Slashdot Mirror
SGI to Scale Linux Across 1024 CPUs
im333mfg writes "ComputerWorld has an article up about an upcoming SGI Machine, being built for the National Center for Supercomputing Applications, "that will run a single Linux operating system image across 1,024 Intel Corp. Itanium 2 processors and 3TB of shared memory.""
Sweet, now we'll be able to run Doom3 at highest detail in *SOFTWARE*-rendering mode!
But does it run--crap. I mean what about a Beowulf--doh!
Damn you SGI!
Why not fork?
Yeah, but can it run Longhorn?
Intel's sales figures for Itanic^Hum CPUs more than doubled as a result.
--
"Open source is good." - Steve Jobs
"Open source is evil." - Microsoft
DOOMIII :-)
"If anything can go wrong, it will." - Murphy
It seems that if they pull this off one of the dtrongholds of solaris (namely massivly parralell computing) will have been conqurered by linux. I wonder how sun are feeling at the moment?
Hmm, quite possibly.
"Proudly Posting Without Reading The Article"
what happend to RISC...
Obviously this would be overkill for doom3(altho I'd still like to have it in my apartment as a space heater/server)! Ok, so it would be more than a space heater; I'd have to run my a/c 24/7/365.25, with all my windows open in the winter. But rendering would be sooooo sweet.
The link to the press release as of July 14.
CC.
TaijiQuan (Huang, 5 loosenings)
so does this mean KDE and Openoffice will finally run at decent speed?
No, you're going to need quantum computing for that.
...how easy it is to install printer and sound drivers?
Microsoft made a statement today reminding everyone that Windows Server 2003 can handle as many as 32 processors, at the same time even.
When shown the report about Linux running on 1024 processors, Gates purportedly responded, "32 processors ought to be enough for anybody."
Unknown host pong.
you mean Doom 3, right?
Join Team Mozilla #38050 Folding@home
yes, according to the project leader "on this supercomputer, OpenOffice will finally *run* at decent speed, but waiting for the JVM to start up will still be a bitch" As for KDE, he stated "we're still waiting for the qt toolkit to initialize, but we're confident we can be fully logged in before August"
can you imagine a beowulf cluster of these ?
I was going to ask why not just build a cluster instead. Then, the Slashdot herd mentality took hold, and this thought crossed my mind:
"My god, imagine a Beowulf cluster of those!"
This space intentionally left blank.
They said Itanium cluster, not VAX cluster!
Scientific computing means data crunching (floating point). Complex, powerful processors are needed. The "stupider, but more" tradeoff doesn't work anymore. Sun processors have fallen behind in this respect.
The Raven
RISC stands for "reduced instruction set computer". It made sense in the 1980's when the "CISC", complex instruction set computers, took tens or hundreds of clock cycles to execute some instructions. With RISC one had less instructions, but each instruction executed in less clock cycles, resulting in a faster computer. Today, CPU's with full-size instruction sets execute most of them as fast as a RISC CPU does, so there is no need to limit the instruction set anymore. Even such complex instructions as multyplying double-precision floating point numbers are executed in a single clock cycle in a Pentium 4.
Sun hardware has additional, wonderful resiliency features like - allowing cpu's to "fail-over" to other cpus in case of failure. The same holds true for memory, network interfaces, etc. Solaris is aware of these hardware features and can "map out" the bad memory and cpus on the fly (or allow swap-in replacements). The engineers can then replace the broken cpus/memory/interfaces WITHOUT BRINGING THE MACHINE DOWN. This lends itself to an environment than can enjoy nearly 100% uptime. Finally, since Sun has been doing the "lots of cpus" thing for many years, their process management and scalability tends to be much better.
I don't work for Sun, I'm just an SA that deals with both Solaris and Linux boxes. You don't pick sun for just "lots of cpus", you pick it for a very scalable OS and amazing hardware that allows for a very, very solid datacenter. If downtime costs a lot (ie. you lose a lot of money for being down), you should have Sun and/or IBM zseries hardware. Unfortunately those features cost a lot and most times you can use Linux clustering instead for a fraction of the cost and a high percentage of the availability.
With the exception of the NUMA stuff, is there software available to re-create this? I'm not even sure what to search for; would this still be considered a "cluster"?
.... to see just how far you can stretch a bit before the MP loses control....
They bought HP's overstock of them for pennies on the dollar.
But does it play Ogg Vorbis?
I wish I had that much disk space...
A small suggestion - get better hardware, and you won't need those features.
:-)
No charge
First Doom 3 now this... coincidence? I don't think so.
"Academicians are more likely to share each other's toothbrush than each other's nomenclature."
Cohen
won't that result in running both KDE and GNOME at the same time, but as soon as you look at it you see FVWM instead?
The Sun hardware is more difficult to deal with, since there isn't a virtual machine abstraction. You can't do everything below the OS. Still, Linux 2.6 has hot-plug CPU support that will do the job without help from a virtual machine. Hot-plug memory patches were posted a day or two ago. Again, this is NOT required for hot-plug on the zSeries. IBM whips Sun.
I'd trust the zSeries hardware far more than Sun's junk. A zSeries CPU has two pipelines running the exact same operations. Results get compared at the end, before committing them to memory. If the results differ, the CPU is taken down without corrupting memory as it dies. This lets the OS continue that app on another CPU without having the app crash.
Worst.. FP.. ever!
"Hey! Unless this is a nude love-in, get the hell off my property!!"
"will run a single Linux operating system image across 1,024 Intel Corp. Itanium 2 processors..."
"The National Center for Supercomputing Applications will use it for research"
1. Make a system that generates more heat than a supernova.
2.Research a solution to global warming.
3. Profit!
SCO gained $715,776
I have replaced Sun Hardware/Software combo's in the core datacenter for many of our customers, and I can tell you that yes - Sun brings some amazing features to the table - most of which are there to serve old technology. Linux on simple CPU's delivers such an amazing price performance (depending on the job, we see an average of 3x to 4x performance increase for 25% of the cost. That means that if I were to spend the same, lifecycle-wise, on a Linux cluster as I would on a big Sun box like the 10k or 15k, I'd end up with 12x to 16x the performance of the Sun solution.
The same functionality in terms of cpu and ram (and other hardware) failure is available on the Linux cluster, albeit in less graceful form - the magic spell to invoke goes like this: if I have 300 machines crunching my data, I can afford to lose a couple, and can afford to have a few hot-standby's.
Of course, the massively parrallel architecture does not work for all applications, and in those cases you would look to use either OpenMOSIX or of course the (relatively expensive) SGI box mentioned in this article.
People who think they know everything are a great annoyance to those of us who do.
i know this is a joke but i don't get it - you guys must have slow pcs is all i can say. kde ain't slow for me..
Will the person who moded the parent "off topic" please put down the crack pipe and post a reply explaining why.
--= Isn't it surprising how badly I spell ?
Hot damn, this is one server that could survive a slashdotting.
As everybody that has read the IBM redbooks about mainframe linux knows, Sendmail is the service of choice! Of course, you could run Postfix on a decrepid old pentium-1 and get the same level of perfomance, but that won't help IBM with their Mainframe income, will it?
People who think they know everything are a great annoyance to those of us who do.
Really, I know this is off topic, but what are the advantages of having more CPUs on a computer system? I have been approached by various vendors explaining to me why a system with dual CPUs would be better. Wouldn't I be able to harness the same performance by using more and more RAM or just a more powerful CPU?
According to top 500 supercomputers 35 TFLOPS would put it ino second place after Earth simulation center - 35-40 TFLOPS
There goes SGI again, violating SCO's copyright! I'm gonna tell Darl McBride on them!
This sig no verb.
Why is NCSA buying big iron like this when the National Science Foundation has told them to start steering away buying machines like this?
And from SGI for crying out loud! What are these people thinking?
I don't think many people realize how _HUGE_ is a 32 processor box (myself at least I can't even imagine it). As someone already told you, win 2k3 supports 64 CPUs not 32. But I bet they could make it run in ej: 128 cpus with no much effort (if any).
Why they don't to it? Plain simple: money. That SGI box is a special case because some "national" thing. Enterprises don't use 1024, or even 128 or 64 cpus in most of the cases: The run clusters of small machines, that's what it sells today and that's where Microsoft is putting their eyes.
...Right on the heels of this too.
It's ok for embedded and other areas (slower CPUs) but with desktop/server CPUs being much faster than memory speeds and remaining so for the forseeable future, having common and popular instructions being shorter than other instructions is actually an advantage despite the complexity that involves.
It's like having on-the-fly instruction decompression. e.g. CISC programs tend to be smaller in main memory+cache, and they travel in CISC/"compressed" form taking up less memory bandwidth over the memory/cache buses to the CPU instruction decoder where they are "decompressed" to RISC micro-ops to be executed.
Look at the mainstream desktop/workstation/server CPUs. Only the SPARC is RISC. IBM POWER/PowerPC is barely RISC[1], some people think it's more CISC than RISC. Itanium isn't RISC. x86 isn't. The rest (Alpha, MIPS, PA-RISC) are either out of the market or on their way out.
As long as CPUs are fast and much faster than RAM (and cache remaining expensive), it's often worth doing the compression/decompression thing.
[1] I believe IBM's POWER chips actually decode their "RISC" instructions to simpler instructions, some of their "RISC" instructions are pretty complex- kinda oxymoronic... But as I mentioned, that may not be such a bad thing.
Pentium 4 reduces the CISC instructions to a series of RISC-like "microops" that, for the most complex of the bunch, can take hundreds of cycles to complete.
Well, we know that the kernel can be made to scale but what about the applications? The same issues the kernel had to face, the applications have to face also. For parallel computing you naturally try to avoid too much sharing by "parallelizing" the programs. For applications like databases, you are talking about a lot of sharing of a lot of data. Not all the techniques the Linux kernel used are available to the applications yet.
is not impressed by these system specs.
Fire up apache and then post a link to it here on slashdot. We love a challenge.
The UNIX made by SGI (the company making the machine referenced in the article) is more scalable than Solaris. Remember, IRIX was the first OS to scale a single Unix OS image across 512 CPUs. And now they've eclipsed that, with Linux.
None of that is unique to Sun.
Better than what? And says who? They've never decisively convinced the market that they're beter at this than HP, SGI, IBM or Compaq.
In addition to ignoring the other good Unix architectures out there in a dumb way with this comparison, you're also totally missing the point of the article. Linux supercomputing isn't just about cheap clusters anymore. Expensive UNIX machines on one side and cheap Linux clusters on the other is a false dichotomy.
Now before I get modded down, I be to remind whoever might read this that what I am saying is FACT. - bogaboga
And what is more interesting, will some of the code that obviously needs to be developed for this kind of system be released under GPL?
After all, in most cases you'd think that the work would be based on GPL'd stuff like the Linux kernel, and therefore the modifications would need to be licenced under GPL also, bringing a lot of technology available for Linux.
http://codeandlife.com
But what about Duke Nukem Forever?
No. Stupid Doom 3 framerate jokes in stories about supercomputers are NOT FUNNY ANYMORE.
They stopped being funny SEVERAL YEARS AGO.
I realize you were trying to defend Sun, but in this case the vendor (SGI) has far more experience with large systems than Sun does. At every point over the last 16 years (since SGI announced the original PowerSeries on 10/4/88), SGI has always supported more processors running a single OS than Sun. Those processors were MIPS based, but the Altix architecture is derived from the bricks/bus of the Origin servers.
The flip-side of this is that SGI has been in decline for several years longer than Sun and may have lost some or all of its edge in this area.
PowerSeries 140 (4x16Mhz MIPS R2000), 10/4/88.
Quake 3 is _already_ out, thats why.
That's almost enough to run Emacs!
Generally, bash is superior to python in those environments where python is not installed.
was just a little too RISCy
Find a job you like and you will never work a day in your life.
Read the article again. This machine will be around 6 TFLOPS, with the total combined NCSA computing power (counting *all* of their machines) over 35 TFLOPS.
It's already out...in Japan
Join Team Mozilla #38050 Folding@home
A beowulf cluster of these .
yeah my 2.6 ghz Pentium 4 is really showing its age.
Join Team Mozilla #38050 Folding@home
Maybe b/c quake 3 has been out for years.
The choice between using a shared-memory or message-passing model depends entirely on the problem domain.
If you have a massive computation which can be divided into independent subtasks, it doesn't need either model. Just compute each subtask on a different processor and collect the results.
Things get more difficult when the subtasks have to share intermediate results. If this information is primarily needed for coordination between specific processing nodes, then message-passing between them is a good model.
A shared-memory model is best for representing algorithms which have to share intermediate results across a large or indeterminate population of processors.
It wouldn't with multi-processor applications if there isn't an os that supports them :)
Privacy is terrorism.
That much hard drive space rivals my porn collection! :O
I know a lot of people have said this is just in time for Doom III, but I think this'll be good for Duke Nukem Forever [In Production].
Where do I sign up to get my hands on one of these?
Government of the people, by corporate executives, for corporate profits.
My opinion is that Linux on a 1024-way is a spectacularly stupid idea, introduced more for the sexiness of having a 1024-way machine than for any practical benefits. Linux is simply not designed for scaling that large. And there is a huge difference between an OS designed to scale that large, and an OS hacked up to support something that large, without actually making the appropriate design choices. SGI may know about those choices (and probably better than Sun), but I highly doubt they'd throw them into a GPLed Linux kernel - they still want to sell their own version of Unix!
I expect (yes, a wild pie-in-the-sky guess) that the advantage of a 1024-way machine over a 512-way machine, both running Linux, is going to be maybe 20-30% performance, far from the 100% the numbers might claim or the 70-80% that might be tolarable. For a supercomputer where that 20-30% is irrelevant because no other machine can crunch the data, cool; for everyone else, two 512-ways running unconnected will be better, cheaper, and faster. [At least, until Linux can scale that large... maybe in 5 years or so?]
A witty [sig] proves nothing. --Voltaire
KDE isn't slow for me either, but it does reserve all my physical memory causing my system to swap.
I have a gig of ram, believe it or not, I don't feel I should ever have to swap, and greatly prefer to leave my physical ram open for other things than KDE and it's various subcomponents. Gnome doesn't do this, I'll stick with Gnome.
If your iSeries or pSeries CPU gets hit by an
alpha particle, your data gets corrupted.
The zSeries protects you with mirrored pipelines.
Of course, for some apps you could simply run
every computation twice without hardware helping.
Who does that though? You can be sure that your
typical database software wasn't written to do that.
"genesis@kurta:genesis$ bc
bc 1.06
Copyright 1991-1994, 1997, 1998, 2000 Free Software Foundation, Inc.
This is free software with ABSOLUTELY NO WARRANTY.
For details type `warranty'.
715776/1024
699"
Don't be silly... you won't have to pay a totally bogus (Bill!) SCO fee... that joke is, like, so old now, it, like, literally, figuratively, as old as the, like, hills. I thought most excellent (Ted!) SGI had licensed SVR4 in perpetuity regardless?
...a part of dtrace? My goodness... they only just got that and now you say they're losing it?
The US Army Research Laboratory will be receiving two 2,000+ precessor super computers running Linux. ARL already has a 256 processor Linux cluster that throws off enough heat to cook the staff dinner...gotta wonder how much these suckers will dissipate.
1024? So what: http://zdnet.com.com/2100-1103_2-5097398.html
The point here is that if performance continues to grow like it is today, they will be selling these machines for $1,000 at Walmart in just 14 years. It will be about the same size as the computer you own now.
The problem with 1024CPU is much more then just the operating system. It is a mess of communication hardware needed to wire everything together. It is about special power feeds and air conditioning, and sometimes floor loading requirements.
Take a quick look at the end of this PDF. It talks about heat output and the need for 3 phase 240V power coming into this computer. It is not unusual to hire both an electricial and a cooling expert when you talk about installing one of these babies. Not for the Home user, and never will be, however, idential compute power comming in just 14 years, so get ready...
I will avoid the tech terms (partly because they would confuse you, partly because I don't know them all but mostly because they ain't needed.
A single CPU computer can execute ONE instruction at the time. Meaning one program thread running at the time. But wait you say, my OS can run multiple programs at the same time. WRONG. It can't. It is a trick. It is running one program at the time but it is switching the program it is running really fast. There is however a problem with this. When it has switched to a program all the other programs are effectevily at the the mercy of the program now running INCLUDING the OS. Wich is why DOS and Windows and Linux and Mac OS and all the others had "hangups". With an extremely well written OS these hangups (when a program doesn't switch back to the OS) can be avoided but it still remains a case that all the programs and the OS are fighting for time on 1 single cpu.
So what happens when you add a cpu? Well a lot less switching PLUS if a program for whatever reason does not switch properly the OS can still be run on the other processor. Just making a windows box a dual CPU instantly makes it far more robuust. I encountered this myself with an old dell P3 that had a dual board but no dual CPU installed. Before I added a second CPU it was the usual windows crap of hangs and reboots and BSoD. Afterwards it ran as stable as a unix machine. Simple things like openeing a complex folder in exploder no longer "froze" the desktop as it could simple run exploder on one CPU and say word or my mp3 player on the other.
Don't forget too that there think like ATA harddrives and CD-ROM need the cpu to drive them. This takes a lot of long cycles and a lot of waiting, not so much CPU power as just time on the CPU. With a second one to do all the other tasks this makes everything run far smoother.
So what is better? Running 1 2ghz cpu or 2 1ghz cpu's? Depends. If you are running 1 program thread go with the 1 cpu. It will take all the cpu time but will not need to share it. If however you are running countless small threads go with the 2 or more solution. Threads will have access faster and you will loose less cpu time on the time needed to execute switches.
Oh yeah that is another problem. Switching between programs takes cpu time as well. It is not unknown for single CPU systems to spend so much time on switching they don't have time to run anything anymore. The old to many running programs problem known from windows but wich affects every OS.
Lastly there is a simple problem. Say you want real power do you go for a quad 2ghz or a single 8ghz. Answer? It is a trick, no such thing as a 8ghz cpu.
If you get the chance buy a second hand dual P3 and install windows 2000+ or Linux on it and be amazed. That old system will respond a lot faster underload then your 3ghz monster.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
Internally referred to as "bricks".
Yes, I have.
Well, if you look at the kernel changelogs, e.g.
linux.bkbits.com:8080/linux-2.4 and
linux.bkbits.com:8080/linux-2.6
you'll see that SGI already contributed alot.
That wasn't the only issue. Compiler technology (especially back-end optimization) was less mature then. CPU architects added lots of support for high-level language constructs (look at all the string ops and BCD support in x86) so that compilers could translate to those ops directly. The alternatives were function calls (with high overheads) or inlining small code fragments to do the complex instructions (which would often not compile very efficiently). David Patterson called this the "semantic gap" (Communications of the ACM, January 1985).
As we now know, this support was costly. My favorite example was Zilog's Z-80 8-bit CPU, which extended the Intel 8080's instruction set with an indexed addressing mode, designed to make it easier for compilers to generate references to local variables in the stack. Unfortunately, the added instructions were a lot slower. With hand-coding, I was able to speed up some critical leaf functions 5x just by avoiding all the indexed instructions and sticking to the old 8080 set, and by loading the variables I needed (not many) into registers at the start.
As compiler back-ends started to approach assembly-code programmers in efficiency, the benefits of such high-level instructions lessened.
This image can reach 140 TiFLOPS, WOW!!!
This image can use 64 TiB of DRAM, WOW!!!
This image can use 64TiBx50= 3200 TiB of virtual memory, WOW!!!
This image can use 16384 TB of HDD, WOW!!!
How increible is this SUPERCLUSTER?
open4free ©
Remember, IRIX was the first OS to scale a single Unix OS image across 512 CPUs. And now they've eclipsed that, with Linux.
u gust/nasa.html
Just to clear things up:
SGI's first 1024p single-image supercomputer was an Origin 3800 running a customized "XXL" IRIX 6.5 kernel. This was in August 2002, almost 2 years ago.
http://www.sgi.com/newsroom/press_releases/2002/a
...how easy it is to install printer and sound drivers?
I don't know about the Altix, but on SGI's Origins you can install a $50 M-Audio Revolution 7.1 PCI card and it'll work right out of the box. Works on any SGI using the IRIX64 kernel. I have one in an old Origin 200 as part of my streaming audio setup. SGI also sells an assortment of consumerish to pro audio cards and video i/o capture cards.
As far as printer drivers, I'm assuming their Altix setup is similar to their Origin setup, which now uses a commercial CUPS printing subsystem. ESP PrintPro I think. Plenty of drivers to choose from if you don't already have a nice laser/colorlaser with built-in postscript and ethernet.
" yeah my 2.6 ghz Pentium 4 is really showing its age.
Join Team Mozilla #38050 Folding@home"
Maybe if you switch off Folding@home you'll actually be able to experience the power of that CPU.
With enough memory there's no need for KDE to be slow. My housemate's P-II 300 with 256MB runs KDE quite nicely, even with OpenOffice and Mozilla both open. Definitely no "slower" than the Windows 2000 GUI
sigaar
And if you want to run Office on top of it, you'll have to double the CPU's and triple the memory.
hehehe got me there. But I don't fold when responsiveness is crucial.
Join Team Mozilla #38050 Folding@home
don't get me wrong, I understand the reliability advatages of the zSeries hardware. You can set the suckers up for 100% availability, if you have the money. Nothing else comes close, with sysplexing, the hardware error detection and recovery, and the sheer stability of the various mainframe OSes (linux not withstanding). But they're not fast machines, and aren't designed to be. They're designed for data processing throughput and reliability.
And they're also kind of expensive. The maintenance contracts, etc. are a lot anually, and you need to find and hire increasingly scarce resources to admin the machines. Sometimes, though, that's worth it; it depends on your needs.
The SCO Group (SCOX) sues both Silicon Graphics and the National Center for Supercomputing Applications (NCSA) for copyright infrigment of their UNIX Intellectual Property.
I've been working all weekend to cluster 4 Honda Civics. When I'm done, I expect it to go 280MPH, get 12MPG and 0-60 in under 3 seconds.
In my opinion, for a 64-bit system, 16 KiB, 32 KiB and 64 KiB are better sizes without big performance's impact. ;)
open4free ©
open4free ©
The UNIX made by SGI (the company making the machine referenced in the article) is more scalable than Solaris. Remember, IRIX was the first OS to scale a single Unix OS image across 512 CPUs. And now they've eclipsed that, with Linux.
Scalability is a complex issue. SGI has put a whole lot of processors together and put a single Linux image on it (so that a single program can use all memory), but this says nothing about how that setup will actually perform for general purpose use. Just because the hardware allows threads on hundreds of processors to make calls into a single Linux kernel, does not mean that there will not be major performance issues if this actually happens.
There are performance issues with memory even on single processor systems with nominally a single large address space, and a developer may need to put a lot of work into ensuring that data is arranged to make best use of the various levels of cache.
Many of the multi-processor architectures require even greater care to ensure that the processors are actually used effectively.
The fact that a single Linux image has been attached to hundreds of processors is no indication of scalability. A certain program may scale well, or not.
Itanic is too expensive and too late.
At 1024 CPUs sharing the same memory bank, I would say they are also making an archiectual mistake. There will be a memory bus bandwidth problem for sure.
Beowolf would be a much better choice. And use AMD 64.
Let's hope we hear stories about a 1025 node machine soon!
- A scheduler that runs out of local data structures as much as possible.
- A memory manager that stores the thread's page tables locally.
- Forcible processor affinity for threads.
- Forcible processor affinity for memory allocation.
- Competent use of communication hardware, esp. semaphores.
All of these problems are either solved or trivially solvable. Not surprising, as the OS's biggest job for this sort of workload is to get the hell out of the way.That hasn't been funny for a long time. Stop posting.
no jerk. I'm tired of all these retarded geeks
that think they are experts on the English language. Any decent student of the English language will know there no such thing as a hard and fast rule in the language. Try reading james joyce. He broke plenty of "rules". There's no ISO or IEEE standard for the english language. There's the MLA but nobody follows every rule in that thing. It's not like programming, wanker, most of the time the idea is to get your point across as clearly and succintly (and if your lucky, subtly) as possible.
I wonder where it will be housed...NCSA's new building isn't complete yet. And it doesn't seem like they would install it only to move it a few months later, does it?
Being an administrator of some 24-way boxes, I have to ask a more detailed question about the error handling. Is the L2 cache in the CPUs just ECC'd, Parity, or fully mirrored? You'll find that on a large installation of CPUs, not being fully mirrored on your L2 will cause quite a bit of downtime over the course of a year with that many CPUs. I don't have those Itanium 2 specs. Anyone?
UPDATE: I looked. Itanium 2's L2 cache is ECC. It'll correct a 1 bit failure, detect and die on a 2 bit failure. Believe it or not, on a large number of CPUs running over a long period of time, it happens more often than you think. It also says it has an L3. No idea on the L3 cache protection method used. Because they don't say, I'd also guess ECC. Wheee! Lots of high speed RAM around the CPU with ECC protection. Well, nobody called this an enterprise solution, so I guess its okay.
Also, you're going to have regular issues with soft ECC errors on that many TB of RAM. And then your eventual outright failures that'll bring down the whole image of the OS. (An OS could potentially handle it 'gracefully' by seeing if there is a userspace process on that page and killing/segfaulting it, but that's more of an advanced OS feature.)
Boy, I'd really hate to be the guy in charge of hardware maintenance on THAT platform.
My opinion is that Linux on a 1024-way is a spectacularly stupid idea, introduced more for the sexiness of having a 1024-way machine than for any practical benefits.
I think there are practical benefits, but only for very specialised applications. As far as I understand it, Linux only really works highly multiprocessor on architectures that aren't really symmetrical (like this SGI machine). Different CPUs don't have the same speed of access to different memory areas as in true SMP.
So, it's not comparable to true SMP as in systems from Sun and IBM.
I am happy to say that I have worked, and continue to work on the current state of the art:
/dev/shm /etc/redhat-release:
http://www.ccs.ornl.gov/Ram/Ram.html
A few notes:
Linux kernel: 2.4.21-sgi240rp04051808_10074
From df, a 1 TB ram disk:
none 1023700704 0 1023700704 0%
From
Red Hat Linux Advanced Server release 2.1AS (Derry)
The machine is actually not nice to work on. It is prone to frequent short freezes (2-15 seconds long; about one every 2-3 minutes, although not evenly spaced out).
at last slashdot can have 1 server no more need for 1024 servers any more
imagine a beowulf cluster of these...
Serenity now, insanity later.
Why is this such a big deal?
Why is this different than say a G5 cluster?
Oh yeah that is another problem. Switching between programs takes cpu time as well. It is not unknown for single CPU systems to spend so much time on switching they don't have time to run anything anymore. The old to many running programs problem known from windows but wich affects every OS.
Like most else you said, it is wrong (read the other replies to parent). A Linux 2.6 kernel does this 1000 times/second, at a cost on the order of 100 cycles. And that's up 10x from the 2.4 kernel. And 100kHz/1GHz = ~0,01% of the CPU time.
The only "too many running programs problem" is related to system memory. If you can't keep it and memory and start trashing to swap, performance will suck to no end. But that has nothing to do with the CPU.
Kjella
Live today, because you never know what tomorrow brings
From what I have read, the Altix machines have an amazing 1:1 relationship of the 'N' of CPUs to improved performance, a computing Holy Grail of sorts.
Thats why they received 'Best of Show' at LinuxWorld last year. It wasn't for a 'me too' implementation.
More recently I had it with windows XP pre SP1, however here the motherboard changed from a asus single P3 to a dual P3, however the change was remarkeble (why XP on a P3 my athlon caught fire, damn those things get hot), previously XP was just the old windows, hanging at times and just not that good. With the dual it truly became good. No crashes or freezes or unexplained reboots.
Of course these are hardly scientific results but I noticed the same on Linux. Dual machines just seem better even when using the same crappy brand hardware (asus is hardly a server grade motherboard maker).
As said I am not an expert am just describing my experiences so that another lay person can understand it. I run and maintain single and dual workstation/gamestations and servers and the duals just perform better.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
fascist raping islam murderer fucker rapist gay father fucking man fuckig boy raping pig fucking islamic shithead. we will fucking CRUSH your religion, fucker , ill get out my .50 bmg and snipe all your fucking clerics for a long ways away.
60 mph = 60 * 1609 = 96540 metres / hour
= 26.82 metres / sec
At 9.8 metres / sec**2, it would take 2.74 sec. to reach this speed from standstill.
Allowing for air resistance, it probably would take longer than 3 seconds to get from 0-60mph if you pushed it off a cliff. Or, if the cliff was less than 36.69 metres, it would hit the ground before it reached that speed.
What would be really impresive would be having sex nine times and getting a baby in one month!