coward
    Happy

    Here we go ....

    ................again and again and again and again and again and again :)

  33. Jesper Frimann

    RE:Kebabbert

    I'll just repeat myself:

    "So again this is a case of not telling the whole story, which is that For small dictonaries Cell is faster for large ones Niagara is faster. And that there is quite a good reason for this, the two processors are targeted at different markeds/workloads.

    But cause you are picking a slam dunk x10 benchmark from a SUN FUD site then you get burned."

    I don't dispute the SUN benchmark results. I just point to the fact that they aren't telling the whole story, but cherry picking where they are best.

    "Sorry no points for quoting BMSEER, this is SUN FUD central, And you get no points for quoting HP the real story or 'IBM's pseudo business technology statements' either."

    Jup, I have no problem quoting SUN/HP/DELL/IBM/Oracle technical sites, there are a lot of serious good people working at all these companies.

    But BMSeer is a FUD site just like the others I mentioned. Sorry.

    // jesper

  34. Jesper Frimann

    RE David Halko on Ellison whips out his Sparc TPC-C test

    Damn, thread "Ellison whips out his Sparc TPC-C test" closed and I just used time to write up an anwer to David... well he's on here so here it comes out of context:

    David Halko wrote:

    "Good applications, however, are not always available for the architecture that may suit it best. The market takes care of the latter situation, over time."

    Look, SUN tried to jump ahead of the other server vendors by using a lot of threads, and they will continue down that road. Victoria falls has what 128 threads per chip, that's just as many as many as a M8000. So to get the maximum throughput out of a single chip you will have the same thread complexity, with locking, memory use, monitoring and amdahl law issues, as you have on a complex Big Tin server, which is many times faster.

    "I covered the main drawback above in a previous quote. I can add additional arguments (production line manufacturing of a single-chip solution is normally less expensive than an MCM - the consumer market size drives the most profitable implementation.) The AMD-Intel battle for quad-core illustrated it best, in the most recent near-term. If I need me to explain this, I can. I am really not trying to pick an argument - technology is what it is and I appreciate it as it is."

    No your original comment was "Some may argue that the MCM's were innovative engineering, disagree and suggest MCM's are a pragmatic business-man's short term solution to a technical problem." That is not "technology is what it is and I appreciate it as it is.".

    And I'll just repeat myself. It is not either CMT or MCM, you can use both. Sure it is faster, cheaper and better to put it all on one chip, but when you've filled the chip up, then an MCM's is next best thing, rather than having to use more sockets. From what I've read here at 'The register' then the next POWER7 will just as POWER4 use MCM's to make high density servers with 2 socket servers that houses 2 MCM each with 4 chip that each houses 8 Cores. Now that will give you 64 cores in a 2U box. Not a general purpose server btw, but a HPC box. And witout MCM's I doubt that that box would exist.

    If you want to say that something is business man like, then it is that IBM _didn't_ make an MCM version of the POWER6, it's not like there weren't enough memory bandwidth in the sockets.

    "Sacrilege is not the issue, accuracy is. Also, Solaris and SPARC are historically based upon community efforts, as Open Communities, which have been guided by Sun, external companies, and external organizations - so inaccurate information used in a slanderous way offends many people who invested their university, research project, masters, phd, and/or life works into it. Offense is to be expected when inaccurate information is used to slander large groups of people."

    Yes I know, at the university I attended, VAX and HP was for the students and SUN was for the researchers and PHD's. But it still doesn't change that it's bloody anoying when you get people in projects that base their decissions on feelings and not cold facts. This does not have to Solaris followers, but might as well be mainframers or Vaxers or windowers or Linuxers.

    "Some would suggest that the behavior merely objectifies a group of supporting individuals and slanders their life work. There are lots of people who feel they can slander groups of people calling them "blood sucking..." (fill in the blank) - it is just the same behavior. It shuts down inter-group dialog instead fostering healthy inter-group competition and later teamwork."

    Well you don't live in Denmark now do you. Here insulting/making fun of people and then yourself (or their work or..), is kind of of a accepted way of bringing people to the same level. Rather than praising others and then yourself almost as much, you 'make fun' of the other and then yourself. Like 'What an ugly dog you have, but it's not as ugly as mine.', could be a good ice breaker. Odd yes I know, but quite true.

    "(i.e. OpenOffice, VirtualBox, NFS, Lustre, Xen, etc.) When you consolidate it all together, Sun contributes a lot."

    Then don't call it linux, OpenOffice VirtualBox .. are all applications that run on pretty much every platform, or can be made to do it. Now if you had said that SUN offers up opensource great applications, I would just have agreed with you.

    I use virtualbox myself and absolutely love it, for it's simplicity and usefullness. I run OpenOffice and I like it, I sometimes get a little fustrated cause it isn't as stable as it used to be. I feel very strongly for open document standards and is actively engaged in promoting these.

    But these applications are not a part of the operating system called Linux. Period.

    // Jesper

    Again Bigbrother cause something is wrong in the state of Denmark.

  35. David Halko
    Happy

    RE David Halko on Ellison whips out his Sparc TPC-C test #

    Hi Jesper,

    You know, I think we are just about done with this line of discussion - I think there is one last piece left in regard to the conversation from this thread which carried to this commentary:

    http://www.channelregister.co.uk/2009/10/12/oracle_sparc_tpc/comments/

    David earlier posts, "I don't think I made fun of multi-chip modules, at least I didn't try to. I merely said it was a business-man's approach. MCM's require less engineering, get faster to production, uses older technology effectively, reduces risk, but does not perform as fast if engineered it onto a single piece of silicon."

    David later posts, "Doubtless, there are many benefits to multi-chip modules, as well as drawbacks"

    Jesper posts, "And the drawbacks are ?... I suspect that the only reason you say it's not innovative is cause SUN doesn't do MCM's."

    David later post, "I covered the main drawback above in a previous quote. I can add additional arguments (production line manufacturing of a single-chip solution is normally less expensive than an MCM - the consumer market size drives the most profitable implementation.)"

    Jesper posts, "No your original comment was 'Some may argue that the MCM's were innovative engineering, disagree and suggest MCM's are a pragmatic business-man's short term solution to a technical problem.' That is not 'technology is what it is and I appreciate it as it is.'"

    I copied the relevant posts, in the order of my making them, into this message. Let me highlight the first example of a disadvantage that I mentioned:

    "MCM's ... but does not perform as fast if engineered it onto a single piece of silicon."

    I added an additional drawback, since I could still not understand what you were trying to get at.

    "production line manufacturing of a single-chip solution is normally less expensive than an MCM - the consumer market size drives the most profitable implementation."

    I am not sure what you are trying to get at, honestly. My discussion was surrounding how multi-chip modules were a pragmatic solution based upon benefits and drawbacks that required less innovation at the engineering level than a single chip solution.

    This does not mean that I believe innovative ideas are devoid in multi-chip modules. I did not mean my position to be mutually exclusive, merely a generalization to give an impression of my leaning.

    Have a good day Jesper!

  36. Jesper Frimann
    Big Brother

    RE: David Halko (I give up.)

    If you can't see it, you can't see it :)=

    // Jesper

This topic is closed for new posts.

Other stories you might like