back to article Sun's Sparc server roadmap revealed

For those of us on the outside of Sun Microsystems, the future of the Sparc processor and its related server platforms has been the subject of much speculation and debate. But for Sun's largest customers, the mystery has been over since sometime in June. That's when Sun's execs made the rounds to the biggest Sparc shops in the …

COMMENTS

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

    Sure

    Sure. Picture Donkey from Shrek singing "I believe I believe"

    So JupiterE (because they are scared to call it Jupiter++) will still be 65nm in early 2011, and at 3Ghz. Will it be water cooled ? And they are scaling from 2.52Ghz to 3Ghz in 2 years ( both are 4 cores with 2 threads each) over a period of 2 years. Look out Speedy Gonzales coming thru...

    This sounds like the story, we canceled UltraSPARC V to bring out ROCK. Now we canceled ROCK to Yosemite Falls. Rinse, Repeat. Just buy our stuff.

    And Intel will be at 32nm by then ? at 1/3 the price ? (ok only to 8 sockets with Nehalem EX, but hey 8 sockets 8-16 threads per socket, and runs every OS, hey even Solaris, if Larry says that's ok.)

    Whatever.

  2. Anonymous Coward
    Anonymous Coward

    No IBM, HP, Intel or AMD?

    So Timothy, do me a favor. Get the same type of roadmap from IBM, HP, Intel (x86 and Itanium), and AMD and post it up here. It would be interesting to see them all.

  3. John Savard

    Probably no hope

    I saw one interesting thing in all this gloom and doom. They're making, for Fujitsu, a version of the Sun with advanced vector capabilities for supercomputer use. If such chips were to become more generally available, it would be a way to provide serious performance competition to Intel.

  4. Anonymous Coward
    WTF?

    Crazy Larry

    "Crazy Larry, his prices are IN-SA-A-A-A-A-ANE!" WebLogic up 47% after the acquisition and you can expect more of the same with the Sun gear.

    No wonder Oracle is putting out desperate Wall Street Journal ads.

    Larry decides to pick up Sun for a paltry $5B then finds out it is in total disarray and the m-class line declines 58% in just one quarter. He also has to wait till next year to finish the deal and beg to Euro trash. All he wanted was to monopolize Java, kill MySQL and milk all the Solaris customers.

    www.sunsucks.com

    That's gotta Hurd

  5. Steve McPolin
    Paris Hilton

    sad

    Other than AC, nobody else can even bother to post about this. Maybe there should be a run-off of all the dead architectures so we at least can know who really came in last.

    Paris; because it really speaks to how pathetic it is.

  6. Matt Bryant Silver badge
    FAIL

    Just to remind the Sunshiners.....

    I know you're all going to start singing the revised Ponytail hymns now you think there is a future, but you really need to realise this is just a Sun marketting ploy to try and boost sales witha little false customer confidence. The key is there right at the end of the article:

    "......All of this is subject to change, and some of it most certainly will once Oracle takes control of Sun."

    This is Sun (the defunct, about to be bought, and not incontrol of their own future) admitting that even in the best of cases this is all they can do, it is not from Oracle. It is not a promise from Larry, it is not an Oracle roadmap, at best it is valid only for a few months if Oracle get clearance to complete the purchase. Then Larry can turn around and retract as much of it as he likes. Well, what little there is left to retract - most of the roadmap is taken up by chips that have already been and failed! So just keep that in mind before you start retreading your old mantra of "only Solaris on SPARC", the rest of us are too busy pointing and laughing to listen.

    /Enjoy!

  7. -tim
    Thumb Down

    Wrong direction for happy customers

    A 45nm Sparc IIIi would outrun all their current hardware by far for most unix apps that their old school clients depend on and be much cheaper than all these other options. It could even run old versions of their OS so there would be no need to port things.

  8. Anonymous Coward
    Gates Horns

    Kebbie Fairy Tales

    Once upon a time, there was a boy called Kebabbert, who used to have a pure heart.

    You could see his sparkling eyes when Sun mages presented the future for him.

    Sun mages poisoned his head, so that he only dreamed about the future.

    Once more, he is dreaming. The future looks promising again...

    Now that besides Sun mages there is an Oracle...

    He smiles, there is a future. He can dream once more...

    Kebbie, wake up man!!! Take the RED pill!!! Follow the rabbit!!!

    Cheers!

  9. WinHatter
    Pint

    wait for it

    Larry the pull cable guy may well pull the sparc plug altogether.

    65nm in 2011 means they are not investing in manufacturing processes so wether the chips are priced at $1 a pop and that may work or ... they don't have a roadmap at all.

    If I were a Sun customer I'd move, even Windows 7 is not as scary as this roadmap. Ooops I forgot, I was ... RedHat is me buddy now ... oh the irony!!!

  10. Morten Bjoernsvik
    FAIL

    just die-shrinks, no real development

    This is just shrinks based on new lasers with shorter wavelength. It is as hard as taking the original candence-file and shrink it to the new size-factor. There may be some out of sync pulses that needs to be fixed because of shorter transportation, but this is no hard work.

    This kind of development can be sustained by 2-3 people.

    The investment is in the die-factory (Texas Instruments and Fujitsu), Sun/Oracle has to

    guarantee a lot of cash or minimum production quantity to get this going. Or share the

    production with someone else, which means far longer from cadence-file to real chip.

  11. Anonymous Coward
    Anonymous Coward

    Interesting...

    You've taken a slide which is clearly marked Copyright and placed it for all to see.

    Can't wait to see what Sun's Attack Lawyers do with this one.

    Good to see Pratt Bryant the Mark Turd-fanboi back. He we conspicuous by his absence during the HP results being posted.

    Typical coward.

  12. Kebabbert

    Mattie Pattie Laddie

    "This is Sun (the defunct, about to be bought, and not incontrol of their own future) admitting that even in the best of cases this is all they can do, it is not from Oracle. It is not a promise from Larry, it is not an Oracle roadmap, at best it is valid only for a few months if Oracle get clearance to complete the purchase. Then Larry can turn around and retract as much of it as he likes."

    Yadda yadda. Of course Larry will turn around and kill Solaris, SPARC and also kill Java - according to you. And all SUN products. You would be very happy then, yes?

    Or, you could just listen to what Larry himself says?

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

    Sorry to see that your dreams and reality doesnt mix together very well. You have a somewhat strange view of things. Do you still believe that a small cache punishes performance, and it is impossible to get high performance? How many times have I told you that you are wrong on this? As many other things, no doubt. For instance, you are wrong in this thread too.

    .

    AC

    I dont take the red pill or follow the rabbit. Do you really want to know why I support SUN instead of for instance, IBM? Would you like me as an IBM supporter? The reasons I support SUN are

    1. SUN is open. Open technology is good for everyone. IBM is closed. AIX is closed. Mainframes are closed. etc. Admittedly IBM contributes to Linux, but Linux sucks. It is not a option for a real Unix.

    2. IBM FUDs and lies. For instance, IBM states that the Power6 CPU has 240 GB/sec bandwidth. That is a lie because IBM has added all bandwidth in the chip: L1 cache, L2 cache, etc. If there is a bottleneck on 5GB/sec then 5GB/sec it is. You can not add bandwidth. That is dumb, or ignorant or deliberate lies. I suspect the latter.

    For instance, IBM claims that a Mainframe consolidates 1500 of x86 servers, That is a lie. Because IBM assumes the x86 servers idles, and the Mainframe is under 100% load. And also, it is well known that 1 Mainframe MIPS == 4 MHz x86. Thus, Mainframe CPUs sucks and can never match a strong Nehalem.

    For instance, IBM claims that the Power6 core is faster than a Niagara core, thus the Power6 CPU is faster. That is clearly wrong. The core might be faster, but if you want to claim that the Power6 CPU is faster, then you have to compare CPU vs CPU. And in many benches, 1.4GHz Niagara is several times faster than a 4.7GHz Power6.

    3. I always support the best tech. I am a geek and love high tech. The company with the best innovative tech gets my support. I mean it. It is that simple. If IBM had best tech, I would support IBM. I promise. Despite me not liking IBM FUDing and lying, I would support IBM. But I would strongly object each time IBM FUDs.

    Now, which company has best tech and is most innovative? Which company can match Niagara? ZFS? Dtrace? Java? This techniques just crush and kills and stomps everything else. Nothing can compete with ZFS. Nor DTrace, Nor Niagara. etc etc etc. If IBM gets better tech, then I switch to become an IBM supporter. I promise. 10 years ago I supported Microsoft and WinNT. I thought WinNT was the shit. Later I discovered Unix and immediately switched to Unix supporter. I will switch again if I find a better tech company. I have no problem with that. I mean it. I follow the best tech.

    And lastly, Oracle will reprice their database so to not punish SPARC. IBMs slow Power CPUs will be punished instead. :o)

  13. morklebork
    Flame

    The rumours of SPARC's death are greatly exaggerated... (or, more politely, suck on this, dummies!)

    The RF chips were never listed as 16 core * 16 thread in any official slide deck shown to me. Way back when you lepers at El Reg reported it, I already knew it was wrong.

    Nice to see you are humble in accepting your mistake.

    Further to this - to the clowns saying the roadmap is not a roadmap at all with the rate of process shrink. Remember - Sun has been using others to do the semiconductor fabbing for ages. The T-series stuff is done by someone different to the M-series - and that special someone is in bed with IBM already making the next few processes smaller... And Sun will be able to leverage that.

    It may be that Fujitsu will take a little longer to get the M-series chips to the smaller process node, but remember - with M-series, it's as much about scaling and correctness as it is about straight-line speed. What happens when an Intel CPU takes a parity error on a register? Pretty sure it won't even notice... I'll tell you what the M-series stuff will do. It'll see the parity error, it'll roll back the current instruction stream, reload and replay it...

    And - saying there is no innovation - You are forgetting; There was tons of innovation in the Rock platform that's still available to them. And - they have learned the lessons in that they have actually built the systems. They just chose not to go full scale production.

    If you are one of the many that look only at CPU straight line speed, you are overlooking all of the other features lacking in commodity platforms today...

    Like:

    - being able to scale from a single CPU to 64 sockets, with no recompile (or even a reboot...)

    - being able to add and remove entire system boards from a running OS

    - being able to add and remove not just PCI cards, but entire IO assemblies

    - being able to have a single OS instance with a 2 or even 4 TB physical address space

    - being able to physically hardware partition your large box into multiple physical systems *without* a hypervisor

    - having OBP (And if you have not used it, you won't get this one...

    - being able to dump the box externally if it stops (more than just an NMI and hoping...)

    - being able to diagnose, the first time, if something fails

    - having an extensive FMA database for that *specific* platform

    - being able to take components offline whilst the system is up

    - being able to configure the system with *mirrored* DIMMS, so you could potentially ride through the complete failure of a whole stick (ok - I'll grant some other platforms can do that...:)

    - having on-ASIC history registers that allow the inspection of bus traffic to understand what's happening

    - having all sorts of telemetry available for inspection (say, through busstat)

    - having 256 and larger sized memory pages available to the system

    - having 16 I/O assemblies, each with 8 * pci-E x8 slots, all available to the one OS image for Mega DB or other large workloads where you need to get bulk amounts of data into and out of the one system

    I could go on and on - but I think you get the point.

    Oh - And don't forget. Larry loves the idea of an appliance... They will own all of the Rock IP, and the FISHWorks group at the end of the acquisition... One of the (sadly never released) 8 socket Rock Boxes would make an ideal system for an appliance... Oracle could sell you one, and use the magic screw-driver to turn up and down the capability of the box based on the size of your requirement / checkbook. :)

    Way more differentiated than just another x86 RAC environment, and you don't need to cross your fingers (and everything else) that people don't cobble queries up that cream the interconnect between nodes...

    And - Remember that it's Linux, not Windows that's been eating away at SPARC marketshare, and a large part of that is due to Oracle supporting more on Linux recently than Solaris SPARC (Validating the Linux platform...). Do you think they will continue to do that when they have their *own* operating system, Solaris, which is more scalable, stable, diagnosable, etc than linux?

    Let's not even *start* with the horrors presented to Oracle if they actually did kill off the SPARC platform. If someone wants a big box to run Big oracle, they only have a few choices these days... SPARC, POWER or ITANIUM.

    Considering how few people buy Itanium, how little software support there is it's really SPARC and Power (Yes - I know this is simplifying, but it's close enough to illustrate the point), where will those customers be driven if SPARC goes away??? To IBM on POWER. Oracle's sworn enemy in the database space... Even Sun haters would have to concede that it would be dumb for oracle to push their biggest customers in to the arms of IBM for hardware when everyone knows that it would also drag in IBM GS, and present the opportunity for IBM's DB2 and other applications to get an airing.

    In short - If you think SPARC can do *nothing* but die... you are wrong. Very wrong.

  14. Anonymous Coward
    Megaphone

    Come on you Mhz

    As soon as these damn chips get up to speed the better. We have an app that runs with 3k threads open and relies on turning round responses quickly. The x86 amd/intel chips did much better with this kind of load in the lower end of the performance range and the Sparcs only began to outpace the x86 chips as things got really busy (basically Sparcs lasted much longer before getting crippled). Our conclusion was x86 until things get busy, followed by a hot-swap to sparc ... except you cant do this :( Sooner the Sparcs are up to speed with the x86 market the better.

  15. Anonymous Coward
    Pirate

    Sun has cancelled Rock publicly (ish):

    From the 2009 10K filing:

    "We canceled development of our next generation microprocessor and recorded vendor claims and impairments of, intellectual property of approximately $30 million associated with the cancellation."

    Where's the Stephen T Colbert icon, for truthiness ?

    Pirates, as Oracle will take the Sun gold and make it shine

  16. Dave 13

    EU proctologists

    With the EU doing way, way more paranoid "due diligence" than any sane person would think normal, Oracle and Sun really *can't* collude on plans until the deal is finalized. Way to go IBM/HP..

    All Oracle can really do at this point is what they're doing now. Saying what they plan without any real ability to execute (yet.)

    I'd not bet too much against Larry and his boys, they have a habit of winning, ans Sun has been declared dead more times than I can remember.

  17. Anonymous Coward
    Go

    Oracle will be best match for IBM

    From the start the problem of SUN was marketing , there was no doubt about their technical superiority over IBM/HP . IBM/Cisco/Microsoft , by the mere size of the companies have so much money to throw in for marketing and bully other suppliers , which some of the Technically Superior product vendors & other sales people are unwilling to do . Recently i was suprised by IBM's move where i was consulting , IBM dropped the price to 4M from 10M (2*P595,2*570,22Blades and 40TB of storage, Tapes and support to go with loaded with everything ) . SUN stuck to their intial offer of 6M not willing to come down . SImilar cases with Cisco whenever they see Juniper on the case (A paranoid Cisco sales guy gave 80% discount for a Global bank when he heard the J word ).

    I beleive Oracle is best matched for IBM in terms of sales skills and a disciplined Sales organisation, which can play the customer around . And larry's vision may not be to sell hardware , he will change the way Computing is sold , So the Databases will be sold more like a appliance for the future (like cisco sells the routers ) no one will question what cpu , what cache and what manufacturing process used . You buy computing power per OLTP , tested and packaged , ready to plug in . You will buy desktop power by number of users. And they will team up with HP , who is has nothing to loose by reselling Oracle Appliances & integrating them.

    So, if i am a customer to buy some Big boxes i will wait and watch atleast till Mid October . Larry is no fool and integrating this two organisation is not going to be a big issues , as even before the buy they worked very close except a short period of Linux marriage .

  18. Lou 2
    Thumb Down

    Stay with cloudy writting ....

    And some non techie reporter said: "And instead of hitting something close to 2 GHz as it should be able to do as it shifts from a 65 nanometer to a 45 nanometer process in the middle of 2010, Sun is only telling customers that it can boost clock speeds to 1.67 GHz with Rainbow Falls.

    This, to use a technical term in the computer business, sucks."

    So improvements in GHz directly relates to improvements in CPU thoughput? Naha!! This is like improving the revs your car engine can do without improving the breathin ... naw, waste of typing - you won't get that either.

    So what you do is - ask Intel and IBM that if next time they "increase" the GHz of their CPU, if the speed of the application using this will improve at the same rate? Shall I type slower? Use capitals?

  19. mattaw

    Die shrinks are now very painful

    Seriously, 90nm was annoying to work with (the first technology where stuff didn't "just work"). 65nm is significantly unpleasant. 45nm and below are downright bizarre.

    This is where teams of backend P & R engineers have to push structural change on the frontend design team. They then have to re-engineer bits of the system to cope. Its not a few guys and some retiming issues anymore - you have to rebuild the blasted things.

    This is why most companies avoid sub 90nm processes like the plague. Too expensive.

    Note I am not talking about Intel, Nvidia, AMD, etc who can run their own foundries and have hundreds of engineers to get things working.

  20. DJUNIX

    Is it too late .. or just in time

    Not sure if it is too late or not, but I suspect most Sun customers are looking at alternative plans while waiting to see what happens. The rock chip might of made the difference at the high end and I'm not sold on the Niagara chip to take them to the promise land (its fine on low end and to some extent to the mid range but not design for the high end). Besides, I know its not who has the best product and the best technology, but rather who has the best marketing.

    So time will tell, but Sun does remind me a lot like DEC. Good products that people once bought but then faded into history. So Oracle can make Sun hardware better in time but I wonder if Larry is more interested in a quick buck or to take a chance and develop some good hardware!!

  21. amanfromMars 1 Silver badge

    Put your Money where your Mouth is ....Stand and Deliver :-)

    "In short - If you think SPARC can do *nothing* but die... you are wrong. Very wrong." ..... By morklebork Posted Sunday 13th September 2009 02:01 GMT

    Talking the talk, morklebork, is not walking the walk, and SPARC has failed miserably in, and been failed by its Executive Administration .... for they are still Playing in the Field rather than Leading IT in a NeuReal Virtual Direction and onto to Higher Planes of Abstraction/Extrapolation/Confusion/Progress. And that is Indicative of a Mentally Blocked Leadership and/or a Sinking Ship with a Crew of Bluffers and Fluffers?

  22. Aaron 10
    Grenade

    @Lou 2

    Who was the tech-ignorant writer that posted that blather about clockspeed == performance? I want to write a note to that bloke.

  23. Jesper Frimann

    Re:Lou 2

    Well if comparing between different types of the same processor, for example from a 3.5 GHz POWER6 to a 5 GHz POWER6 processor, for example on a power 550, then it's actually just as fast or rather a bit faster than the increase in GHz suggest.

    Why ? Cause the interconnects and the busses scale with the CPU frequency.

    // Jesper

  24. Matt Bryant Silver badge
    FAIL

    RE: The rumours of SPARC's death are greatly exaggerated...

    Ah, a few more Sunshiner myths wrapped up in the usual feature sell.

    Exactly what "innovations" were in the failed Rock platform that actually have (a) been proven, and (b) would benefit Niagara? The problem for you is that Sun are keeping very schtum about exactly what wasn't working and what was on Rock, but the industry analysts are pointing fingers at all those "innovations" like Sun's implementations of transactional memory and scout threads. If these couldn't work on a "large core" SPARC design, why on Earth do you think they'll work on the smaller, simplified-SPARC cores of Naigara, where there is even less die space to add in more transistors? That's even if Sun/Oracle could get those "innovations" working? And if they did, would they add anything to Niagara that the competitiors haven't already surpassed? One of the key problems with Rock was insiders were hinting it just couldn't match the current generation of Itanium and Power, let alone the ones it would have to compete with upon its delayed release, so it doesn't sound like there would be much point in attempting to carry anything over to Niagara. That sounds more like a plea from the Rock team to keep their jobs.

    Then I really enjoyed your list of features you say are lacking in commodity platforms, without mentioning there also lots of the points stated missing from Niagara. For example:

    ".....- being able to scale from a single CPU to 64 sockets, with no recompile (or even a reboot...)...." Niagara only scales to four scokets, period. Even Xeon has been able to scale higher than that for years in generic Windows servers, let alone more specialised designs like those from Unisys.

    ".....- being able to add and remove entire system boards from a running OS....." Seeing as all Niagara systems have one motherboard with the sockets directly on it, there is no way you can remove the system board as without it there is nothing to run the OS!

    "....being able to configure the system with *mirrored* DIMMS, so you could potentially ride through the complete failure of a whole stick (ok - I'll grant some other platforms can do that...:)...." Good thing you did becasue even the old IBM xSeries have been able to do that for years and I can just imagine the IBMers have a field day if you hadn't put in your little retraction. If you knew it was a bogus claim, why put it in the list?

    After more and more feature sell (what, no-one else can monitor their IO or remove IO chassis? - yeah, right!), you say you could go on and on? Please don't, because if I laugh any harder I'll fall off my chair. At least you didn't sprout the normal male bovine manure clowns like Kebabbert come out with, such as ZFS's supposed superiority (hey, remember BTRFS, Oracle's own technology, properly open-sourced and not stolen from NetApp, and already doing more than ZFS - which do you think is the less painful route for Oracle?). Maybe you Sunshiners are learning. Well, except Kebabbert, he's still running to catch up with the rest of the herd.

    The reality is that Sun has just gone through a storm of bad publicity during the last year, where their continual denials of any issues or of a sale when they were secretly hawking themselves around, publicly bleeding red ink and trying to hide that they were dying in the development labs, all made them look the more untrustworthy. In the years to come, customers will be very sceptical about taking on anything associated with the old Sun.

    /SP&L

  25. John Chadwick

    Here's something to make you chuckle...

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

    Well if that's true why are so many of the existing specialists voting with their feet, if they haven't already been made redundant.

    And why can Rock?

    PS - f this doesn't make sense then they changed the splash page.

  26. Anonymous Coward
    Anonymous Coward

    Re:Lou

    "...then it's actually just as fast or rather a bit faster than the increase in GHz suggest.

    Why ? Cause the interconnects and the busses scale with the CPU frequency."

    Wow! Who's been drinking the IBM Koolaid? That's the most ignorant piece of drivel I've ever heard - OK, that's exaggerating, but it's close! IBM went from Power6 4.2GHz to Power6 5GHz (20% increase in GHz), but when looking at IBM provided rPerf numbers this increase only yields 11-15% increase in Performance. The numbers start looking even worse when you compare the doubling of GHz between Power5+ and Power6 where IBM could only get a 35% performance increase...

    Don't believe me? Check out IBM's performance numbers:

    ftp://ftp.software.ibm.com/common/ssi/pm/rg/n/poo03017usen/POO03017USEN.PDF

    This isn't an indictment of IBM, it is just a matter of physics. Increasing GHz only gets you so far. That's why everyone except Itanium are going heavily multicore.

    ftp://ftp.software.ibm.com/common/ssi/pm/rg/n/poo03017usen/POO03017USEN.PDF

  27. Anonymous Coward
    FAIL

    Re: Here's something to make you chuckle...

    It still doesn't make sense... and they haven't changed the splash page.

    Oracle hasn't actually bought Sun yet, so how could anyone be redundant

    yet?

  28. Jesper Frimann
    Happy

    RE AC on RE:LOU

    AC wrote:

    "Wow! Who's been drinking the IBM Koolaid? That's the most ignorant piece of drivel I've ever heard - OK, that's exaggerating, but it's close! IBM went from Power6 4.2GHz to Power6 5GHz (20% increase in GHz), but when looking at IBM provided rPerf numbers this increase only yields 11-15% increase in Performance."

    Well let me guess, math isn't your strong side. Or at least you don't really bother checking out the facts before you make an unfounded statement.

    rPerf: for the Power 550

    2x4.2Ghz does 18,36 rPerf equals 2,19 rPerf/Ghz/Core

    2x5.0Gh does 21.18 rPerf equals 2,12 rPerf/Ghz/Core

    That is 97% scaling.

    4x4.2Ghz does 36,28 rPerf equals 2,16 rPerf/Ghz/Core

    4x5.0Gh does 41,81 rPerf equals 2,09 rPerf/Ghz/Core

    That is 97% scaling.

    6x4.2Ghz does 52,24 rPerf equals 2,07 rPerf/Ghz/Core

    6x5.0Gh does 60,20 rPerf equals 2,01 rPerf/Ghz/Core

    That is 97% scaling.

    8x4.2Ghz does 68,2 rPerf equals 2,03 rPerf/Ghz/Core

    8x5.0Gh does 78,6 rPerf equals 1,97 rPerf/Ghz/Core

    That is 97% scaling.

    That means that the scaling in performance is 97% of the scaling in frequency. That's better than putting more cores on the same die, in scalability.

    Now if you look at specint_rate numbers:

    16 way 4.2GHz power 550 does 212 in specintrate which means 3,15 specintrate/Ghz/Core

    16 way 5.0 GHz power 550 does 263 in specintrate which means 3,29 specintrate/Ghz/Core.

    Again my statement stands.

    AC wrote:

    "The numbers start looking even worse when you compare the doubling of GHz between Power5+ and Power6 where IBM could only get a 35% performance increase..."

    And I also specific wrote that you could not use my statement to compare between different versions of the Processor.

    AC wrote:

    "Don't believe me? Check out IBM's performance numbers:"

    I did and you were wrong.

    // Jesper

    Big brother cause something is wrong in the state of Denmark.

  29. morklebork
    FAIL

    I'm glad you enjoyed the list...

    Matt Bryant said - Then I really enjoyed your list of features you say are lacking in commodity platforms, without mentioning there also lots of the points stated missing from Niagara. For example:

    ".....- being able to scale from a single CPU to 64 sockets, with no recompile (or even a reboot...)...." Niagara only scales to four scokets, period. Even Xeon has been able to scale higher than that for years in generic Windows servers, let alone more specialised designs like those from Unisys.

    Clown. Did you forget that you can move a binary from a T-series to an M-series box and it'll just work? Ooh.

    ".....- being able to add and remove entire system boards from a running OS....." Seeing as all Niagara systems have one motherboard with the sockets directly on it, there is no way you can remove the system board as without it there is nothing to run the OS!

    Did you forget about M-series? Ooh. And the E/F-series the preceded it?

    Seems to me that overall, you don't know anything much about the M-series or are trying deliberately to ignore it... Either way, I think it's clear that in any case, you are just looking to poo poo the Sun kit no matter what. Nice work.

  30. Anonymous Coward
    Stop

    You should be ashamed....

    ... so should the individual that gave you the slide. This *IS* company confidential information and as such is protected by law. I wish we would sue you for using it, and the individual that passed over the slide.

    You have no right to publish.

  31. Matt Bryant Silver badge
    FAIL

    RE: morklebork

    "....Clown. Did you forget that you can move a binary from a T-series to an M-series box and it'll just work? Ooh....." True, but why would I move a binary form a T-series to an M-series, could it be because T-series doesn't scale? And if the app has been tuned for T-series (in other words, tiny little multi-threads), then it is going to perform like a dud on M-series unless I pay massively over the oddds to get the same number of cores in the M-series server. I'd be better off migrating to RHEL on x64, especially with the new Nehalems. And going the other way - M-series to T-series - is a nightmare with the heavy-thread apps typical to traditional datacenter SPARC, which was the fact that forced Sun and Fujitsu to bring out the M3000. Was that "ooh" at the end the reality of the situation striking home?

    "....Did you forget about M-series?...." Did I mention M-series? Please show me the M-series server that has Niagara CPUs, which is what I was discussing. Oh, you can't. Try reading before posting in future. In fact, just try not posting unless it's under a humour tag.

    "....Seems to me that overall, you don't know anything much about the M-series or are trying deliberately to ignore it...." Seems to me that, overall, you just can't read or comprehend; I expressly stated my post was a critique of the limitations of the Niagara-based servers. Of course, the M-series are quite easy to ignore seeing as how my experience is that IBM's and hp's offerings trounce them in use. And you really don't want to bring up the E- and F-series, especially not when it comes to performance or reliability! I spent three years replacing our E10Ks and E25Ks, I know exactly what their limitations were.

    In summary - go back and try again, clown.

    /SP&L

  32. David Halko
    Thumb Up

    @WinHatter: wait for it #

    WinHatter Posted on Saturday 12th September 2009 15:57 GMT, "65nm in 2011 means they are not investing in manufacturing processes"

    The SPARC road map posted by The Register shows 40nm in 2010!

    http://regmedia.co.uk/2009/09/11/sun_sparc_roadmap.jpg

    Does that mean you believe Sun is planning on investing in the manufacturing processes?

  33. Jesper Frimann
    Big Brother

    RE:@WinHatter: wait for it

    "Does that mean you believe Sun is planning on investing in the manufacturing processes?"

    SUN isn't actually investing in the manufacturing processes, they don't manufacture processors.

    T2 line is manufactured by TI and the SPARC64 is made by Fujitsu and Manufactured by TSMC.

    Furthermore Rainbow falls and onwards is also going to be manufactured by TSMC, as TI is exiting that marked.

    So basically Snoracle is relying on TSMC, who are going to make their first server processors, with these new SUN and Fujitsu deals.

    So I wouldn't be surprised if some delays crept in, with SUN going from their long time partner to a new one.

    // Jesper

This topic is closed for new posts.

Other stories you might like