Slashdot Mirror
Cray CTO: Linux clusters don't play in HPC
jagger writes "Linux clustering was touted as the next big thing by many vendors last week at ClusterWorld Conference & Expo 2004. But supercomputer vendor Cray Inc. scoffed at the notion of putting Linux clusters in the high-performance computing (HPC) category. "Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer," said Dr. Paul Terry, CTO of Cray Canada."
While Paul Terry makes some good points, in his statements, including the partial quote from the post, "Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer, said Dr. Paul Terry, CTO of Cray Canada. "At best, clusters are a loose collection of unmanaged, individual, microprocessor-based computers."
Remember to take this with a grain of salt. The inflammatory nature of the comment is nothing more than a marketing ploy to increase visibility of, and sell, the new Cray XD1
Kinetic stupidity has a new brand leader: Allen Zadr.
...a Beowulf... cluster... thingy... doesn't that count?
A quote I've seen before:
"If you were plowing a field, which would you rather use? Two strong oxen or 1024 chickens?"
Maybe he meant penguins?
Nuke Gay Whales for Jesus.
How could Cray be wrong. I mean just becuase linuxis running some of the top 500 computers there is no reason to consider HPC right. What a self serving statement Cray makes....they still dont get it .... there way is a dead-end...
. I love the sound of burning women and screaming rubber....
Dr. Terry's assertions remind me of a Seymour Cray quote I had as my
I'm not picking a side, it just seems interesting that the Cray CTO would echo Seymour's thoughts. I guess it's for business and marketting reasons though, sadly.
Trolling is a art,
You did notice he is the CTO of Cray... Canada??
Those who can, do. Those who can't, consult.
Company officer claims competitor isn't as good as his product. Film at 11.
Seems to me to be a bit snobbish.
Geek Hillbilly
Oracle disclaim MySQL and PostgreSQL as "toy databases", Microsoft claims that "Apache cannot be used for real web serving", and Sun announces that "Intel and Linux simply cannot be used for enterprise computing".
So all those supercomputing labs that use Linux clustering (that invented Linux clustering, even) have been wasting their time?
Ceci n'est pas une signature
I guess that the simple problem is just that the algorithm applied is usually not suitable for massively parallel computing.
Cray CTO: Linux clusters don't play in HPC
By Jan Stafford, Editor
12 Apr 2004 | SearchEnterpriseLinux.com
SAN JOSE, Calif. -- Linux clustering was touted as the next big thing by many vendors last week at ClusterWorld Conference & Expo 2004.
But supercomputer vendor Cray Inc. scoffed at the notion of putting Linux clusters in the high-performance computing (HPC) category. In fact, Cray showcased a system -- Cray XD1 with Active Manager -- that will compete in performance and price with some Linux clusters upon its release..
Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer, said Dr. Paul Terry, CTO of Cray Canada. "At best, clusters are a loose collection of unmanaged, individual, microprocessor-based computers."
Businesses shouldn't expect supercomputer performance from Linux clusters, Terry warned.
"Cluster vendors would have you believe that their performance is the linear sum of each of their respective GFLOPS [Giga Floating Point Operations Per Second]," he said. "Most cluster [experts] know now that users are fortunate to get more than 8% of the peak performance in sustained performance."
Linux clusters do have a place. "For applications that require low performance, they are a cheaper solution," said Terry.
With XD1, Cray intends to make HPC a cheaper solution, too. "With the Cray XD1, Cray will introduce new price points that should make HPC solutions more available to industries that before couldn't afford such devices," Terry said.
Cray XD1 was developed by OctigaBay Systems Corp., a Vancouver, B.C., Canada-based company acquired by Cray on April 2. Formerly OctigaBay 12K, Cray XD1 will be released to some companies for testing in May. Full release is expected later this year.
The acquisition of OctigaBay's technology will allow Cray to move into new markets by "doing supercomputing on a smaller scale with some commercial, off-the-shelf components," said analyst Richard Partridge, vice president of Enterprise Server Solutions for DH Brown in Port Chester, N.Y. "Cray just can't shrink its custom-built supercomputer designs," he said. Having the ability to put a value-added HPC solution on AMD processors is a good way to move downmarket.
Cray XD1 marries the performance of large SMPs with the economics of cluster solutions, according to Terry. It will also pair new interconnect and management technologies with AMD Opteron 64-bit processors in a direct-connected processor architecture. Its parallel-processing capabilities will directly link together processors to relieve memory contention and interconnect bottlenecks found in cluster systems.
"The Cray XD1 is not a traditional cluster; it does not use I/O interfaces for memory and message passing semantics," said Terry. "For HPC, the most important thing is application performance, and the Cray XD1 is specifically designed to maximize application performance."
In some situations, XD1 would be a good replacement for very high-end Linux clusters, Partridge said. He sees the XD1 providing more "compute performance for the dollar" for organizations that do heavy number and data crunching and analysis. He noted, however, that Cray has shown analysts a limited amount of information about the new products.
Terry believes that individual copies of Linux used for HPC today are intrinsically "heavy" and run independently on multiple processors, significantly adding to the difficulty of managing clusters.
XD1's integrated management software -- Active Manager -- will eliminate the "FCAPS" management ills common to clusters. "Fault, configuration, accounting, provisioning and security" are not handled well by current cluster management solutions, he said. "Often times, [management] appears to be done as an afterthought instead of being designed into the system from the ground up," he said.
Active Manager, which was demonstrated at ClusterWorld, offers a single-point of system administration and contr
main(0)
Regardless of whether I agree with the article or not I feel compelled to point out that:
The 1100 node Apple G5 cluster in virginia has yet to run any real scientific code. So far it has only ran benchmarks.
Just to mention, when pointing out an object: 'There is a dead-end', is fine. However, 'their way is a dead end', is probably what you are striving for.
"We are dropping our line of Cray supercomputers and replacing them with rack mounted Beowulf cluster of 486's!"
I am not saying Cray isn't worth it, but there is something to be said on replacing/fixing your supercomputer with over the counter parts.
REading the article it's fairly obvious that Cray's CTO has an agenda, however, assuming he's right, what does play in HPC? Cray Prorpritary Cluser OS (TM) or what?
In other news...
"Despite assertions made by Toyota salesmen, a Lexus sedan is not a luxury car," said Bill Taylor, CEO of Mercedes-Benz.
This space intentionally left blank.
Clusters can get high performance on some types of tasks. But sometimes, you need fine-grained parallelism that just isn't available on a cluster.
On the other hand, high performance usually comes through special hardware. And on that hardware, I think Linux could be the right thing (modulo some patches).
A message from the system administrator: 'I've upped my priority. Now up yours.'
Cray can kiss my shiny, metal ass.
Now, if the CTO of Cray Canada started talking about your mother than I think you're morally entitled and required to respond.
Of course he is going to say this. He is an exec at Cray, what would he say "Oh yeah, our machines are good at HPC but of course you could build a Beowulf cluster fairly cheaply and efficiently and you wouldn't have to rely on us to do it."
Cray used to be a big name in computing but unfortunately for them, they are a relic now. They had their day and it hard to believe that they will be able to compete effectively against Beowulf clusters and Linux mainframes that IBM is pushing. With IBM's public love and more importantly, financial, affair with Linux at the high end, I wouldn't want to compete with them.
No trees were harmed in the composition of this; however, numerous electrons were inconvenienced.
First they laugh at you... check!
I guess all those universities using Linux clusters are a figment of our imaginations.
I'm guessing it went something like this:
"Our Cheaper competition is nether competition nor cheaper"
It's not the vendors who are claiming that Linux clusters are real supercomputers, it's the people who are using them to do real supercomputer work. They sell themselves based on actual price and performance.
Methinks Cray is feeling a little threatened...
Surely he has never seen OctigaBay's computers:
http://www.octigabay.com/
I do not work for them, I was simply amazed by what you can do with these things and how they interconnect with up to 1,000 boxes.
Oh, shit. I just went to their web site and THEY WERE BOUGHT BY CRAY!!!!
Hahahaha!!!! The ultimate Linux HPC is now a Cray product.....This is too funny...
"Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer,"
Maybe so but not everyone can pull a Cray out of his ass when they need horsepower. A Linux cluster is affordable, a Cray is the thing of wet dreams..
Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer, said Dr. Paul Terry, CTO of Cray Canada. "At best, clusters are a loose collection of unmanaged, individual, microprocessor-based computers."
I guess they're not happy about being only #19 on the Top 500 Supercomputer List. Linux is considered faster than they are according to the list.
The 'ol ad-hominem attack of "if you can't beat them ligitimately, attack them personally" just doesn't cut it Paul. Build a better computer.
Ruby on Rails Screencast
The Cray CTO makes the point that Linux clusters get, at best, just under 10% peak as sustained performance and uses this as a justification that Linux clusters are not HPCs. This is a reasonable criticism. Let's take the percentage he cites as real for a moment. Now what is the cost difference between a Linux cluster and a Cray (not some future offering, but today) and how much more of a Linux cluster could you afford? Would that offset the quoted inefficiency? Would the flexibility of being able to use commodity components further offset any advantage Cray might have? What about 24hr or same-day parts replacement without a hyper-expensive service contract? At the end of the day, I suspect the Linux cluster wins out even given the sub-10% efficiency figure Cray cites. --Pat / zippy@cs.brandeis.edu
that Linux first appears at the 6th spot and Cray appears at the 19th.
Who doesn't play in what?
Running with Linux for over 20 years!
SGI came out with some fine high performance linux boxes with up to 256 CPU per kernel check out www.sgi.com
well i guess now we'll have to add them to the list...after SCO and Microsoft of course
Cray CTO: Linux clusters don't play in HPC
By Jan Stafford, Editor
12 Apr 2004 | SearchEnterpriseLinux.com
SAN JOSE, Calif. -- Linux clustering was touted as the next big thing by many vendors last week at ClusterWorld Conference & Expo 2004.
But supercomputer vendor Cray Inc. scoffed at the notion of putting Linux clusters in the high-performance computing (HPC) category. In fact, Cray showcased a system -- Cray XD1 with Active Manager -- that will compete in performance and price with some Linux clusters upon its release..
Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer, said Dr. Paul Terry, CTO of Cray Canada. "At best, clusters are a loose collection of unmanaged, individual, microprocessor-based computers."
Businesses shouldn't expect supercomputer performance from Linux clusters, Terry warned.
"Cluster vendors would have you believe that their performance is the linear sum of each of their respective GFLOPS [Giga Floating Point Operations Per Second]," he said. "Most cluster [experts] know now that users are fortunate to get more than 8% of the peak performance in sustained performance."
Linux clusters do have a place. "For applications that require low performance, they are a cheaper solution," said Terry.
With XD1, Cray intends to make HPC a cheaper solution, too. "With the Cray XD1, Cray will introduce new price points that should make HPC solutions more available to industries that before couldn't afford such devices," Terry said.
Cray XD1 was developed by OctigaBay Systems Corp., a Vancouver, B.C., Canada-based company acquired by Cray on April 2. Formerly OctigaBay 12K, Cray XD1 will be released to some companies for testing in May. Full release is expected later this year.
The acquisition of OctigaBay's technology will allow Cray to move into new markets by "doing supercomputing on a smaller scale with some commercial, off-the-shelf components," said analyst Richard Partridge, vice president of Enterprise Server Solutions for DH Brown in Port Chester, N.Y. "Cray just can't shrink its custom-built supercomputer designs," he said. Having the ability to put a value-added HPC solution on AMD processors is a good way to move downmarket.
Cray XD1 marries the performance of large SMPs with the economics of cluster solutions, according to Terry. It will also pair new interconnect and management technologies with AMD Opteron 64-bit processors in a direct-connected processor architecture. Its parallel-processing capabilities will directly link together processors to relieve memory contention and interconnect bottlenecks found in cluster systems.
"The Cray XD1 is not a traditional cluster; it does not use I/O interfaces for memory and message passing semantics," said Terry. "For HPC, the most important thing is application performance, and the Cray XD1 is specifically designed to maximize application performance."
In some situations, XD1 would be a good replacement for very high-end Linux clusters, Partridge said. He sees the XD1 providing more "compute performance for the dollar" for organizations that do heavy number and data crunching and analysis. He noted, however, that Cray has shown analysts a limited amount of information about the new products.
Terry believes that individual copies of Linux used for HPC today are intrinsically "heavy" and run independently on multiple processors, significantly adding to the difficulty of managing clusters.
XD1's integrated management software -- Active Manager -- will eliminate the "FCAPS" management ills common to clusters. "Fault, configuration, accounting, provisioning and security" are not handled well by current cluster management solutions, he said. "Often times, [management] appears to be done as an afterthought instead of being designed into the system from the ground up," he said.
Active Manager, which was demonstrated at ClusterWorld, offers a single-point of system administration and cont
Wish I'd been there so I could have slapped him after about 3 seconds of stunned silence.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
Marketdroidspeak again. Linux clusters have been a pretty common technique here for many (10?) years. Back then, you could call a bubble sort algorithm "research" if you ran it on a beowulf.
It all depends on the problem you are trying to solve. I have been doing some work of late that would not complete in my life time on the 108 node cluster that we have. But when programmed for and run on two Cray X1s I should complete inside of a week.
Granted there are many codes (and more every day) that will run on clusters, the big iron will never die.
Just because we love Lunux doesn't mean that clusters are HPCs.
There are real issues that differentiate mainframe/supercomputers from large, powerful, clusters.
Of course this all depends on your definition of an HPC. But I believe that it's reasonable to say that if parts of your computer are connected with low bandwidth connections (10/100,gigabit) they just can't handle the same kinds of transactions that a computer with parts that are connected by 10 gigabit or 1000 gigabit connections or whatever it is nowadays.
As far as I know if you're deploying a large database it's still advisable to have a big huge IBM mainframe or a Unisys box or a Sun 10k instead of 4,8 or 16 clustered 8 proc machines.
My point is there are valid arguments for not including clusters of commodity hardware in the HPC category.
In my mind they aren't High Performance Computers... they are High Performance Clusters of Commodity Computers.
~foooo
Where's your God now hippies? Where's your God now?
How well a cluster will do depends on the application that it is performing. Some problems can be divided into several small problems with little reliance on other parts of the problem (SETI / Encryption breaking). These things can be easily distributed to hundreds or thousands of "small" boxes for processing and are what a beowulf cluster would be good at.
Other applications require the breakneck interconnect speeds that large Cray / Sun / etc.. build on. When the data being calculated on one CPU requires data from CPU2 to continue its calculations you don't want to have it wait for 100mbit or even 1gbit ethernet speeds. Even quicker interconnects such as SCALI are going to be slowed by PC bus speeds.
Cray fills an important niche for those who can afford it.
He just wants to pump up OpenBSD.
The comment was stupid, yes, but not all jobs that you'd use supercomputers for can be broken down into many threads as others can. A linux cluster will do well for some jobs, a cray box will do well for others. There *will* be times when a Cray system is so far superior to anything you could do with Linux that it becomes the only real option.
However, dismissing linux cluster technology automatically is dumb. In many cases, it provides more than enough cpu power and I/O bandwith to support your reason for getting a supercomputer, and probably at less cost than the other options.
Its all a matter of determining what you need the computer to do, determining your budget, and get the best system in your budget for the uses you have for it. Sometimes that will be a Cray, sometimes a Linux cluster.
The simplest rebuttal is that its not what you call it that matters but what you can do with it. And, judging by the ubiquitous deployment of linux clusters, the answer seems to be "almost anything under the sun".
Who is this guy and what does a company like Cray know about... oh... never mind.
Clusters can rival a supercomupter when they are assigned is a task that's suitable for distributed computing. That is, work units can be divided up and worked on in any sequence... the result of segment 45 doesn't depend on knowing the result of 44 and such. Effectively, you can have the sum of all of the processors minus just a little overhead for the clustering.
What Cray's rightfully pointing out is that for most business applications, however, distributed computing is not a viable option. When processing on a transaction basis, the transactions often need to posted in the exact order they were recieved, which means they must be taken serially. In those situations, the programs can't multithread work out to the other processors so well, and the cluster will end up running at roughly the speed of just one processor while the others waste clock cycles waiting for something to do.
The cluster isn't the solution to everything. Nor is the supercomputer. You've gotta think about the job, then figure out which tool is right for the task.
Everyone knows THIS is the world's fastest computer.
Further info.
I await the flames from the Jobsians
His rhetoric is quite predictable, actually. He talks at some length about how and why clusters of PCs can't get the job done, and how clustering is inherently inferior to a REAL SuperComputer, then goes on to describe how their new product (which sounds suprisingly like a cluster of propreitary machines) can work. Repeat the above as it applies to the management software.
If clustering doesn't work, and Supers are better / cheaper, explain why large companies (Pixar, NVidia,
Note that this does NOT mean that clusters are suitable for ALL traditional SuperComputing tasks. It really depends on the problem. If the problem is better solved with a vector processor, then a vector machine (like a Cray) is what you want. If the problem is solvable in parallel, then a cluster might be the right answer.
I'm a layman...I have no idea what I talk about, but of course that doesn't stop me.
I know I keep coming back to Virginia Tech, but isn't all those G5's linked together to make the 3rd fastest supercomputer itself a cluster? Or is it considered something else?
And if it IS considered a cluster, then why wouldn't a Linux based (along with the *BSD based G5s) be able to make a fast supercomputer?
If so, then what Paul Terry is spouting is just FUD and marketing to help sell his product, yes?
Just wondering.
"Music is everybody's possession. It's only publishers who think that people own it." - John Lennon.
In truth, such machine will always have a certain performance advantage over traditional clusters. The question is, will the price point be low enough to invalidate the idea of just adding more boxes to the traditional cluster.
"Who are in control, they are not in control of anything - they don't even control themselves!" - Glen Beck
The correct answer is "Two strong oxen." Have you ever seen 1024 chickens plow a field? It's absurd. And so is the analogy. You use what is appropriate for the job at hand.
Cray is losing market share and want people to believe that their expensive to maintain and operate machines are better and 'cheaper' than a clusteer of regualar pc's running together. My question is Does anyone with experiance with both systems back him up.
While I certainly disagree that you can't build a very high performance computer out a cluster of computers (Linux or otherwise), there is a lot of merit to the fact that clusters just don't scale well for certain classes of applications. Hence the renaissance of the vector supercomputer (ala the Earth Simulator ).
Obviously, this guy is plugging the new Cray X1 architecture, which really is quite promising. For instance, check out this paper by some folks at Oak Ridge National Lab that appeared in Supercomputing 2003.
Of course, since this is Slashdot, I expect that there will be a deluge of posts decrying everything about the new Cray machine because it commits the cardinal sin of NOT USING LINUX. Oh, the horror!
I'm sure Hank Dietz would disagree : http://www.aggregate.org
The "interconnect" latency (especially) and bandwidth in a cluster, even using very high-end network hardware, is much worse than that of a Cray-style supercomputer. This does make certain applications run slower, especially if not specifically tailored to clustered architecture. Some applications are very difficult to break down into small pieces and require extensive memory sharing between nodes, which clusters just can't do well.
Cray is one of the few last behemoths left. If they can survive, surely this is not the way. This is borderline-SCOish, or Microsoftish. Oh well, at least the last thing they do won't be selling Crays @ WalMart, as Sun apparently wants to do.
Cray is in denial, and that won't solve anything. What is certain is that they will fall.
Must-not-watch TV!
To build a HPC supercomputer from a cluster you need a parallel network, you can chose between Quadrics, Myrinet or, (if you don't mind it not working) Infiniband. All of these work with Linux and have made large Linux clusters.
What cray are doing is interesting and is going to result in a big computer but it's in the same ballpark as current Linux supercomputers. They are competitiors, they do not dominate the playing field by any means.
And remember folks, Gnu's *not* unix.
I happen to work in a facility that has large had both large supercomputers (cray t3e, j90, sgi) and linux and *nix based clusters (beowulf/linux, compaq/Tru64). The Cray CTO is correct that you can't just call every linux cluster out there HPC. Just about anyone with networking and linux knowledge can build a linux cluster.
What really makes a difference between an HPC cluster and your normal every day cluster is the hardware interconnects used. There is a comment in the artical that refers to not using I/O for memory and message passing. I am not quite sure what he means by that, but I am guessing that he is saying that the network is not used for shared memory/message passing (MPI/openMP/SHMEM).
If a cluster can limit the impact of latency between nodes either through smarter software or faster interconnects then I can't see any reason not to concider a linux cluster as HPC.
Clusters without smarter software tend to be a real difficult coding platforms. Some developments with things like globally shared memory might make the difference, but there will still be the problem of latency between nodes.
Since when has asking a head employee at a company whether they think their competitors products are good ever been a good idea?
the old sheppard said "there are no wolves in these parts, no need to build a fence" and then lost all of his sheep later that night.
sure, go ahead and turn a blind to Linux, but as MS is noticing now, it will bite you in the arse.
A cluster isn't a computer in the traditional sense. It's better for many tasks but worse for some as well.
It's better in that it is not a single monolithic device that can fail and can be built and scaled as needed. It's worse in that it requires different management tools and techniques due to the multiplicity of systems involved. It also fails in the tasks where a single large shared memory is required.
The great thing is that almost all problems today can be handled using a cluster arrangement and thus resources can be saved. The bad thing is that less Cray's will be sold.
--- I wish I could hear the soundtrack to my life. That way I'd know when to duck.
Imagine an Eciton Cluster ....
This is not a political statement. This is not legal advice. It's a frick'n Slasdot post. However: I'm Running For
"Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer,"
That's like saying that the automobile is not a high performance team of clydesdales. That's true, but it may be irrelevant. If it can get you there faster or better, I guess it doesn't matter.
Disconnect your television. Do your own research. Draw your own conclusions. They're probably lying. Don't be a sheep.
VENDOR: Ya our stuff rul3z! Linux Sux0rz!!!
REPORTER: Ya but why are so many people using it?
VENDOR: Those p33ple ar3 like so stooopid dooood! I'm so not just saying this because we s3ll competing products m@n. We rul3z!!! Linux sux0rzz!!! At b3st Linux is like two handicapped dwarves playing NES dooodz!!! We iz L333t!!
The SGI Shared-memory linux clusters are genuine HPCs.
Note: SGI uses Itanics in the Altix, possibly because Intel gave them access to everything they needed to build the memory interfaces they needed for this chips. I'm wondering when we'll see an AMD/64 Altix, and if not, why not.
When all you have is a hammer, every problem starts to look like a thumb.
My friends and I at the university create a beowulf cluster as a project for our Linux class, (just installing software, nothing impressive really, basic system administration) and benchmarked it (24 P200s 32MB ram each 10baseT network) against a reasonably fast computer running the non-parallel version of the same code... (HP C station) and ours finished faster, with the overhead of networking! as to being "a loose collection of unmanaged, individual" We managed it pretty tightly, using nfs to provide the binaries that would be used, thought we did not use NFS for the root partition, as this would put to much stress on our poor fileserving node. If I remember properly, we used dist for password files, and aside from pushing the power switches, we could manage everything in the room without leaving our seats (24 computers running 'xlock -mode matrix" hehe).
Basicly, I disagree with Dr. Terry.
our project writeup is here.
(Please forgive any mistakes or stupiness therein, we were 15, 15, and a 30something non-geek at the time.)
Less look fast, more go fast.
In a sense, what Cray is asking for in the record books is an asterisk to say that "100 clustered computers times 3 MFLOPs each" is as faster as a single 300 MFLOP supercommuter because you can only get the 300 MFLOPs out of the cluster if the task is capable of being broken out into 100 threads that can each be independantly considered.
There are some tasks that still require a true supercomupter rather than a cluster. Imagine an environment simulator... each atomic time-unit in the simulation cannot be started until the previous unit has been completely computed. Such a serialized task doesn't chop up into 100 threads very easily... so most of the speed of the custer will be wasted, while the supercomputer can still go at full speed on-task.
So, a Linux cluster is a powerful tool that can do most of the things supercomputers have been asked to do in the past... but there's still some specific cases where it's just not going to work. Therefore, a Linux cluster is a Linux cluster... and can't be directly compared to supercomputers in a speed race.
The reason that Cray only holds 19th right now is because they have only deployed X1 systems using up to 256 nodes. When the number of nodes is increased, you will certainly see the Cray moving up the top 500 list -- the architecture is VERY scalable.
_PLEASE_ don't compare "supercomputer" clusters with real supercomputers. How fun is a "supercomputer" cluster compared to a real SGI Origin/Altix? Not fun at all for most things.
are they? or are they not? owned by SGI right now?
SGI has bought and sold the company so many times I lost track.
The funny thing about that is, now the same problems Cray is having, SGI is having as well: (trying to sell single supercomputer machines in a market that is heading to clusters because of price.)
For some tasks distributed clusters are better, for others ultra-high-bandwidth Cray-type monsters are better. So what's new?
"'I pass the test,' she said. 'I will diminish, and go into the West, and remain Galadriel.'"
- JRR Tolkien.
So I was at Supercomputing last year, which is one of those big trade shows for this sort of thing. There's a reception on opening night and everybody gets two little drink tickets to redeem for crappy chadronnay or their fizzy domestic swill of preference. Problem is the drink counters "open" way after everybody starts to show up, so there are long lines waiting for the starting bell (mooo, moooooo). I'm standing in one of these lines waiting for a delicious coors light (okay, maybe "delicious" is only in that sentence for the rhythm of it). Been standing there for a Really Long Time when it finally starts to get going, people trucking through getting their drink on, the tantalizing coors light getting closer with every agonizing babystep forward. Just about to get my beer when this weenus walks up, straight to the front of the line, engages in a some poinless chitchat with somebody he pretends to recognize, gets a beer and walks away. In many years at supercomputing, I had actually never seen that happen before. Neither had anyone else in line, judging from the dropped jaws and colorful language that ensured after the guy walked away. The weenus in question was Bob Bishop, CEO of SGI.
My purpose in relating this tale is to point out that Cray is _also_ insulting SGI when it says things like this, and SGI deserves it.
And the XD1 is a cluster; it's a cluster of 6 2-proc Opteron boards with a custom interconnect inside the box that sits directly on Hypertransport and a coupled Infiniband switch. And there's some FPGAs in there too if you care write code to use them. It's still a Linux cluster, just with lower latencies within a box than most (cross-box isn't that amazing).
Computing with the XD1 is like plowing your field with 170 of Doctor Mephisto's six-assed chickens. I think it's cool, but Seymour would not be impressed.
but practically, the performanc is "high enough" and certainly a helluva lot cheaper than buying a custom system.
It's just like the old days, except more so:
and you can end up paying a lot of money to squeeze out that extra performance.Given that Linux clusters can achieve speeds in excess of a teraflop, that available dollars for computer purchases are finite, and that per processor performance and price performance is increasing, the market size for the world's highest performing machine is rapidly vanishing to a set of measure zero.
"Provided by the management for your protection."
Okay. So on the
... umm... yep. Linux.
CPU % Used / Theoretical CPU power you own
The cray owns. Whoopee.
Methinks that's a totally useless number. I mean, Engineering and efficiency is fun and all, but we don't write everything in Assembly now, do we? I'll put money even CRAY don't. Nobody'd get anything done that way.
How about doing the good'ol BANG FOR BUCK? A method whose benefits people can actually measure and enjoy?
Someone feel like dividing CPU benchmark by TCO?
A good TCO, including office space rent to park the setup. Including service and parts. Including finding, training and hiring the people who know how to run the setup. Including the amount of money you pay when your sysadmin[s] gets a better job offer and dump you on a week's notice. And of course, including downtime - especially the kind you get as punishment for not using off-the-shelf standard hardware (the other side of which, as someone here in an earlier post mentioned, is uber-expensive service contracts).
For people who're shopping for clusters and don't have latency on their minds, a PC-based beo cluster will probbably give the most horse per dollar, by a very, very, very long shot.
And people who shop for low-latency-interconnect superclusters tend to buy SGI origins or very big SUNs. Both companies having or being in the stages of letting their own UNIX push daisies in favor of
Way to go, Mr. CTO. FUD on.
-
The Cray XD1 System operating system is Linux
Cray could easily be at or close to the top of the top500 list, their X1 architecture will extend that far. However, for a lot of really important supercomputing codes, it's no contest: The cray will trounce the clusters (linux or otherwise). Those #19 crays are only 256 processors. To get similar performance a stack of xeons requires thousands of processors. Some tasks just can be split appart that easily.
A cray processor has eight floating-point units running at 800Mhz. The big Mac cluster (for example) uses G5 processors which have 2 FPUs at 2000Mhz. Thus the cray has a ~40% advantage. However, the G5 processor has ~4GB/s memory bandwidth. The Cray has ~50GB/s memory bandwidth. If you have a problem that needs to do a HUGE amount of math on a tiny amount of data, the G5 will rock. If you have a problem that needs to do a HUGE amount math on a GINORMOUS amount of data, buy the cray. (for a GINORMOUS amount of money too)
Similaraly infiniband (ala the big mac) is really hot in the cluster interconnect space because it gives 2.5GB/s per node. The Cray gives you 51GB/s.
You need to move a little data, buy a cluster. You need to move a lot of data, buy the Cray.
There's no one solution for all problems.
Cray is dying. The days of fat government sales have been over for a long time. It's only logical to discredit your competitors, especially if you stand to lose a lot because of them.
This is nothing new, nor anything special. For instance, if you've looked at the latest computer magazines, Microsoft is doing the same kind of "it sucks" argument to anything related to Linux in a wide front. For example, Apache lost to IIS in a review, and IIS became the Editor's Choice in one magazine. In the next issue of the magazine there will be somekind of "debunking Linux myths" article. (This certain computer magazine is nothing special, even though it has become a nothing short of an unfunny joke paper written by people who don't have a clue. Some of their readers do have a clue, that's why they cancel their subscription.)
So, to sum it up: it doesn't matter what the reality is, the people who decide only see the image which is created for them. Even if that image is wrong, that's the only thing they decision makers are going to see. They don't have the time or the energy to investigate things thoroughly on their own. This is why Microsoft pays the magazines to write garbage. This is why a Cray executive talks garbage.
Lobbying is pretty powerful stuff.
I do not moderate.
The Japanese Earth Simulator shows that the Japanese chewwed up and spat out the US competition using vector CPUs and a Super-UX UNIX-based OS. Not Linux and not scalar CPUs which Linux is aimed at.
The joke being that Seymour Cray always was a vector processor innovator but the narrow minded US goverment pushed towards scalar.
Nah, sounds like you have a scalar product to flog now and are playing catchup by childishly poking fun at the competition. So far Linux has worked very well in what many would consider to be valid HPC environments of film production and at a lot lower cost than what you've made to date.
When you get a product that Digital Weta can use cost effectively then you can say Linux isn't a valid HPC platform.
A lot of what he said isn't much of a surprise, however definitive statements about clusters not being supercomputers and being unmanaged loose collections of machines are a bit overblown. Management software exist for clusters and they are rather easy to program for with available popular and industrial strength libraries.
Moreover many HPC applications actually scale quite well on clusters of Linux systems. Affordable interconnect infrastructure is increasing in bandwidth and reducing in latency, further broadening the scope of the problems these clusters can tackle. In addition each node can now comfortably have 2 or four processors giving even better bandwidth between CPUs sharing a node. With 64 bit processors and operating systems now available the final barriers to very impressive easy to use HPC Linux clusters have been removed which is exactly why Cray now sees them as a threat. Now is probably the worst time to talk of how a cluster is not a supercomputer. Clusters form a class of supercomputer that can now handle most supercomputer tasks. True there are classes of problems that the dedicated supercomputer systems CRAY sells will excell at, however clusters are useful workhorses in the supercomputer world and hold their own.
Todays supercomputer problems are tomorrows computer problems and Cray must continue to find new classes of problems to solve as they always have, rather than attacking competing technologies, people will use clusters where the clusters meet their needs.
True Parallele Programming with computer with over 16 or so CPUs is a slightly different mindset then the way most people program. In PP you can write a sort routiene that runs in O(log(x)) While with one processor system you can only do it in O(x). Most programs today that are threaded tend run a buch of code on one processor and its own memory. That is much the way that linux clusters work, by writting programs that minimalize the amount of comunications needed so then they provide high performace. But crays and the like super computer allows all the processors to comunicate with each other and the shared memory a lot faster. Thus making some algorithms run in Maginatudes faser.
An example is when I took a course in Parrallel processing we used a MassPar system which had 1024 processors in a grid formation. Now woring on that system I was able to sort a list of a million random numbers way faster then my Duel Processor PC could.
But on the flip side when I ran a program on the MassPar that wasn't designed parallel processing (emacs) it took upwards of 3 minutes to load it due to the age of the computer. While my PC could open up emacs in a split second. So on the clusters even the fastest in the world a Cray that may not be the fastest could actually beat it on many applications because of the faster bus comunication.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
Weird- first time I read the subject line I thought it said "Crazy CTO: Linux clusters don't play in HPC" - and I was like, yeah, that guy is nuts! :)
You know Goodwin's Law? Well, here is Heironymous' Corollary to Goodwin's Law:
"Anyone using the terms 'zealot' or 'FUD' in a Slashdot discussion is immediately declared the loser of the thread and discussion stops at that point".
Of course I'm force to break my own corollary to make this point.
But to call me a "Linux zealot spouting FUD" (and excuse me for paraphrasing your lucid comment) because I mock a commercial vendor who says that the free alternative is no competition... WTF?
As it happens: I have 20+ years of experience in IT and I've used every one of those packages (except the Cray). Oracle, MySQL, IIS, Apache, Sun, Solaris, Linux. And hundreds of other platforms, as well.
My opinions are not those of a zealot, but pretty impartial and generally very accurate. There is a good reason, for instance, why the most critical servers in my business all run Debian Linux, why the desktops use Xandros, why our applications use MySQL, and why we're phasing our out Microsoft/COM+/IIS/SQLServer platforms. Zealotry has little to do with it, but good sense does.
The facts are these: open source, free, commodity IT has become good and cheap enough to exceed the capabilities (at any price) of many commercial systems. Most specifically, Cray, Oracle, Microsoft, and Sun find themselves spot center of the area that has been commoditized.
Ceci n'est pas une signature
"No, a Kilochicken is a 1000 chickens. You're thinking of a kibichicken. Check it out at http://www.nist.gov" Somebody had to, right? Right?
I'd like to see Paul Terry say this in front of everybody at the Super Computing conference where they announce the Top 500 Computers. Its worth noting that he is not bashing Linux per se, but "Linux Clusters", which is pretty arbitrary, because he should be saying "all clusters", because the OS really doesn't have too much to do with it. Supercomputing apps run in userspace, not kernel space, and the hardware, including interconnects or some kind of interprocessor communication drive the performance.
The Cray XD1 looks like a nice system, but there are only theoretical performance values given, and noone can go out and buy one of these things yet. I also don't know how much these guys cost.
I love this statement:
Linux clusters do have a place. "For applications that require low performance, they are a cheaper solution," said Terry.
Yeah, when we spend a million+ dollars on a supercomputer, we are thinking of low performance, because our applications require it. Thanks.
I'm guessing this guy is a wannabe marketer who got stuck in a CTO position. There are plenty of HPC vendors out there, and trust me if this XD1 has a good price/performance and they work (this is key), then people will buy them with little questions asked. Otherwise, this whole article is just an advertisement that makes many statements without any evidence that the XD1 is any better than 4 Xboxes connected together over a serial connection. Next....
Or prove that no such problem exists :).
All problems solved by computers are solved by a series of steps or computations (c_x). Each c_x has input and output. Say the input of a given c_x doesn't depend on the output of c_x-1. In this case, we can easily parallize the problem by computing c_x and c_x-1 in parallel.
Lets say that c_x does depend on the output of c_x-1. Let the number of possible outputs of c_x be n. We can run c_x in parallel for each of the possible outputs of c_x-1. All these computations are done at the same time as c_x-1. When they complete, we can use c_x-1's output to choose the solution of c_x without running any extra computation.
Well, it's a sketch anyway.
doesn't this CTO of cray remind you of someone?
"There IS no Linux in high-performance clusters."
"There IS no Americans in Iraq."
OMG! It's the former Iraqi mis-Informed-ation minister!
Especially when 2004 has been dubbed the year of the penguin, it's wreckless to claim that Linux can't be used in HPC's.
Hell, just look at the current top500 list. There's no Cray in the top 10 but there are two Linux based clusters there (and one based on OSX [FreeBSB based]).
Here's a few:
NCSA's IA32 Linux cluster
NCSA's IA32 Linux cluster
Space Simulator Clust at Los Alamos (SS51G based; makes me proud as I have a SS51G too)
Beowulf - used in many Linux clustering projects
Linux clusters at Los Alamos (they seem to have more than one)
Virginia Tech's Supercomputer X
Think of two large corporations with 10,000 employees. Corporation A spends a lot of money on this new Cray, and corporation B spends about the same amount of money (within a factor of ten) on a Linux cluster "supercomputer". Right off the bat, Corporation A probably realizes some real gains, whereas Corporation B spends a bit of time playing (R&D) with the Linux cluster. The geeks at Corporation B stop and think "wait a second, we have 10,000 computers sitting around doing nothing for 16 hours a day ... maybe we could find a way to tap into that power." ... and they set about to find a way to start using this idle computer power.
Which company has made the wiser investment, in terms of the long-term picture?
I have a rack right here housing a linux cluster. 36 1U dual-Xeon servers. On the Cray XD1 site it details the "Exceptional Performance" of the XD1 system. It details the performance of a system with 12 AMD Opteron processors, and the performance of a rack of systems with 12 AMD Opteron processors. I understand that the underlying architecture of those servers may be vastly different than the servers in the rack next to me, but fundementally aren't they both multi-processor PC servers operating in a cluster? If so why does their rack full of multi-processor systems qualify as an HPCbut mine does not?
"We can't solve problems by using the same kind of thinking we used when we created them."
We have two linux clusters right now going on to a 3rd. Granted we have had major headaches with installation but I blame that more on IBM.
They work perfectly fine for us. I'll take them any day of the week over a windows cluster or any other type cluster.
Despite assertions made by Linux vendors, a Linux cluster is not a high performance computer, said Dr. Paul Terry, CTO of Cray Canada. "At best, clusters are a loose collection of unmanaged, individual, microprocessor-based computers."
Although this statement reeks of FUD, he's right about one thing: a cluster is not an HPC... that's why its called a cluster. But to say that a cluster is 'unmanaged' is one hell of a stretch IMO. All in all, he's just arguing semantics: nothing to see here, put down your flamethrowers, move along folks.
Since this is slashdot, I'll add that the rest of the article is full of choice quotes all of which point squarely at basic FUD + marketing spin for their new cluster-cost-like product.
It seems to me that Cray is just plain bitter that Linux (through all the cluster solution providers) has managed to steal Cray's thunder at a mere fraction of the cost. Cray's probably even more bitter that folks are willing to sacrifice performance (at least from Cray's perspective) just to save a buck.
Okay, this is Cray we're talking about here: people are saving millions of bucks all over the place by using clusters instead of big expensive machines.
And guess who wants 'their' slice of the pie back.
...there. Have phun now.
Chickens are cheap and easily replaced. Oxen are expensive and require lots of training. If one dies you will have lots of downtime.
Plus chickens lay yummy eggs.
So long as feed is cheap, I'd go with the chickens.
VOS/Interreality project: www.interreality.org
>Let the number of possible outputs of c_x be n Well, if n is a 32 bit number you need 4 billion computers working on c_x.
There are certain types of computing which simply cannot be done with microprocessor based platforms including clustering. One of these calculation types is vector processing. A Cray supercomputer is a vector processing based unit. When comparing a cluster of PC systems being used to calculate what a single Cray is designed to calculate, the Cray CTO is perfectly correct in his statement.
The problem is simple, seed a random number generator with the processor id number and sum up the first N numbers generated from that seed, where N is the number of computers in the cluster. Why are we doing this? I have no idea, it's just my example problem.
If you want to do this in parallel you have to grab the seed from the machine which is initiating the problem, as the seed is tied to its processor. Then you have to scatter that seed. Then each node in the cluster has to generate M random numbers from that seed where M is the computer number of that node in the N node cluster. Then we can gather the values, adding the values that are gathered each step of the gathering for (lg N) time of actual addition. The scatter and gather also take about (lg N) time if you consider the transfer to be on the same order of time as an addition (if I remember right...). At this point we're looking good. Sadly, remember there are N operations for the last node in the cluster to get the Nth number from the random stream. O(3 lg N) + O(N) is on the order of O(N)
Now consider doing this on a single machine. N operations to get the full set of numbers to add plus N-1 additions is O(N). There is zero speedup by doing this in parallel. If you consider that data transfer is hardly on the same time scale as computation (maybe with uber-expensive interconnects, maybe) then you're actually wasting time by sending the problem through the cluster, resulting in a slowdown rather than a speedup. I suppose you can argue that it's parallelizable, but I'd call anybody who wanted to do this in parallel a fool. No matter how you split it, some machine will be forced to do N operations to get the last number out of the stream.
I'll grant you that the example is contrived. However it clearly shows that there exists a problem which cannot be sped up by being run in parallel. It is by no means a stretch to think that there are interesting problems that rarely see a speedup of over 8% when run on a cluster.
If not now, when?
I'm seeing alot of single threaded versus multi-threaded arguments.
That's great and all, but for a single threaded application a cray isn't even going to smash your modern top of the line home pc by too terribly much.
crays are massive smp systems, they need a multi-threaded app to take advantage just as much as a cluster does. The difference is in the bus speed. A cray has a much faster bus, and with equivelent processing and memory it will excel with a number of small quickly terminated threads, whereas a cluster will as well or better with larger more processor consuming threads.
Why would a cluster ever do better? Simple, although a cluster has a drastically slower bus, there is memory local to the processor in question so there is much less congestion on the bus, and since if your shelling out for a cluster you will be switching rather than hub style whatever you do there will be almost without collisions and bus contention. Each node has it's own ram so there isn't much of an issue with contention for the bus and much greater memory throughput.
So like I said, it's all about how fast threads spawn and terminate, because if your rapid firing threads then you will doing alot of communicating between nodes over the slow bus (network), if your sending good sized chunks of data do something and keeping your nodes busy they will spend more time working and less time communicating results and your cluster will tromp all over that cray.
Octigabay bay was recently purchased by Cray. That is how they are able to produce their new XD1 which is an opteron based linux cluster at an affordable price point.
For great justice take off every sig.
And writing code for parallelism at that level of granularity is not trivial.
Cray is dead, their old methods are dead, When VT, and Apple can build the 3rd fastest machine ever and do it for under 5 million (not including other facility expences). Where is Cray on the list? This guy is all about the marketing approach and spreading the FUD to the masses to try to gain a market share again.
MacOSX, because making *NIX better is a lot better than waiting for Micro$loth to fix Windows
Either way, two racks of the new Cray thing is just as fast as all those racks of G5's. Cray's (or anyone elses) custom interconnect is always going to be faster then a traditional cluster configuration. And it acts as more of a unified computer then a bunch of seperate nodes AND you don't have to buy things like video cards, sound hardware, etc, etc, that are just wasting money and OS resources in your cluster.
If you are going to go cheap, get a cluster and a hundred people to manage it. If you're going to do it right, get the right equipment and 1-2 people to manage it. Sorry, I'm with Cray on this one!
-m
PS: Why do you think all the hardcode UNIX servers are operated through a serial port? Virtually no overhead!
http://www.invisik.com
Is a cluster actually a low-cost solution in the long-run, when factoring in electricity costs and additional requirements to disperse heat?
Hate Penguin Computers.
Less is more.
If I wanted to remove insects from a field, would I prefer 1024 chickens or 2 power shovels?
Once there was a one, then there was another one, then another one, until there was enough of them to call themselves a cluster, but then another one started, lo an behold it multiplied into another cluster, then another, then another, then another 'till one day there were so many of them they had to call 'em a cluster, this situation was getting messy we were ending up with clusterings of clustered clusters, then one bright spark said instead of calling it a cluster, lets call it a one :)
What's the state of the art of proofs of parallelizability [and non-parallelizability]?
Is there a standard list of problems that have been proven to be non-parallelizable? Are there any problems that have been proved to be parallelizable, but for which no parallelizing algorithm has yet been discovered? Is there anything analogous to the NP-completeness conjecture in this field?
That Cray and there supercomputers had disappeared years and years ago
Saying Apple is better than MS is like saying Botulism is better than rabies.
And you figure Cray's "Cluster in one box" approach will take less power/cooling? Weren't the dam things cooled with liquid nitrogen at one point because of the Nuclear power plant like temps the things produced... Bob
MacOSX, because making *NIX better is a lot better than waiting for Micro$loth to fix Windows
Cray is owned by Sun, isn't it?
Yeah, but slash puts spaces in long strings with no spaces, so you have to do a hyperlink for people to be able to go there without editing it. so this beowulf link works but this http://www.ibiblio.org/pub/Linux/docs/HOWTO/other- formats/html_single/Beowulf-HOWTO.html gets an extra space in it and doesn't.
That which is done from love exists beyond good and evil
At what price point does the Cray XD1 come in? While huge clusters are (supposedly) cheap individual computers -- I would argue that G5s are not inherantly cheap -- how many G5s that make up the Virginia Tech cluster would you have to get to before you've paid for a Cray XD1?
I mention this because the article implies that Cray is planning on selling the XD1s at a price point cheaper than equivelant clusters. If they succeed at making the XD1 cheap enough, then it may be more cost effective to [[ effectively, cluster ]] a couple of these Crays, with less power consumption, heat dissipation and plain old real-estate.
It seems to me that TCO would be cheaper for the Cray, especially considering that the best clusters expect 5% of the member computers to be broken at any given time.
So, does anybody have Cray XD1 pricing? That, seems to me, to be the only way to rationally decide on the 'better' solution.
Kinetic stupidity has a new brand leader: Allen Zadr.
I'd be quite curious to see it (as I didn't get this interpretation out of the article). Though I, personally, make no claims to having intimate knowledge of Cray's offerings (which is why I'm honestly asking for info).
Kinetic stupidity has a new brand leader: Allen Zadr.
Those other products are not as good as ours...no really we are better, ... so buy our stuff.
What would you expect him to say other then that. That would be like 7UP saying "sure Sprite gets the job done".
Repeal the 17th Amendment TODAY! Also Please Read http://www.gnu.org/philosophy/right-to-read.html
As the subject says, you don't really cluster in the database world for performance. You do it for a number of reasons: 1) Cause in Oracle is cool tech to work on (most importantly) 2) If one node goes down your database keeps chugging alot, just slightly slower. 3) Cause its cool. Saying that the largest that I know of run off of big SMP systems like the HP Superdome. Oracle did a presentation on linux clusters and told us it was cheaper than large SMP boxes and could scale better. Well, recently we had a new project interested in RAC and for our test server we wanted 2 dual cpu blade servers, hardware wise cost 20k. Well, the Oracle licencing on top of that added 80k. I have no idea how running RAC could be cheaper than a single SMP box cause the licensing will kill ya. But anyways, you cluster so you reduce the points of failure and for scalability. Its well noted that Oracle scales excellent in a clustered environment, so you buy a couple of cheap boxes and just toss more in as you need extra capacity.
"Thanks to the remote control I have the attention span of a gerbil."
Sorry.
"Thanks to the remote control I have the attention span of a gerbil."
And it's worth noting, for the uninformed, that the reason Slash puts in the spaces is that there used to be a problem with page-widening fucks screwing up the comments sections with deliberately wide strings.
It's not a 'bug' in the slashcode.
resigned
Put up yore boxen and weee'll seeee
MSBPodcast.com The opinions expressed here are my own. If you don't like 'em... Think up your own stuff.
Every other machine in the top 10 is built from standard processors. The old DEC Alpha, PowerPCs, and IA-32 predominate, with a few Itanium machines.
Because supercomputers today have several thousand processors, they can't even be big shared-memory multiprocessors. Speed of light lag in the interconnects would slow everything down. It just takes too long for the signals to make it across the room.
So all supercomputers today are clusters of one kind or another, fast machines with slower interconnects between them. The hardware architecture revolves around interconnect schemes. The software architecture revolves around working around the limitations of the interconnect schemes. Tightly coupled problems don't map well to such machines.
Bear in mind that we're talking about clusters of uniform machines located near each other with gigabit or better interconnects. We're not talking about "clusters" consisting of spare-time programs out at the end of Internet connections. Those are useful only for problems with almost no coupling between parts. Such problems are usually low hit rate search problems, like cryptanalysis, SETI@HOME, and such.
Yes, there's the Cray X1, the last of the liquid-cooled monsters, but it looks like the only customers who bought one were Government agencies with old Cray machines.
Sure, if you can afford spending exponential space, everything will be done trivially in constant time. But can you?
Cray supercomputers are a big number of processors all in the same machine, and more importantly all sharing the same memory. Each processor has the same delay to access any memory content.
In general, that's not entirely true. You usually have local and non-local memory for each cpu (hence the name N[on]U[niform]M[emory]A[ccess], for when you actually make this distinction). Still, access to a non-local memory location is a lot less expensive than non-local access in a cluster (different machine).
Most readers have the right idea - you don't listen to a competitor's opinion when judging whether something is viable or not. It is very easy to twist the words to be "true" while misleading.
A cluster isn't a supercomputer, by definition, but for many jobs can be equal or better. In other words: Those 2 oxen cost more, consume more resources, are only useful for the one job (pulling a plow) and only benefit a single owner. Those 1024 chickens cost less, consume less resource, are useful for many jobs besides the one (including laying eggs) and benefit their many owners.
I was taking one day at a time, but then several days got together and ambushed me. (from a Rhymes with Orange comic)
For example, in the SETI project, there is a hugh amount of data which needs to be analyzed, but one chunk of work has no dependence on another, so the work can be spread to thousands of workstations all over the world. This is the hallmark of an "embarssingly parallel" problem.
Does anybody actually know the price on the new XD1? Not the X1, but the new machine that Cray is pushing here.
Having a comparison price point would really help a lot of us feel more informed.
Kinetic stupidity has a new brand leader: Allen Zadr.
oh come on, you'd rather have the 500HP 9620 Deere than the 95HP 6003 series surely! i'm not sure what 500 horsepower is in chickens, but it's a lot...
What people seen to be missing is that Linux doesn't only do multi-processor architecture via clustering a' la' Beowulf.
In the traditional clustering architecture you have a lot of standalone machines that operate each in their own memory space an each use their own operating system image.
This is great when a problem can be broken into a lot of nearly independent pieces.
When you have a lot of interconnected pieces as might happen in matrix problems - finite elements or big matrix inversions it is best to have a single block of memory accessed by many processors simultaneously. This cuts down on the penalty for inter processor communication.
Linux has had NUMA from since the mid 2.4 days.
The SGI Altix uses this technology with 256 processors operating with a single Linus OS image on a single block of memory!!!
I can just imagine some ad involving flying penguins zipping past a jet or something..
Hmmm... Sandia and several other US government labs seem to think different. Exactly *what* class of computations can a linux cluster not handle?
UC Irvine
University of Cyprus
Linux supercomputer for Los Alamos
AMD Tapped for Gov. Linux Clusters
Installing, Running and Maintaining Large Linux Clusters at CERN
And more....
"Who are in control, they are not in control of anything - they don't even control themselves!" - Glen Beck
aid Dr. Paul Terry, CTO of Cray Canada.
I never claimed it was a bug, I simply described the behavior in a non-judgemental manner. However you are right, it's a feature not a bug. I rember the long string assholes fucking up entire pages, and the slight inconvinence is well worth it.
That which is done from love exists beyond good and evil
There's no one solution for all problems.
In this case, there really is: money. You can go on and on about theoreticals all you like, but whether you use 256 CPUs to get a #19 or 2200 to get #3, all systems have a cost associated with their resulting performance. Big Mac made news because they did so much with comparatively so little money. If Cray wants to make a hard number claim at bang-for-buck, let them. Until they do, or if they can't, their business will continue to dwindle in favor of cheaper clustering solutions.
$5 / month hosted VPS on linux = awesome!
The idea behind tracks, instead of tires, is that tracks reduce soil compaction by spreading the weight load over a greater surface area (in the same way a heavy snowmobile floats along over deep snow while a comparatively lighter human sinks to their hips). Bonus side effects of the greater contact area include a greater overall production of pulling power (ie. it's easier to cut through tough ground with a plow or other implement) and less tendency for the tractor to "power hop" when digging deeply, which is common with standard tires (a bouncing effect produced as the drive wheels cycle through breaking free and then digging in).
Rubber-tracked tractors are generally only used in the early stages of standard tillage when a field is being turned over (ie. with a plow) or broken up (ie. with a set of discs or toothed cultivators). The headlands of the field are torn up more with tracks when you turn, but at that point in the season it doesn't matter much. When you're done working the field, you just finish off the headlands as best you can and move on. When it comes time to plant, fertilize/spray, and then harvest, you go back to standard tires.
This being said, tracked tractors are only really used in areas where soil compaction is a big issue, like wet/low-lying areas or areas where the land is really heavy. I've never seen anyone around our part of the country (Southern-Central Ontario) using a tracked tractor.
Me? Debunk an American myth? And take my life in my hands?
As Clusters and Mainframe computers alike, in educational and general research environments usually do multiple classes of things with their computing power (including renting out CPU time) - the raw processing power is often more important than the views necessary for specialty computing.
In the end, most super-computing facilities will want to have both Clusters AND Super-Mainframes, but which to get first could be quite a price/performance question. Either way, once you've sunk how much?? money into purchasing one or the other - there's little chance that most computing needs will be custom written for that platform anyway. I don't see a lot of OSS for Cray-Frames, that's for sure.
Kinetic stupidity has a new brand leader: Allen Zadr.
From the description there it appears that while this guy is bashing Linux's ability to cluster, they seem to have no problem with it's ability to SMP against a specialized chassis. It also appears that their system a half-breed between what clustering offers and what a single computer can offer. Honestly, pretty cool technology, but still no excuse to bash the capabilities of Linux clustered.
Too bad that they couldn't have released with the much faster (for SMP, especially) 2.6 Kernel.
I wonder where we can download their Kernel Patch sets.
Kinetic stupidity has a new brand leader: Allen Zadr.
He's probably actually pushing the XD1, formerly the OctigaBay 12K. Paul Terry used to be the OctigaBay CTO, remember.
Looking up from my laptop, I SWEAR I can see a 192 node cluster - Bunyip at the ANU - running Linux. It says "high performance computing cluster" though, so it must be a figment of my imagination.
NOT.
OzymandiasNTTTTtis it custers llasts stand?//
It's called an elephant's trunk whereas it is in fact, an elephant's nose, a nose by any other name would smell as sweet
From Cray web site: "The position involves developing and maintaining infrastructure, utilities, and applications for providing a single system image (SSI) over distributed computing resources running the Linux operating system." Interesting, isn't it ?
"Think globally, act locally".
African or European chickens?
I found this interesting article on google when looking up some X1 specs. Brings up some other majpor factors (space, electricity, heat, etc.) for comparing the two.
Sadly, PS/2 was yet another victim of USB, which doesn't care what you plug into it, the electrical slut.
Which leads to my pondering thought, I wonder if all non-parallel algorithms can be rewritten in a parallel form? For example, I've personally worked the proof that all recursive algorithms can be written as iterative algorithms (and vice versa) as part of my coursework. I've got a gut feeling that the same "equivalence" won't hold for parallel vs serial algorithms, but I'd like to know if it has been formally (dis)proven.
I asked this earlier in the thread: Provably non-Parallelizable?
Allow me to ask it again: What's the state of the art of proofs of parallelizability [and non-parallelizability]?
Is there a standard list of problems that have been proven to be non-parallelizable? Are there any problems that have been proved to be parallelizable, but for which no parallelizing algorithm has yet been discovered? Is there anything analogous to the NP-completeness conjecture in this field?
I asked this earlier in the thread: Provably non-Parallelizable?
Allow me to ask it again: What's the state of the art of proofs of parallelizability [and non-parallelizability]?
Is there a standard list of problems that have been proven to be non-parallelizable? Are there any problems that have been proved to be parallelizable, but for which no parallelizing algorithm has yet been discovered? Is there anything analogous to the NP-completeness conjecture in this field?
I asked this earlier in the thread: Provably non-Parallelizable?
Allow me to ask it again: What's the state of the art of proofs of parallelizability [and non-parallelizability]?
Is there a standard list of problems that have been proved to be non-parallelizable? Are there any problems that have been proved to be parallelizable, but for which no parallelizing algorithm has yet been discovered? Is there anything analogous to the NP-completeness conjecture in this field?