backtop


Print 42 comment(s) - last by Performance Fa.. on Jul 2 at 2:17 PM


IBM researcher Shawn Hall inspects a new Blue Gene/P supercomputer (Source: IBM)
IBM's Blue Gene/P triples the performance of its previous supercomputer

IBM has announced Blue Gene/P, the second generation of the world's most powerful supercomputer. Blue Gene/P nearly triples the performance of its predecessor, Blue Gene/L – which also held the title of being the world's fastest computer.

The Blue Gene/P scales to operate continuously at speeds exceeding one petaFLOP – or one-quadrillion operations per second – and can be configured to reach speeds in excess of three petaflops.

The performance jump from Blue Gene/L and Blue Gene/P is due to several factors. In hardware, the Blue Gene/P supercomputer doubles the number of processors per chip, with each processor operating at a higher clock speed. More memory is added along with an SMP mode to support multi-threaded applications. This new SMP mode moves the Blue Gene/P system to a programming environment similar to that found in commercial clusters. The system’s software is also upgraded for Blue Gene/P with refinements to system management, programming environment and applications support.

"Blue Gene/P marks the evolution of the most powerful supercomputing platform the world has ever known," said Dave Turek, vice president of deep computing, IBM. "A new group of commercial users will be able to take advantage of its new, simplified programming environment and unrivaled energy efficiency. We see commercial interest in the Blue Gene supercomputer developing now in energy and finance, for example. This is on course with an adoption cycle – from government labs to leading enterprises – that we've seen before in the high-performance computing market."

Four IBM PowerPC 450 processors running at 850 MHz are integrated on a single Blue Gene/P chip, with each chip capable of 13.6 billion operations per second. A two-foot-by-two-foot board containing 32 of these chips churns out 435 billion operations every second, making it more powerful than a typical, 40-node cluster based on two-core commodity processors. Thirty-two of the compact boards comprise the 6-foot-high racks. Each rack runs at 13.9 trillion operations per second, 1,300 times faster than today's fastest home PC.

The one-petaFLOP Blue Gene/P supercomputer configuration is a 294,912-processor, 72-rack system harnessed to a high-speed, optical network. The Blue Gene/P system can be scaled to an 884,736-processor, 216-rack cluster to achieve three-petaflop performance – though a standard Blue Gene/P supercomputer configuration will house 4,096 processors per rack.

Not only is the Blue Gene/P designed to be blazingly fast, it is also energy efficient. IBM says that the Blue Gene/P supercomputer is at least seven times more energy efficient than any other supercomputer today.

The power of the Blue Gene/P could be applied to the medical field, such as modeling an entire human organ to determine drug interactions, for example. Drug researchers could run simulated clinical trials on 27 million patients in one afternoon using just a sliver of the machine's full power.

Some of the world's leading research laboratories and universities have already placed orders for Blue Gene/P supercomputers. The U.S. Dept. of Energy's Argonne National Laboratory, Argonne, Ill., will deploy the first Blue Gene/P supercomputer in the U.S. beginning later this year.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

:D
By ForumMaster on 6/27/07, Rating: 0
RE: :D
By Kuroyama on 6/27/2007 12:41:54 PM , Rating: 5
pet-a-flop = a dog that flops down when you pet it

So, no, unless your dog is very talented then it probably can't play Doom.

PS. Yes, I have a ridiculous sense of humor.


RE: :D
By DeepBlue1975 on 6/27/07, Rating: -1
RE: :D
By MobileZone on 6/28/2007 2:19:23 AM , Rating: 2
Actually, I was looking for a new machine to use with 3DMax. Does anyone know if they give discount for cash payment?


RE: :D
By xphile on 6/28/2007 3:12:29 AM , Rating: 2
My ridiculous sense of humor has always found it funny that supercomputers are measured by how fast they FLOP. I've generally rated a quick computer on how fast it WORKS - but what the hell do I know :-)


RE: :D
By FITCamaro on 6/27/2007 12:54:52 PM , Rating: 5
It's so fast it can play Doom 5.


RE: :D
By OblivionMage on 6/27/2007 5:01:46 PM , Rating: 1
You don't even need to buy games for this baby. You just type in, no think of a games name and this machine will make the game for you, as your thinking of it.


RE: :D
By UNCjigga on 6/27/2007 5:05:45 PM , Rating: 2
It doesn't care much for Doom, but I bet it likes a mean game of chess against Kasparov now and then.


RE: :D
By rninneman on 6/27/2007 1:06:38 PM , Rating: 5
I haven't heard about whether it can play doom, but rumor is that it can boot Windows in just under 4 minutes! :D


RE: :D
By MADAOO7 on 6/27/2007 1:13:10 PM , Rating: 5
But it stills struggles to use Aero


RE: :D
By Master Kenobi (blog) on 6/27/2007 2:02:42 PM , Rating: 2
Doesnt support Pixel Shader 2.0, not AERO capable :D


RE: :D
By Zurtex on 6/27/2007 3:35:41 PM , Rating: 2
That kind of power could probably emulate graphics card functions, probably :P.


RE: :D
By adamf663 on 6/28/2007 2:07:08 PM , Rating: 2
In 2015 such power will be on the average desktop
In 2020, it won't be fast enough for a desktop refresh for the current version of microsoft's operating system.


Great computer is it capable of good AI
By MarcusJ on 6/27/2007 1:04:01 PM , Rating: 2
After reading this post I'm wondering if this new supercomputer can do basic or middle of the road AI. If so It probably has areadly made a Doom game. Not just have it run on it's desktop. Besides the game reference this is a major leap in computing. The trillion commands the computer can process is amazing.

Now, some people might think that this computer is just an upgrade but its more! I'm waiting when we finally get a computer with good AI. The power to think on it own and follow complex commands.

This will help the human race greatly and pull us closer to computer acting on controlling themselves. Maybe that idea is far away but I believe we are only 10 years away from making "super thinking computers" not just "super computing servers" in which we have now.




RE: Great computer is it capable of good AI
By Hare on 6/27/2007 1:59:28 PM , Rating: 5
I don't think the problem with AI is the processing power. Really advanced AI is extremely hard to code (software).


By Master Kenobi (blog) on 6/27/2007 2:04:16 PM , Rating: 3
Yeap. We haven't figured out how to properly code a real living AI, computing power shouldnt be an issue by the time we eventually do figure it out.


By MarcusJ on 6/27/2007 4:14:49 PM , Rating: 2
I agree but disagree. Your right with software but whats also needed is a great amount of processing power to keep up with the language being written. Even if your able to code AI you need what ever sort computer your working with to respond and excute the command. That comes with having more processing power and the ability of the AI to finish a task given when task without taking a great amount of time.


RE: Great computer is it capable of good AI
By masher2 (blog) on 6/27/2007 11:23:47 PM , Rating: 5
Until we have a better idea of how "I" (intelligence) works, we're never going to be able to code a real AI. The mind and how it works is still a mystery to researchers.


By Adonlude on 6/27/2007 4:07:07 PM , Rating: 3
You should probably read about Technological Singularity here: http://en.wikipedia.org/wiki/Tech_singularity


By CollegeTechGuy on 6/27/2007 4:20:17 PM , Rating: 1
The problem with AI like stated, is not processing, but rather software. We've had many discussions at my College about AI, the ideas behind...and believe it or not, but what would actually be considered AI. Many computer geeks have very different definitions of what an AI is. My personal opinion on what computers can't do that we can, is learn and apply to different applications. Something like lets say a game AI is programmed to actually learn from how a player plays. Then adept its soldiers to act accordingly to how that person plays. Yes, the computer can "record" all the data on how a player reacts to certain situations, but it can only adapt to those situations based on the programmers input.

Basically the computer can't think for itself, it only does what it is told. I belive we can program the learning part of computers, because thats just storing data, but actually getting the computer to react to different situations and "think" for itself is a whole different ball game.

Another arguement that i've had with fellow students and Professors about AI is initial programming. True AI wouldn't have any some think, because they say we don't know anything when we are born. I disagree though, I think that we are born knowing how to do certain things. This is somehow "programmed" in our DNA.

Then you start thinking about the DNA programming and you begin to wonder, or at least I do, about Human Beings. How much different are we from actual computers, besides the fact we have a good AI program. I mean we store data, process it in our brain, and react upon new stuff based off of stored data. And technology is just beginning to use Organic materials for data and imaging(OLEDs). Although I don't believe in religion personally, perhaps there was a God like figure who did create us just like we are trying to create AI and incorporate organic material into computers. Who knows, maybe we created ourselves created Adam and Eve and then sent them back in time.

Alright, i'll stop my ramblings of a college student with too much to think about :P


By Lazlo Panaflex on 6/28/2007 1:40:12 PM , Rating: 2
quote:
This will help the human race greatly and pull us closer to computer acting on controlling themselves.


Nah, Skynet & the Terminators will take care of that! Ahhhhnold! ;-P


By Drexial on 6/28/2007 3:55:48 PM , Rating: 2
wait till it discovers fire, then well talk about it programming Doom games. considering its a computer, would it be fighting off zombie servers and PC possessed by some unseen force (aka people) that are trying to shut it down as an intelligent entity?


AI
By vladio on 6/27/2007 4:43:24 PM , Rating: 2
AI -- Not needed FAST computer, needed very SMART computer(period)




RE: AI
By DeepBlue1975 on 6/27/2007 6:49:14 PM , Rating: 2
But hey! We already have smart phones!
Ain't it the same?


RE: AI
By MobileZone on 6/28/2007 2:18:15 AM , Rating: 2
Some have a higher IQ then others... depends on the brand. Apple for example is the less smart one, cos it only cares of appearance.


RE: AI
By theapparition on 6/28/2007 8:20:40 AM , Rating: 2
So Apples are the playmate blondes of the industry. Fun to play with on the side, but not what you want to move forward with.

I like the analogy!


RE: AI
By GlassHouse69 on 6/29/2007 12:36:32 AM , Rating: 2
pwnd


Curious
By KristopherKubicki (blog) on 6/27/2007 12:47:29 PM , Rating: 2
I'm curious how this relatively "off-the-shelf" solution stacks up against, say the billion-dollar LANL supercomputer:
http://www.dailytech.com/article.aspx?newsid=2243




RE: Curious
By KristopherKubicki (blog) on 6/27/2007 1:07:07 PM , Rating: 4
I dont like replying to myself, but I just found out the Top500 was published recently:

http://www.top500.org/list/2007/06/100


RE: Curious
By Viditor on 6/28/2007 2:53:35 AM , Rating: 2
The list doesn't include Ranger (due in October).
http://news.yahoo.com/s/nf/20070626/tc_nf/53340

quote:
When Ranger is up and running at the Texas Advanced Computing Center in Austin, and joins with fellow supercomputers on the TeraGrid national network in late 2007, it is expected to deliver a peak performance of more than 500 teraflops. Constellation environments can eventually be configured to provide as much as 1.7 petaflops...
When it is complete, Ranger will have 1.7 petabytes of storage capacity using Sun Fire X4500 data servers, and over 15,000 quad-core microprocessors connected by a Sun InfiniBand switch


In essence, the tallies will be:
IBM: 294,912 processers, 72 racks, 1 Peta-flop
Sun: 15,000 processers, ? racks, 500 Tera-flops

Also, the Ranger system will use off-the-shelf components so it can be broken apart or built up


just wondering
By Xajel on 6/28/2007 2:50:34 AM , Rating: 2
Why IBM do't use Cell CPU's in there Supercomputer things ??

I know too much diff. for example, the CPU's used in current BlueGene clocked at les than 1GHz, while Cell is 3.2GHz beast, this will need very hard cooling thing, but it will rock the Peta flop for more than 10 or even more ( I didn't caclulated it, will someone make some math here :D )




RE: just wondering
By lompocus on 6/28/2007 3:44:42 AM , Rating: 1
Amazingly cell is scalable up to 6GHz :D.

I am curious as to how they cool the things, I mean there's only so much air conditioning the building in which the supercomputer is housed can have, right?


RE: just wondering
By Hoser McMoose on 6/28/2007 4:51:26 PM , Rating: 2
The SPEs in the Cell only do single precision calculations, which is often pretty much useless for this sort of device. If you look at it's double-precision performance the Cell is pretty ho-hum since it can only do that in the PowerPC unit. So it's score in terms of Petaflops would be rather abysmal (worse then an equivalent number of Opteron or Xeon chips). IBM has done some work on an HPC design using Cell processors, but from what I've seen there isn't too much interest.

IBM could take steps to correct the single-precision-only issue as well as modifying the chip somewhat to make it more useful for HPC workstations, but in all likelihood they would end up with something that looks an awful lot like their current BlueGene design.


RE: just wondering
By Schmeh on 6/29/2007 1:15:17 AM , Rating: 2
This is not at all true. The SPEs are fully capable of doing both single precision and double precision floating point operations. The only caveat is that there is a 7 cycle latency for DP on the SPEs.

From RealWorldTech.com:

quote:
A quick glance at the microarchitecture of the CELL processor reveals that the SPE’s are capable of performing 4 (non IEEE754 compliant) SP floating point multiple-add (FMADD) operations per cycle or 2 (IEEE754) DP FMADD operations every 7 cycles. Consequently, the 8 SPE’s alone can achieve the 256 SP GFlops rating at 4 GHz without the aid of the PPE. Presumably, the (DD2) PPE can also produce 4 SP FMADD’s per cycle, and the (DD2) CELL processor should instead be rated as 288 Gflops at 4 GHz when the compute power of the PPE are taken into consideration. Similarly, the 2 DP floating point multiply-add operations every 7 cycles results in 18.3 DP GFlops per second for the 8 SPE’s at 4 GHz, and the PPE can sustain a peak throughput of 1 DP FMADD operation per cycle, producing 8 DP GFlops at 4 GHz. The total of 26.3 GFlops matches nicely with IBM’s claim of > 26 DP GFlops.


http://www.realworldtech.com/page.cfm?ArticleID=RW...


PetaAI
By Brassbullet on 6/28/2007 1:57:31 PM , Rating: 2
Sweet news, I live about a 1/2 mile from Argonne, I wonder if there is any change the guards will let me watch them build it?

In other news, to all those people talking about AI, I believe the key is still in processing power. The software for AI must be well written to mimic basic human instinct, but the amount of senses in an animal would require a huge amount of digital processing power to simulate, and at this point is more out of reach then programming certain reactions for sensations.




RE: PetaAI
By Performance Fanboi on 7/2/2007 2:17:10 PM , Rating: 2
True AI does not 'mimic' instinct - it must posess it. Creating AI will, as others have stated, required understanding and defining intelligence and we haven't even scratched the surface of how our brains work.


Q
By jay401 on 6/27/2007 2:16:10 PM , Rating: 3
quote:
The performance jump from Blue Gene/L and Blue Gene/P is due to several factors.


You mean the performance jump from /L to /P, right? Or do you mean to imply there a third computer that those two are jumping from?




RE: :)
By kaosrise on 6/27/2007 2:37:40 PM , Rating: 2
Wow this is just great, but, does it have YouTube integrated?

*waiting for the iGene*

teehee




A quadrillion ops per second
By microchip on 6/27/2007 4:40:07 PM , Rating: 2
How many problems remains to be solved. To decode our DNA.
Or to calculate from the big bang how many chances we have in order to be created and to exist. I think 0.000000000000 infinite%.




Clarification needed.
By crystal clear on 6/28/2007 3:31:05 AM , Rating: 2
quote:
The Blue Gene/P scales to operate continuously at speeds exceeding one petaFLOP – or one-quadrillion operations per second – and can be configured to reach speeds in excess of three petaflops.


But not so as per-

EDIT: Our original post stated that the Blue Gene /P had already broken the 1 petaflop record. After carefully reviewing IBM's statements, it appears the system is projected to break the record, but has not actually done so.

http://arstechnica.com/news.ars/post/20070626-ibm-...




Not Big Deal
By smartpuppy on 6/28/2007 4:04:00 PM , Rating: 2
When will everyone understand this is total waste of money. Not even in the most dreaming will work for every problem of most people. Could total of execution equal be more than the sum? What else could it mean? All answers will not be coming I think. Is there more to say?




"Well, there may be a reason why they call them 'Mac' trucks! Windows machines will not be trucks." -- Microsoft CEO Steve Ballmer

Related Articles
IBM's Next-Generation Supercomputer
February 28, 2007, 1:52 AM













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki