Latest Top 500 Supercomputer List Released 130
chrb writes "BBC News is reporting on the release of the June 2010 Top 500 Supercomputer list. Notable changes include a second Chinese supercomputer in the top ten. A graphical display enables viewing of the supercomputer list by speed, operating system, application, country, processor, and manufacturer."
Computers keep getting faster (Score:1)
Re:Computers keep getting faster (Score:5, Interesting)
I think power requirements are probably the main problem, rather than the hardware. It must be pretty trivial to add more cores to a system that's already using tens of thousands of them, but you're going to need a lot of power.
These systems are only really getting "faster" for parallel tasks too - if you gave them a sequential workload then I assume they would fare worse than a high end gaming machine!
Re: (Score:3, Insightful)
These systems are only really getting "faster" for parallel tasks too - if you gave them a sequential workload then I assume they would fare worse than a high end gaming machine!
I doubt it. A good fraction of them use POWER6 processors, which are still a lot faster than any x86 chip for most sequential workloads. On top of that, they typically have a lot more I/O bandwidth. They might only be a bit faster, but it would have to be a really high-end gaming rig to be faster.
Re: (Score:3, Informative)
"A good fraction" in this case means: Less than 10%. In fact, only 42 out of 500 use POWER.
Re: (Score:1)
What i wonder is, what % of flops they are making verses all the others... if it wasn't for the top 5 having close to 1mil cores, they might make up more then 10% of the computational contribution, no?
I would like to see this graphic chart to include that...
That, and perhaps a distribution chart too... i'd like to see how the PS3 is fairing now in distribution
Re: (Score:2)
I would like to see this graphic chart to include that...
They got 18% of the performance: http://www.top500.org/overtime/list/35/procfam [top500.org]
Re: (Score:2)
Re:Computers keep getting faster (Score:5, Insightful)
Parallel tasks are the whole point of using a supercomputer. The gains made in speed for sequential tasks really haven't been that great; Moore's Law for sequential tasks fell apart a while back.
Being able to parallelize a task is a prerequisite for putting it on a supercomputer.
Re:Computers keep getting faster (Score:5, Informative)
Parallel tasks are the whole point of using a supercomputer.
Well it is now. The original supercomputers were based around a single very fast processor, and had a number of co-processors whose sole purpose was to offload IO and memory prefetch, so the CPU could churn away without interruption. Modern out-of-order CPUs are effectively an old style supercomputer on a chip. Heavy use of parallel processing didn't really take off until the late 80s. This paradigm shift is what caused the supercomputer market crash in the 90s, as development devolved from custom CPUs, to throwing as many generic cores at the problem as you can and using custom interconnects to mitigate parallel overhead.
I thought it was evolved pseudocode. Yay DEC Alpha (Score:1, Interesting)
I always thought computers (and supercomputers) were nothing more than proprietary implementations of someone's attempt to simplify their pseudocode. It all boiled down to memory and bus bandwidth issues, not the speed of the processor. That's where the DEC/Compaq/HP Alpha was retired as was HP PA-RISC, yet theoretically the Sun SPARC and IBM Power designs should succeed. Instead we see these astonishingly bogus processors that you call "general purpose" when they are nothing more the biproduct of bad ma
Re: (Score:2)
Back in early 1980 I headed up a team of techs to install Wang VS systems all over China. I had to take a train to Harbin (far north) from Beijing because it was still too cold to fly there.
I was a visiting American scientist.. and as such in each province i visited the governor would have banquet in my honor and we would all drink wu-shing pigu (5-star beer) and eat great food. For those who know me personally know that I do not like being the center of attention so this was really out of my comfort zone.
D
Re: (Score:2)
Actually, having been involved in a fair amount of work with the CERN and Fermilab parallel computing groups, I can attest that they are using the supercomputers at their disposal for highly sequential data processing, mostly parsing a data set looking for a given pattern. For the amount of data they collect, this sequential processing is paramount.
Consider trying run a SELECT on a non-indexable 2TB database table. You have to look at every row, and it takes time. But if we can parse the table by several ma
Re: (Score:2)
I wouldn't worry too much. We've been pretty good at finding things to help us keep up with the effects of Moore's law.
June?! (Score:5, Funny)
Holy crap, the supercomputers are so fast they're in the future!
Re: (Score:2)
It's like an early Christmas ...
And speaking of that, a nice present would be an account on a supercomputer for running whatever I want. A Top 5000 would do, presumably.
Hopefully the universe won't mind if we call May June and thereby manipulate Moore's Law in our favor. Year over year if Moore's Law holds while the calendar grows shorter, the light speed barrier shall be overcome.
At any rate, this is the age of the Internet and global news updated by the minute. Supercomputers are expensive to upgrade so k
A 2nd "Chinese".... (Score:1)
Looks like a 2nd NSCS supercomputer located in China is in the top 10. Does that make it "Chinese"?
Re:A 2nd "Chinese".... (Score:4, Interesting)
Re: (Score:2)
Funny every CPU I see on that list is manufactured in china...
Re: (Score:2)
Linux (Score:5, Informative)
Ya for Linux!
Seriously, if this doesn't make every PHB take notice I can't imagine what would. (Hey boss, its free too!)
Re: (Score:1, Insightful)
All our admins and all of our users only know Microsoft systems. Training isn't free.
Re:Linux (Score:5, Insightful)
All our admins and all of our users only know Microsoft systems. Training isn't free.
So your users can't use Linux on the server? Or is it that all the users use super computers on the desktop? Our biz has all MS on the desktop and all Linux on the server. Obviously it is completely seamless. As for the admins, any admin worth their salt is always learning new things to just keep up with technology as it changes. Learning Linux by installing it on one system to start is trivial, and in certain situations, much easier to setup than Windows, such as DNS servers, web servers, etc.
If your admins can only work on a server if it uses a mouse, you need new admins.
Re: (Score:1, Insightful)
If your admins can only work on a server if it uses a mouse, you need new admins.
Agreed. Often times you can't count on morons simply being canned or replaced though. The fact is there's a lot of fools out their that think "system administration" simply means knowing which button to click in the right order. Any understanding beyond that simply doesn't exist, and is lost on them.
This limitation isn't simply one of "GUI vs CLI" or "Windows vs Linux". It's really one of wanting to understand something bey
Re:Linux (Score:4, Insightful)
In my experience Windows admins require *MUCH* more training than Linux admins. There is much more "black magic" that they need to know to be good at their jobs.
A Windows admin needs to know all the secret registry hacks to make things run well. They need to know all the non-intuitive places that Microsoft hides the settings for whatever services need to be configured. They also need to know how to recover things when it all goes horribly wrong.
Most Linux systems have text files to configure things. The files are in a predictable place. Updates are pretty easy and clear.
But Microsoft has scammed people into believing that leaving it harder than just putting up with the same old crap. In this case I just wish that people did get what they pay for...
Re: (Score:3, Informative)
I've done systems administration on both platforms for years and I don't think that there is any real appreciable difference between the amount of knowledge and training needed on one vs. the other when comparing systems that perform similar functions. Compare Active Directory to OpenLDAP+Kerberos 5, for example. They are very, very similar in a lot of ways; so much so, in fact, that OpenLDAP+Kerberos 5 can be used to host the directory portion of a Windows domain.
Re: (Score:2)
From my experience (mainly Linux) it is much easier to setup Linux boxen if you are using them for dedicated tasks, as it is pretty easy to only install the software you need for that task, thus reducing the amount of maintenance in the long run, and narrow down the possible causes of problems. Overall, I would tend to agree that the learning curve is likely equal on both platforms, although finding answers to common Linux issues online is pretty damn easy and fast.
I'm not against Windows on the server per
Re: (Score:3, Interesting)
I'm afraid I'd have to completely disagree with you there. My preferred example isn't Active Directory, but CIFS...
With Windows, you do all the user management, then click-through the painful server setup wizard to turn on file sharing, and everyone's happy and thinks it's oh so easy to do... Right up until some guy tries
Re: (Score:1)
CIFS? Okay, well...
The type of problem you describe can just as easily happen with Samba as it can with Windows Server. (Not sure about the specific problem you're describing, but protocol negotiation p
Re: (Score:2)
Nope. I happen to know the described problem cannot happen with Samba.
Re: (Score:2)
I've never did Windows admin work, but plenty of Solaris and Linux, mostly Linux and some other UNIX OSes over the years. I've also worked as a Windows developer for a while, and I kinda agree that Windows is harder to develop on and I would assume to admin as well.
I've always held high regards for Windows developers. They don't have the internals (source), nor the POSIX standards (yes, I know there are some POSIX stuff for Windows). When I was a Windows developer, I felt like I always had to reinvent th
Re: (Score:2)
I guess you are pretty well !#@%ed, but then again the world still needs ditch diggers. ;)
Re: (Score:2)
So... you don't have internet access? I don't know of any Microsoft routers, switches, firewalls, etc.
And I'd respond to your statement by saying that admins aren't free either. If you're using Windows on your servers, the overwhelming majority of studies say you have a lot more admins than you would need if you switched to some non-Windows server operating system.
Honestly, for the cost of a couple idiot MCSEs, you could
Re: (Score:1)
Re: (Score:1, Insightful)
How is this relevant to the environment most PHBs control? We're talking supercomputers here.. Ferraris.. Lamborghinis... not super reliable diesel trucks. Most PHBs want uptime, not go-fast-real-quick.
welcome to 1995 (Score:3, Informative)
um. you want a Beowulf with that?
Linux has been in the supercomputer lists for decades.
Google is a much better example of how you can use Linux to take over the world; which is what every self respecting middle manager want's to do.
I.e. Shit loads of cheap compute power. Got any tasks which need that?
Re: (Score:2)
A nice round inflation adjusted number would be $1000?
Darl McBride is handed a list of the naughty and very naughty.
By Processor (Score:4, Interesting)
Re: (Score:3, Insightful)
Re: (Score:3, Insightful)
If you're Intel you have more money to spend on marketing, which means "we'll give you a cut rate on a lot of 10000 processors just so we can have the bragging rights."
Re: (Score:1)
It's quite likely that they can offer a hefty discount and still make a profit on the transaction.
Re:By Processor (Score:4, Interesting)
System and component vendors don't make money on these "lighthouse account" supercomputer sales. My experience, having worked in the past for a vendor that did this a lot, is that they're a money-loser. The motivation is bragging rights, though that can be fleeting. I know of several times that my employer declined to bid on a supercomputer deal as it would just be too expensive.
Typically, these systems are actually sold by system vendors (Dell, HP, IBM) and not processor vendors, though the processor vendor will support the bid. That #1 "AMD" system is actually a Cray. Software also plays a large part in success or failure.
Re: (Score:1, Interesting)
What's more interesting is is that the Chinese supercomputer is second overall with only 55680 cores (Intel) and 1.271 peta FLOPS.
That's almost 170000 cores less than the number 1 (AMD), and only 500 tera FLOPS less.
And it's 70000 cores less than the number 3 (IBM) and 200 tera FLOPS faster.
Re:By Processor (Score:4, Informative)
It's especially interesting for two reasons. Firstly, because at that sort of scale interconnect throughput and latency can make a much bigger difference than processor speed. With HyperTransport, AMD has had a huge advantage over Intel here (IBM also uses HyperTransport). It looks like QPI might have eliminated that advantage. Beyond that, you have the supporting circuitry - you don't just plug a few thousand processors into a board and have them work, you need a lot of stuff to make them talk to each other without massive overhead.
The other interesting thing is that the Chinese are using Intel processors at all. I would have expected them to use Loongson 2F chips, or Loongson 3 if they were out in time. I'm not sure if Loongson wasn't up to the job, or if they had some other reason for using a foreign-designed chip.
Re: (Score:2)
The other interesting thing is that the Chinese are using Intel processors at all. I would have expected them to use Loongson 2F chips, or Loongson 3 if they were out in time. I'm not sure if Loongson wasn't up to the job, or if they had some other reason for using a foreign-designed chip.
Loongson has great TDP but isn't all that ballsy. If you're trying to do the job with less cores, it's not in the running. So what if something else takes twice the power? It's the people's money. Same as here.
Re: (Score:3, Informative)
Wikipedia [wikipedia.org] shows the highest performing Loongson system before April scored 1 teraflop peak, and "and about 350 GFLOPS measured by linpack in Hefei". Sounds like they are focusing on performance/watt more than being the fastest, from a read of the rest of the article. Still pretty fast stuff, considering their newest system has 80 quads and is claimed to have a peak around 1 teraflop.
Re: (Score:2)
"and about 350 GFLOPS measured by linpack in Hefei".
Ouch. Most new desktop computers score 35-70gflops, right? That's only ~5-10x faster.
I suppose if it only used 500 watts, it might be worth bragging about - but I can't find any hard power consumption figures.
Re:By Processor (Score:4, Informative)
What's even more interesting is that the nVidia chips that made Nebulae so fast seem to have escaped your notice.
Re: (Score:2)
The flash thingy is really neat: There is a demo [prefuse.org] of this free library [prefuse.org].
How about a direct link... (Score:5, Informative)
How about a direct link to the actual site [top500.org] - or even the actual list? [top500.org]
Re: (Score:2)
Yeah, most "supercomputers" are distributed systems, just like SETI@Home. The only real difference between a traditional supercomputer and a network like SETI@Home is how spread out the nodes are and the amount of bandwith between them.
I just can't stop thinking about a beowulf cluster of those!
Re:SETI@HOME has 3 million or so nodes... (Score:5, Insightful)
Re: (Score:2)
can seti@home run linpack?
Re: (Score:3, Insightful)
If you're gonna open it up like that, Folding@Home would almost certainly take first place.
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=osstats [stanford.edu]
LINPACK (Score:3, Interesting)
I wonder what kind of score these beasts would get on 3DMark ?
Re: (Score:1)
- Steve J
Re: (Score:2)
I guess ACs don't like to name names.
Re: (Score:2)
There are two answers to that.
1. In standard English, it doesn't mean that. Anyone using it in that sense in formal writing is asking for trouble.
2. In many American dialects, that sense has been common for a very long time. A quick Google shows up this paper [jstor.org] from 1975 that says it attracted widespread attention in the 1930s, and it's bound to have been around for many years before that.
Should Say "Top 500 Publicly-Acknowledged Supers" (Score:5, Insightful)
Re: (Score:2)
Re: (Score:2)
Linpack doesn't stress interconnect by that much, however. But yes, there are quite a few systems not on that list.
Re: (Score:1)
Cores (Score:2)
Re: (Score:2)
Food? What food? (Score:5, Interesting)
Re: (Score:1)
Calculating the caloric density of turkey twizzlers?
Re: (Score:2)
Simulations of chemical processes? Estimations of future harvests and researching chemicals used for agriculture? I can't know if that's it, but there you go - some examples where it might be worthwile.
Re: (Score:3, Funny)
Re: (Score:1, Informative)
Because it's delicious, seriously! Don't knock it till you've tried it. It's not conceptually much different from a big sausage, anyway.
Re: (Score:2)
Weather? (Score:2)
Of the UK entries in this list, the first few are Hector (the national supercomputing facility), ECMWF, Universities, financial institutions etc. But there are also some labelled "Food industry". I wonder what I am eating that requires a supercomputer?
Weather simulation, perhaps? Weather has a huge impact on crop yields.
Or perhaps bioinformatics for genetic tinkering.
Re: (Score:1)
On the other hand, I must admit to being curious about what the 'perfect' pizza, matched exactly to me by one of the world's fastest computers, would actually taste like...mmmm...pizza
Re: (Score:2)
I wonder what I am eating that requires a supercomputer?
Doesn't the fast food industry use supercomputers to count the calories of its products, and to annually calculate the number of clogged arteries of its patrons?
Re: (Score:3, Insightful)
missing (Score:1)
Would be interesting to see how the botnets compare
Why do we keep giving China all these advantages? (Score:1, Troll)
Re:Why do we keep giving China all these advantage (Score:1, Insightful)
Do you actually think that everything was and is invented in US? A man that doesn't know the history will lose the future.
Re: (Score:1, Troll)
Thats why I boycott Chinese goods. I don't boycott any other nations stuff, and actually I am better for it. Chinese goods are insanely shoddy. I tend to get much better quality
Largest Pirvately Owned Supercomputer? (Score:3, Interesting)
I was curious if any privately owned(non-corporate or government) machines made the list, and where they placed.
Re: (Score:2)
TOP500 List (Excel) [top500.org]
TOP500 List (XML) [top500.org]
Re: (Score:1)
I was curious if any privately owned(non-corporate or government) machines made the list, and where they placed.
No, botnets are not part of the list.
actual purpose (Score:3, Interesting)
Jaguar -- general research (http://www.nccs.gov/computing-resources/jaguar/)
Roadrunner -- security research (http://www.lanl.gov/)
Kraken XT5 -- general research (National Institute for Computational Sciences/University of Tennessee)
Tianhe-1 -- unstated
Pleiades -- security research (nukes)
"Recently expanded to accommodate growing demand for high-performance systems able to run the most complex nuclear weapons science calculations, BGL now has a peak speed of 596 teraFLOPS. In partnership with IBM, the machine was scaled up from 65,536 to 106,496 nodes in five rows of racks; the 40,960 new nodes have double the memory of those installed in the original machine"
Intrepid -- General research
Ranger -- General research
Red Sky -- General research
It makese me wonder whether the machines for nuclear research went underground or maybe it just doesn't take a top ranking supercomputer to calculate a nuclear explosion anymore.
Re: (Score:2)
perhaps nuke simulations have indeed reached a level where more crunching power isnt worth it anymore, why build a complete new system to do a blast-sim if your existing machine does it in two days. Perhaps there isnt a market for more then X blast simulations per year..
anyway WOW, 40960 NEW nodes... If every BGL node is a single U of rackspace, then even ignoring network/UPS/etc requirements, that means adding 1000 racks, to the already existing ~1500...
Re: (Score:2, Informative)
pleaides isnt nukes, its nasa. airplanes and weather.
the others some are nukes some are open unclassified uses.
noaa/nsf/etc
Re:actual purpose (Score:4, Interesting)
As I understand it most of the nuclear research simulations that it would be nice to run simply cannot be done on any modern machines. If it's only a few particles they can be simulated on a laptop but the interesting interactions need to simulate millions or billions of points with every single one of them influencing every other one in the simulation.
As a simple example, a genetic algorithm was used to program some reconfigurable FPGA chips. A layout was grown on the chip the did the job but broke just about every rule for FPGA design. There were parts of the layout on the chip that were not connected to any circuit but removing them made the device fail to work. Transferring the layout to a different chip got you a non-working circuit. It would be great to be able to simulate this ... not a chance it's too big, by so very many orders of magnitude.
http://www.netscrap.com/netscrap_detail.cfm?scrap_id=73 [netscrap.com]
I don't understand your numbers (Score:2)
Are you counting the entire list of computers or just the top 10? Is the first list supposed to be ones used for security research and the second for general research? If so, Red Sky and possibly others are used for security research.
The change is that most super computers at the national laboratories are not single-use, and are thus listed as general research even if they spend a large proportion of their cycles on security research.
Crysis (Score:1, Redundant)
All this talk of high end computers, and no mention of Crysis?
Not sure how I feel about that.
Treemap (Score:2)
Re: (Score:2)
Re: (Score:2)
Interesting... (Score:2, Funny)
Next bulletin:
"Vista-based benchmark testing complete - converts Jaguar to big pussycat"
And Linux passes the 90% mark (Score:1, Informative)
"Linux family" operating systems went from 89% in the previous list to 91% of this one [top500.org].
Not that the field wasn't already dominated, but it's an interesting milestone. (FWIW, Linux passed 75% in 2006-11, 50% in 2004-06, and 25% in 2003-06.)
Snarkback (Score:2)
This is my chance to snark back at the SPARCophiles at my former employer, Sun. You'll notice that Sun has a respectable presence on this list, lagging just behind SGI. And not a single Sun system on the Top 500 runs SPARC. They're all x64.
The SPARC Uber Alles mentality at Sun in its last days was really frustrating. I was working on x64 systems that were widely considered the best in their class. But you couldn't get the marketing and sales people to make an effort to sell them. They'd march into sales mee
Wow, farms putting NZ on the map ... (Score:2)