Follow Slashdot stories on Twitter

 


Forgot your password?
Close
typodupeerror
×
IBM

IBM Launches p690 231

Posted by michael from the things-to-blow-the-budget-on dept.
edyavno writes: "IBM just announced the launch of their new high-end Unix server p690. It's based on its new Power 4 chip, and is in the same category as just announced Sun's SunFire 15K. It also includes some mainframe level features and can be used either as a single large server or divided into up to 16 "virtual" servers, running any combination of AIX 5L and Linux. Here's yahoo article, and here it is from IBM itself."
This discussion has been archived. No new comments can be posted.

IBM Launches p690 More

IBM Launches p690

Comments Filter:

  • The Marketing Buzz... (Score:1, Interesting)

    by FortKnox ( 169099 )
    "Buy the new p690, with Fast Pipes [slashdot.org]!!"

    Wonder why that pipe test was just released..... hmmmm.....
  • It's nice to see someone giving Sun a little competition in the "very high end" of the Unix spectrum.
    • HP's Superdome actually compares quite well in the "high-end Unix" realm. Dig it: Superdome [hp.com].
      • Very true, HP is another player in the high end Unix market, HP just gets overlooked by a lot of people, since Sun is #1 and IBM is #2 (in the Unix market).

        Additionally, it's been my experience that HP is usually a step or two behind the "best available", although I admit I've not investigated their offerings as closely as those of the other two.
  • Can it play chess?

  • Sure, it LOOKS powerful... (Score:5, Funny)

    by UserChrisCanter4 ( 464072 ) on Thursday October 04, 2001 @01:01PM (#2388501)
    [sarcasm]But if you'll take a look, neither Photoshop Filter or Final Cut Pro benchmarks are even obtainable... obviously it pales in comparison to a Dual G4-800![/sarcasm]
    • thank you!

      finally, someone else that chuckles at the apple ads.

      and for the record, I hate the "feature" of slashdot that won't let you post after X mins when you just posted, and then it waits N seconds after you select reply before it lets you go.

      I'm aware that it does this to prevent spamming and bots being set up to post, but it is too sensitive now and the worst part is that the textbox gets cleared, so you lose what you wrote. which in my case is *always* brilliant.
    • Yeah, but what we really want to know is how fast it can compile Cobol and whether it includes a card reader, 9 track tape and interfaces to a 1403 printer... we gotta upgrade soon.
    • Lets work him the real way.

      1 / Send one to Anand Tech
      2 / take a picture while Anand lose consciousness seeing the beast in his yard (Tom Hardware compatible 8)
      3 / wait 5 days

      4 / Learn at least that it IS possible to play Quake @ 1600*1200*32bits @ 1Gfps in software mode. 8)

      Gosh, someone gives me a joystick 8)

    • I know that's a joke, but Mac apps would probably run pretty fast under MacOnLinux on this machine.
      • Can you run MOL without a proper Mac ROM, though?

        Also, how close is POWER4 to PowerPC, anyway? I know the PPC 601 will run POWER code just fine, but how divergent is the G4 series, for example?

        /Brian
        • Can you run MOL without a proper Mac ROM, though?

          Yes.

          Also, how close is POWER4 to PowerPC, anyway?

          Power4 uses the 64-bit PowerPC instruction set, which is a superset of the 32-bit PowerPC instruction set used in G3s, G4s, etc.

  • *Warning* Rumor... (Score:4, Interesting)

    by FatRatBastard ( 7583 ) on Thursday October 04, 2001 @01:03PM (#2388511) Homepage
    I know the new IBM kit is supposed to go head to head against the new Sun kit, but if The Inquirer [theinquirer.net] is correct [theinquirer.net] this may be the last salvo in the Sun vs. IBM unix war. Of course, I take this with a HUGE grain of salt, but stranger things have happened (*cough* *cough* HP/Compaq).
    • The article makes some rather misleading comments, such as the stock price. . .

      You can't directly compare stock prices without taking into account things like the number of shares outstanding.

      For example, Sun has almost twice as many shares of stock out there, so even though it's stock price is lower, it's not quite as far off as it appears. (Market capitalization for IBM is $168B, and for Sun is $29B).

      As things stand right now, I'd be very surprised if IBM made a bit for Sun, (although, as you say, stranger things have happened). IBM's been gaining in the Unix market for the past year or so, I think they'd be better off to wait a good bit longer before doing anything so drastic.
      • You might also want to look at P/E ratios. IBM is 20.70 while Sun is still up at 31.03. Also while you are right that you can't directly compare stock prices, you might want to take a look at the charts for the last year for each company. Sun has been dropping, shedding over 80% of its value. IBM has remained relatively steady, bouncing around between $80-120.
    • Well, I don't really believe it. There's another company that would probably be biding against IBM if they tried to buy Sun, and they've got significantly deeper coffers as well. (No, it's not MS.)

      Besides, it just doesn't sound 'right.' Gut feeling is that it's just a rumour, no more.

      • Deeper than IBM or MS? In the IT industry? I don't see how that's possible, unless I'm missing somebody really obvious.

        • Hello in there! He's obviously talking about eMachines.
        • Just don't forget the old beast.
          Yes they are specialized. and quite good to see at work.
          They also were the firsts to come out with a WORKING 32 CPU. At the time where SMP for 2 proc in NT was considered as a Technical Mastercraft.
          (*Nix 8)
          And I think they have "some" money 8)
          (something like 9874632 * 10^36 what I'll ever earn 8|)
      • I am trying to figure out who you mean. GE? VA Linux? Apple? Nintendo? Ok, so GE is the only serious one.

        What I hear is that IBM would be happy to get control of Java. I can't think of another company that would have as much to gain from gaining control of Java as IBM. Except for MS and they would just kill it.

        Note: I happen to work for IBM. This post is pure speculation on my part and is not the opinion of IBM.

        • Except for MS and they would just kill it.

          Actually, if MS could gain full control over Java they wouldn't kill it. They'd just make some subtle changes to tie it directly to Windows and go to town.

          But what are the chances of Scott McN. selling to Bill B? About as good as me shitting a gold brick (even though I posted the rumor I still think the chances of Scott McN. selling to Lou G. are about as good as me shitting a bronze brick).
      • Well, I don't really believe it. There's another company that would probably be biding against IBM if they tried to buy Sun, and they've got significantly deeper coffers as well. (No, it's not MS.)

        I think somebody must be underestimating IBM's resources here, but the only other companies I could strain to think of that would buy Sun right now in terms of fit would be Fujitsu, Siemens, or GE. The first one has (I think) the best fit given its investment in SPARC, but I have no idea of whether they could or would want to buy the company. Siemens is (I think) the only big European contender to anything like this, while GE can, as always, do whatever the heck it likes.

    • Regardless of Share prices I don'think that Sun is any worse off than any of the other big tech companies right now. Which is to say they are all hurting. With this in mind I don't think any of the major tech companies want to try and spend that much money with the way the market currently is. It would require a serious amount of cash to purchase Sun and with most companies in the tech sector announcing profit warnings and layoffs I don't see that anybody is ready for a merger of this size. Finnaly having been using Suns for years this just does not "feel" like a very Sun thing todo. It's just not them.
      • With this in mind I don't think any of the major tech companies want to try and spend that much money with the way the market currently is.

        Except for HP, you mean?

      • If the past decade, and especially the past few years have thought us anything, it's that it is no longer about what feels right, if it ever was.
        And any company but especially a tech company couldn't care less about your feelings of loyalty and 'right', or those of their empolyees, if it impacts their bottom line.
        In the capitalist economy, the ever shrinking margins have long since squeezed conscience out of the corporate picture.
    • I did somewhere that there was some special rules about Sun stock that protected them from hostile takeover.

      I have no idea whether such a thing is possible, and even if it isn't, it probably would not prevent a volunteerly merger.

  • My question has been answered! (Score:5, Funny)

    by scott1853 ( 194884 ) on Thursday October 04, 2001 @01:05PM (#2388519)
    "achieves leadership in business, scientific and Java performance benchmarks"

    I had always been wondering what kind of system was needed to run Java apps at a decent speed.
  • The power4 is kicking everybody's ass on SPEC 2000. 783 SPECint2000 base and 1098 SPECfp2000 base. Check out the comparisons on Ace's Hardware [aceshardware.com].
    • SPECint2000 Base and SPECfp2000 Base are single-threaded applications and the results only tests one CPU. I'd like to see how well it scales in performance at 32 processors. That thing could be one very powerful Oracle database server or an application server.

      • Re:SPEC numbers (Score:5, Funny)

        by King Babar ( 19862 ) on Thursday October 04, 2001 @02:52PM (#2389042) Homepage
        SPECint2000 Base and SPECfp2000 Base are single-threaded applications and the results only tests one CPU. I'd like to see how well it scales in performance at 32 processors. That thing could be one very powerful Oracle database server or an application server.

        For what it's worth, the 32 processor 1.3GHz models claim an rPerf of 50.56, and all the number seem to scale about as you'd expect (e.g., not quite linear in # of processors).

        But, sicko that I am, where I think the Power4 might really get down and dirty is in supercomputing applications. None of this sharing, caring relational database stuff. No siree; I'm into a much more serious kind of scene. This freaky baby was just built for LINEAR ALGEBRA, friend, and I've heard that she can keep on doing it it all night long! We're talking a *serious* FLOPhouse here, folks. But you've got to talk dirty to it, something like Hey, take a look at this generalized eigenvalue problem--have you seen a longer or harder one than...

        Uh, excuse me; I guess I was losing my composure a bit back there. I've got both hands back on the keyboard now.

        But doesn't anybody else ever wonder why there aren't more LINPACK benchmarks posted on alt.sex.stories?

  • It's not affordable... Even on a software developer's salary, I can't buy one... bah!

    I guess I'll just have to wait six months before it comes down to the sub $3000 market...
  • Will it fit in my bedroom, and does it need special cooling?

  • Overview from IBM's website (Score:3, Informative)

    by isj ( 453011 ) on Thursday October 04, 2001 @01:08PM (#2388534) Homepage
    I grabbed this from IBM's website (http://www-1.ibm.com/servers/eserver/pseries/hard ware/datactr/p690.html):

    - Innovative, mainframe-inspired, datacenter-class UNIX
    server.
    - 8- to 32-way 64 bit SMP server utilizing the first ever
    POWER4 dual processor on a chip which uses IBM advanced
    silicon-on-insulator (SOI) copper technology.
    - Up to 256GB of memory, 160 PCI slots and over 4.6TB of
    internal storage.
    - Supports up to 16 logical partitions (LPAR), helping to
    consolidate workloads, reduce footprints and lower cost of
    ownership.
    - A dedicated Hardware Management Console that provides a
    graphical user interface for configuring and operating the
    system including a set of functions for managing LPAR
    configurations
    - State-of-the-art self-managing capabilities that improve
    reliability, availability and serviceability (RAS) and help lower
    costs.
    - Packaging in a new 24-inch rack with an integrated power
    subsystem which accommodates a pSeries 690 system and up
    to four I/O drawers.
    - AIX clustering and future to attach to SP systems.

    It looks very good. I just wonder what you would use 160 PCI slots for?
    • Imagine my IRC mates jaws drop when I do a cat /proc/cpuinfo running that!
      • -:- SignOff cdraus (Excess Flood)
    • 160 PCI slots... that would be great for having 100 joysticks to have a massive multiplayer flight-sim game on about a 500 inch monitor!
    • I can't speak for all Slashdot users, but I know for certain what *I'd* do with 160 pci ports.

      yeah, that's right, fill 'em with peanut butter.
    • It looks very good. I just wonder what you would use 160 PCI slots for?

      It is very good - and that list doesn't even include the ability to run Linux in an LPAR, which is also possible.

      Don't think of it as 160 slots - think of it as 10 slots per LPAR. Suddenly it looks a lot less.

      Zack

  • 16 virtual servers? (Score:3, Funny)

    by sporty ( 27564 ) on Thursday October 04, 2001 @01:10PM (#2388544) Homepage
    I must admit this is kinda cool.. but can you run say, 16 copies of NT and have them all blue screen at once?(/joke)


    Just curious...

  • These are both Motorola chips based on the same Power PC architecture, right? Will someone more informed than I am explain the differences between them, and explain why IBM is using Power 4's instead of G4's?
    • There are two variants on the CPU architecture, which was co-developed by IBM and Motorola.

      The PowerPC (with recent releases being known as G3/G4) variant is a 32 bit CPU, primarily found in Macintosh computers.

      The Power4 (and Power3 before it) is a 64 bit CPU (such as an Alpha or UltraSPARC), intended for use in high end workstations/servers, and found in IBM's RS/6000 line, AS/400 line, etc.

      So, they both stem from the same root, but they reflect different intended purposes.
    • These are both Motorola chips based on the same Power PC architecture, right? Will someone more informed than I am explain the differences between them, and explain why IBM is using Power 4's instead of G4's?

      The POWER4 is an IBM chip, for one, and a different chip architecture as well (thought IBM's line is an offshoot of the IBM/Motorola chip allicance). The POWER4 is an industrial-strength chip designed for high-test number crunching and processing power--IOW, it's designed with machines like the p690 in mind. And since IBM makes it, it's a no-brainer.

    • Overview of the Power4 [ibm.com] processor:

      Power4 is the processor that will be used in the next-generation RS/6000 and AS/400 systems (IBM eServer i-series and p-series). It is a high-performance VLSI chip that includes two 64-bit PowerPC microprocessors, connected at high bandwidth to an on-chip memory subsystem consisting of a shared L2-cache memory plus the directory and interface for a large off-chip L3, and with high-speed busses and I/O to enable efficient 8-way systems to be built on a single 4-chip module. The microprocessors will operate at > 1 GHz clock frequency and have processor-L2 cache bandwidths of 100 GB/s. The Power4 chip is divided into 12 units, some of which are being designed by multi-site teams. The Research team focuses on all aspects of VLSI design as well as design tools and methodologies. For the Instruction Fetch and L2 Cache Control Units, the circuit and physical design of the logic circuits (about 2M transistors for each unit) are done in Yorktown, the array designs in Poughkeepsie, and the logic and verification in Austin. Performance exceeding 1GHz is achieved at acceptable power levels using mostly static, custom-designed CMOS circuits for the dataflow. Synthesized logic, implemented using circuit books from a standard cell library, is used for most control circuits. The circuits are designed to be fabricated in IBM's 0.18 CMOS 8S2 Silicon-on-Insulator technology with 7 levels of copper wiring.

    • Yes, they are both PPC chips but the POWER4 is made by IBM. The PPC consortium was founded in the early 90s by Apple, IBM and Motorola. The main difference between them is that the POWER4 [ibm.com] is a server chip while the G4 [motorola.com] is a consumer chip. Because of this, cost, performance etc are VERY different. Basically the POWER4 is much better than the G4 but it costs a hell of a lot more.
    • The PowerPC series of processors is a descendent of the POWER series. A big distinguishing charateristic of the PowerPC vs. POWER series is that the PowerPC is always a single core on chip implementation. However the POWER series, esp. the POWER4, can be/is a multi core implementation on a single chip. Which makes the POWER series better but more expensive, thus great for servers. For more information I would check out Motorola & IBM's site. Here is a piece of info from IBM on the history of the PowerPC project.

      http://www.rs6000.ibm.com/resource/technology/ppc_ arch.html [ibm.com]

    • Re:Differences between PPC G4 and Power 4 ? (Score:1, Insightful)

      by Anonymous Coward
      Actually do you have any idea why the apple kids just dont get together and build a box w/ macos10.1 or whatever the flavor of the day is on it? Sure it would cost 25,000 but it make a great advertising campaign.. yes no matter how you cut it macintosh 50 times faster than intel/windows, or some such (I am making this number up). Incidentally back the intel chips were at ppro 200 macintosh briefly allowed clones. Well we snagged a slew of the compaq clones and installed AIX on it.. the xl* compilers all port code fine (pwr2/ppc etc options) and the generic compile was *binary compatible* across architectures.. that is fucking cool, you could take SP2 code and run it on a mac :). In any case the practical upshot was since ppc was a subset (at least back then) of the pwr* chips, it would be relatively easy to do a port to the pwr4. A few high end graphics people will buy it since they need their tools, and *much more importantly* it really would make for some kick ass advertising. IE the cost of that port
      would easily pay for itself.. -avi

  • Heeeeeyyyy....a new IBM machine.... (Score:3, Interesting)

    by mystery_bowler ( 472698 ) on Thursday October 04, 2001 @01:11PM (#2388546) Homepage
    First off...from IBM's site:

    Self-healing architecture -- Built with technology from IBM's Project eLiza initiative, the p690 is the industry's only UNIX server that offers multiple layers of self-healing technologies that allow the server to continue operating, even through major failures and system errors.

    Gaaah! I blew a hole through p690! But wait...it's...healing itself! :)

    On a more serious note...which marketing direction is IBM taking on these things? I'm sure they're trying to sell at least some of their existing customers on these and keeping those customers on whatever OS they were running (AIX most likely). But for the new customers, are they pushing AIX harder than Linux? Are they actually pitching any Linux conversions to their existing AIX customers?

    • AIX and Linux (Score:3, Insightful)

      by Coz ( 178857 )
      The AIX 5L release is "Linux-capable" - it should be able to run most "vanilla" Linux apps with recompilation.

      That said, IBM has been pushing several of their AIX selling points into Linux, like their Journalling File System logical volume manager. Their system management tools are pretty good (no SMIT cracks, please) and they have good network management tools. I got out of the crystal ball business a while ago, but I imagine IBM would like to be spending their money "productizing" Linux on their platforms rather than supporting their own OS.

      I'd say they're trying to take a piece of Sun's pie, and maybe try to keep some folks from moving to Win2K. Looks like a good price/performance system if you need that much to start with.
    • That mutant healing factor should come standard on all server hardware, I think.

    • I don't see them pushing Linux on this machine. At the moment Linux on IBM is cool for mainframes (consolidate a zillion lightly loaded boxes onto one huge one) and for little rackmount babies.

      This baby screams "massive database server". If you partition it you might break it into "devel", "test", "production", but not into a lot of tiny partitions.

      I do think they'll go whole hog into pushing Linux on all their platforms eventually, but we still need a few more scalability improvements before you'll want to put Linux on a 24 processer SMP machine, plus the Linux LVM and journaling filesystems need to mature a little.

    • Self-healing architecture

      i thought that the integration of Borg technology wasn't supposed to happen until the 24th century......
  • I can recall a local ISP (pre internet) that was basically a home business at that point, and had an insane number of modems in the basement. Had to keep the windows open in the winter just for proper coooling, and had no real heating bill. It was all the heat from the computers that kept the house warm.

    I think it was Channel One in boston, back when it was a BBS.

    I can see this with some of these babies

  • I realize that the power 4 is not exactly the same as the PPC that one might find in the Apple G4 series, but I'm curious about something that this touches on: if IBM has customers that might choose to run a linux virtual server on one of these beasts, they must have a java runtime available. I was wondering if, perhaps, this runtime was also available for linux/ppc. The only one that I'm aware of is the blackdown port, which I find lacking. I've seen that IBM has their own JDK for linux/x86. Do they have one for PPC? If not, why not?

  • Wrong Comparision (Score:4, Informative)

    by Doctor_D ( 6980 ) on Thursday October 04, 2001 @01:28PM (#2388642) Homepage Journal
    IBM's comparing their new server against the wrong Sun server. Here's why:

    For unmatched UNIX system performance, the pSeries 690 can scale to a 32-way symmetric multiprocessor (SMP) helping to provide the scalability required to drive a UNIX datacenter.

    A Sun Fire 15k contains up to 106 processors (72 with max i/o), a Sun Enterprise 10k contains up to 64 processors, and a Sun Fire 6800 contains 24 processors. Honestly this IBM server should be compared with either the 10k or 6800. It just can't scale as high as either the 10k or the 15k.

    LPAR support for up to 16 UNIX or Linux partitions

    Humm, first generation unix partitioning from IBM, or 5th generation partitioning from Sun (with help from Cray early on). BTW, a 10k can be in 16 partitions. No it doesn't require a domain to contain 4 processors--that's the max. A single board domain can have 1 i/o card, 1 cpu and some memory--typically a gig. The 15k and 6800 are similar, although the cpu/memory cards are typically maxed. It is *very* rare to find a company who would buy these sorts of systems to not max them out.

    AIX 5L offers support for systems with up to 32 processors and 256 GB memory.

    Wow, Solaris scales to 106 procesors in a single domain, with at least 1/2TB of memory. Besides, I'd bet there are more apps for Solaris than AIX.

    *Note all of the quotes are from IBM's web page regarding the p690.

    • AIX 5L is designed to run Linux apps with a recompile, or AIX apps without. Find me a Linux box that I can stuff 32 dual-core processors into.

      Unless you're insane I don't believe you're going to tell me that there are more Solaris apps than Linux apps. AIX 5L runs the latter kind.

      Besides, IBM techies have usually struck me as better qualified than Sun guys, although both are leagues ahead of almost any other company's.
      • AIX 5L is designed to run Linux apps with a recompile, or AIX apps without. Find me a Linux box that I can stuff 32 dual-core processors into.

        Humm, recompile, eh? Yeah, if you're going to recompile, most things that I have run under Linux also work under Solaris...except for some things that expects certain things in the kernel, or so other Linux specific quirk. Yeah, I've compiled tons of GNU apps on a Solaris machine. ./confiugre;make;make install done.

        But if you don't want to recompile and are running on Solaris x86, then you can use lxrun [sun.com].

        Yeah, I'd venture to say there are more Solaris apps than Linux apps. How often do you see commerical support for Linux? Yes, it's increasing, and now when you talk with a commerical vendor, most have at least heard of Linux. Some even have software to sell you. If you're talking GNU or Open Source (i.e. gpl, lgpl, artistic, bsd licenses), then sure there's more...but ya know, typically those apps aren't Linux specific.

        This doesn't mean I dislike Linux, or am bashing it. I have used Linux for years, and knowing it got my foot in the door as a sys admin, and helped me figure out how to admin HP/UX and Solaris. Yes, Linux is far more useful today than it was a few years ago, supports more hardware, but internally it hasn't fully stepped into the big leagues...it's getting closer thanks to the larger memory support and journaling filesystems.

        So, unless I'm insane, could you please prove me wrong in this?
    • Last time I checked, the World's most powerful super computers were IBM POWER3 machines running AIX, one of which has 8192 processors. I believe they meant that on the p690 you can have up to 32 processors.

    • Re:Wrong Comparision (Score:3, Interesting)

      by heimdall ( 44846 )
      I think you're missing a few things. First of all, yes it has 32 CPUs... but each CPU has two cores on it, i.e... it's more comparable to a 64 CPU box. Secondly, the CPUs have considerably higher benchmarks than Suns USIIIi's. The 6800's can only support 4 domains, total, split between two hardware segments. This will handle up to 16 domains. We still need to compare it to an E10K or a F15K. While an E10k can handle a single CPU on a board, the F15K CPUs are soldered onto the CPU/Mem boards, and thus are only available in unit of 2 or 4. Also, as IBM invented partition (xx/360-390 and AS/400), I don't know that I'd compare this "first generation" partitioning with Sun's "fifth generation" partitioning. (And they didn't get help from Cray early on... Cray WROTE IT early on, sold it to SGI, who then sold the C6400 to Sun and was renamed the E10K.)

      I bet you're right... I bet there are more apps for Solaris, however when it comes to the apps that actually run on boxen this large, you're typically talking about an Oracle or DB/2 database, which are available for both platforms.

      Both the F15K and p690 appear to be fantastic boxes. Only time will tell how they fare against each other.

    • Re:Wrong Comparision (Score:2, Informative)

      by eclarkso ( 179502 )
      Humm, first generation unix partitioning from IBM, or 5th generation partitioning from Sun (with help from Cray early on). BTW, a 10k can be in 16 partitions. No it doesn't require a domain to contain 4 processors--that's the max. A single board domain can have 1 i/o card, 1 cpu and some memory--typically a gig. The 15k and 6800 are similar, although the cpu/memory cards are typically maxed. It is *very* rare to find a company who would buy these sorts of systems to not max them out.

      Without addressing the rest of your post:

      It has been well publicized that the p690 contains a number of features adapted from IBM's mainframe legacy. So "1st generation UNIX partitioning," while literally true, discounts 30 years of IBM big iron partitioning experience.

      Addressing the rest of your post:

      You are leaving out the IBM's performance claims that their processors vastly outperform Sun's. If nothing else, some benchmarks validate IBM's assertions.

    • You've fallen for the equivelent of the Mhz argument. "But this one goes at 1 Ghz, that's a lot faster than the PPC 500Mhz"


      This IBM machine has what they call, a "server on a chip." It's got memory, processing, and I/O on the same piece of silicon. That means it can process the information a lot faster than comparable sun processors. Check out this [msnbc.com] article comparing them for more information about why IBM sees this as a competitor to the Starcat, and why you just can't go by the number of processors.

    • The 15K has 106 CPUs, but they are the slowest CPUs on the market in SPEC benchmarks. The UltraSPARC III are slower than the 1300 MHz Power4, slower than the 833 MHz 21264B, slower than the 2000 MHz Pentium 4, slower than the 800 MHz Itanium in FP but faster in integer, and not any faster than the 1200 MHz Athlon MP. The only CPU that is slower is the 552 MHz PA-8600, and the UltraSPARC III barely beats it. The PA-8600 will be replaced imminently by the 750 MHz PA-8700, while the 833 MHz 21264B will soon be replaced by the astonishingly fast 1000 MHz 21264B.

      But wait! you say, SPEC numbers aren't everything. Yes the 15K has some seriously inter CPU bandwidth and big-time scalability. Problem for the 15K is that the Power4's inter-CPU bandwidth makes the 15K look like a beowulf cluster running over appletalk. The Power4 has a two cores sitting right next to each other on the same die and can/does have four or more of these double cores wired together in the same package with 128 MB L3 memory bank. The Power4's system bandwidth is 92 GB/s, or 38 times higer than the UltraSPARC III.

      It's time to euthanize the poor old UltraSPARC CPU line.

    • Humm, first generation unix partitioning from IBM, or 5th generation partitioning from Sun

      IBM has been doing LPARs and Virtual Machines a long, long time on mainframes. According to this [techtarget.com] article, IBM started studying logical partitioning in 1976. I think Sun qualifies as the newcomer if you are just talking about partitioning technology.

      See Linux on the IBM ESA/390 Mainframe Architecture [linas.org] for info on running Linux on S/390 LPARs

    • Re:Wrong Comparision (Score:5, Informative)

      by Tower ( 37395 ) on Thursday October 04, 2001 @02:41PM (#2388992)
      disclaimer: I'm an IBM employee (and have played with systems using the POWER4)

      First of all, let me mention that the RS/6k S80 (two releases ago - prior to the p680) outdid the 10k, at reduced cost, with the previous generation of procs. 16 processors outdid 64 in many, many tests (including ones with real-world data movement).

      As for partitioning... hmmm... let's think. IBM has been doing logical partitioning in AS/400 for a while, and on the S/390 (now the z-series) for quite some time... a few decades now. A lot of that experience went into this.

      Regarding the POWER4:
      Scalability: The eServer p690 is able to marshal up to 1,000 processors for high-performancesupercomputing duty, in applications such as Business Intelligence or seismic data interpretation. (think - the big supercomputers are right now POWER3, with several POWER4 systems in devel)

      Raw power:
      Our POWER4 processor can handle seventeen times more data than the UltraSparc III chip used in Sun's brand-new "Star Cat" top-of-the line F150000. Only 32 IBM processors outperform double the number of Ultra Sparcs - which draw much more power, create more heat and are less efficiently packaged than ours, which use modules developed for the eServer z900 mainframe. (The CPU numbers can be found in SPECmark, and from other benchmarks, including TPC-C, Javamarks, and some other fairly useless comparisons).

      • Re:Wrong Comparision (Score:3, Informative)

        by Tower ( 37395 )
        Just to back some more of that up:
        Power4 (1.3GHz):
        SpecInt2000: 783 - 808 (base - peak)
        Specfp 2000: 1098 - 1169

        UltraSPARC III (900MHz):
        SpecInt2000: 438 - 467
        Specfp 2000: 427 - 482

        That, and the clustering of the cores on the MCMs allowing for massive inter-processor bandwith when all bound together makes quite a powerful machine.
      • <disclaimer> I'm drunk.</disclaimer&gt

        IBM, Digital (sorry, Compaq) and HP can happily toast the pants off Sun in any benchmark there is.
        Total bragging rights go to the PowerPC4, the Alpha, the PA-RISC.

        So, why does Sun have so much market share?

        Two simple reasons. It's reliable, and it's cheap.

        OK, I'll justify that. When you buy a box from Sun, you give it a power cable, a network cable, you fire it up and it works. (OK, this magic is done by a sysadmin. But it's a commodity sysadmin. Because *everybody* knows how to make a Solaris box work) So, It's reliable.

        It's also cheap. Sun sold you a box. They didn't *give* you a box in return for unspecified (but expensive) "services". You're not tied in for years. Companies don't like being tied to multi-billion dollar bullies.
        Here's a test: Ask Sun to sell you an E450. Then, ask HP to sell you an L-series. Once you've bought it, I *bet* you get fewer solicitations from Sun for "consultancy" than you get from HP/IBM/...

        Sun build reliable, fast, affordable kit. Then, they support it. That's all they do,

        IBM/HP/Compaq/ make faster, more expensive, less affordable kit. Then they nail your management to the floor with incessant 'Services' crap.

        Before you all correct me, *yes* I do know about real computers. I work with the things. And I'd happily trade CPU performance for bus throughput any day.

        I'd love to see an IBM box running Solaris...

    • Re:Wrong Comparision (Score:2, Insightful)

      by Anonymous Coward
      If you want to talk scale let's talk SGI: 1028 processor system with SSI (Single System Image), and >700GBs (not Gbps mind you) interconnect speed. SGI really has their act together when it comes to scaling.

      Now onto IBM's POWER4--

      Since IBM bought Sequent they are in the process of migrating the NUMA technology into the pSeries servers. If you look a little closer at the specifications they mention that it will be possible to interconnect 16 p690s together. The overall question remains, will this be an instance of NUMA interconnect (like SGI et al) or will it use the SP Switch2. If this is NUMA and they can do SSI then they will be pretty close to SGI's sweet spot: 512 P SSI systems.

      Now onto a comparison of the UltraSPARC 3--

      The first problem is packaging size. Essentially Sun has a huge package for 1 USPARC III processor which is about the same size as 8 POWER4 processors packaged in an MCM. This means that for the same size your processor density is 8 fold greater. Additionally, in the labs at my work place we have a 280R with 2x USPARC IIIs in them, and the machine sucks at least 560 Watts of power--if you count the fact that it has 2x power supplies internally means that there is at least the potential to suck 1160 watts, ouch that really hurts in the Bay Area! Not only that what is Sun's story after USPARC III, and take into account the notion that it took them FOREVER to get the StarCat et al out the door.

      For a comparison--

      Now if you look at the numbers from the p680 (S85) it is obvious that it is possible to directly compare this machine with the E10k and perhaps the E15k (maybe not here though). IBM's notion is that even though you have a whole lot of things if you do not use them efficiently then why do you got'em? IBM was able to beat out the E10k on stadard performance benchmarks. Even if you take a more conservative stance and say that in reality you would see 50% of the performance of each of these machines, with almost 1/3 the number of processors the S85 (and even the S80) were able to hang with and beat the E10k. This is impressive, and states that Sun really does not have the engineering leadership they claim. In fact all Sun really has is mind share and market share. If they fail in maintaining market share, their number is up because they lose the mind share game after that!

      Feel free to send any flames/observations to: Michael.Hay@hds.com!

    • Re:Wrong Comparision (Score:4, Informative)

      by cartman ( 18204 ) on Thursday October 04, 2001 @03:02PM (#2389084)
      I'm not meaning to flame, but your analysis was so flawed that I'm amazed it got moderated up to level 4. IBM's hardware is vastly superior to anything Sun has to offer.

      First, even though Sun's E15k has 106 processors, only 72 of them are even directly connected to the memory fabric. The others are just PCI cards, with staggering latency, consequently they won't help transaction processing performance. The E15k actually has 72 usable processors; the others are there to impress people who measure system performance by "counting processors."

      Second, the UltraSparc III is a notoriously weak performer. It can't even execute instructions out of order!! It is quite likely that IBM's POWER4 would outperform it by more than a factor of 2.

      Third, although the p690 has only 32 processors, it has 64 cores. Each "processor" has two 4-way CPUs with a very low latency interconnect. The IBM product would be more accurately characterized as a 64-processor machine.

      Fourth, the products from other Unix vendors (hp, ibm) always vastly outperform Sun's product with dramatically fewer processors. HP's new 16 processor box gets almost the same tpc rating as Sun's 64 processor E10k. IBM's old p680 with 24 processors almost doubled the performance of the E10k w/ 64 processors. Sun's most recent comments of "we've decided not to use industry-standard benchmarks any more" is likely because they always lose badly.

      Sun's real benefit is in software (Solaris is way better than AIX and has more apps) and in the fact that they have one OS and one proc architecture (Solaris/Sparc) across their entire range of computers.

    • Comment removed based on user account deletion
  • I was one of 2 industrial designers involved on this project. We determined the outward appearance and graphics for the thing. It's real nice to see something you designed finally come on the market. Woo!

  • Oh yeah, these are great! (Score:4, Funny)

    by Sir_Real ( 179104 ) on Thursday October 04, 2001 @01:51PM (#2388747)
    I'm running one of these right now as my MASQ/NAT box... You should see how quickly fortune executes...

    ;-)
  • Well what's with all these P6 90mhz machines that IBM is repackaging??? Sounds like alot of baloney to me, unless I can build a Furbeowulf cluster [trygve.com] with them.

    I hate to say it... but imagine a.... nah... I don't wanna go there.

  • Eggplants! (Score:1, Troll)

    by Ace905 ( 163071 )
    I would like to er.... uh, Hmmm...

    Eggplants! [eggforge.net]
  • simply go to this page [ibm.com], select the type and number of each server that you want and proceed to check out.

    Remeber: Do not click twice!
  • Just for everyones information, this is something BIG inside here. They've been coming over the PA and telling us about it, how great it is, and everything. The stories on the internal news sites are flying, touting the new system (which btw is made right here in Poughkeepsie)

    They even celebrated by giving all the employees free cake today (from 11:30-1:00)

    mmmm. free cake.
  • shouldn't it be called the T-1000 instead of the p690?

    "Holy shit! It's healing itself!"

    "Have you seen this boy?"
  • So, what's the world's least powerfull Unix server (in recent history)? Does it run Linux too?

  • that the Alpha 21364 is not here to provide the Power4 with a worthy competitor.

    At least it sets a standard of performance so that Itania (cubic Zirconia?) cannot be simply passed off as "great" just as they are.

    Intel will have some real hard work to do to match the Power4, which not only has some good processing speed, but some BW to memory that should propel to the top of the heap in some benchmark categories.

Slashdot Top Deals

The Tao is like a glob pattern: used but never used up. It is like the extern void: filled with infinite possibilities.

Close