 |
|
| View previous topic :: View next topic |
| Author |
Message |
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
|
| Back to top |
|
 |
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
 Posted: Wed Mar 19, 2008 8:18 pm Post subject: |
 |
|
A preview of the next International Supercomputing Conference in June. Note that both the US and Japan are scrambling to build a 10 petaflop system by 2010. If completed, such a system will surpass the combined computing power of all the supercomputers currently on the "top 500 list" (the latter of which has a combined computing power of 7 petaflops).
http://www.supercomp.de/isc08/content/e2/e10/e22/e1671/e2908/index_eng.html
| Quote: | | At ISC’08 we are curious to see whether the first Petaflop/s system will have made it into the TOP500. One of the hot candidates certainly is IBM’s Roadrunner at LANL in Los Alamos, USA. Currently, the number one position is held by BLueGene/L eServer at DOE/NNSA/LLNL, USA, with a best Linpack performance of 478.2 Teraflop/s. We are also interested in how Sun’s consolidation system Ranger at TACC at the University of Austin, USA, is doing. Ranger missed the 30th list in Reno due to delivery problems with the AMD processor chips. Ranger has a peak performance of above 500 Teraflop/s. Can Ranger help Sun Microsystems catch up with the other HPC manufacturers again? |
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
| Posted: Sat Apr 19, 2008 1:48 am Post subject: |
|
|
Intel's Pat Gelsinger's vision of where the supercomputing industry is heading for the foreseeable future:
http://www.heise.de/english/newsticker/news/106017
| Quote: | | In his keynote address at the IDF in Shanghai, he said that the performance of supercomputers would be measured in zettaflops (10 to the 21st power floating-point operations) per second by around 2029. With that power, he said it would be possible to make weather forecasts that would be sufficiently accurate for 14 days. He expects by 2017 it will be possible to create a complete genetic simulation of a cell, which would require an exaflop (10 to the 18th power floating-point operations) per second. |
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
| Posted: Mon Apr 28, 2008 11:11 am Post subject: |
|
|
Intel now definitely getting back into the HPC market in a serious way:
---------------------------------------------------------------------------------
Cray and Intel Collaborate to Develop Future Supercomputing Technologies
Apr 28, 2008 12:00:06 (ET)
SANTA CLARA, Calif. & SEATTLE, Apr 28, 2008 (BUSINESS WIRE) -- Cray Inc. (CRAY, Trade ) and Intel Corporation announced today they signed a multi-year agreement to advance high-performance computing (HPC) on Intel microprocessors while delivering broad new Intel and Cray technologies in future Cray server systems. The collaboration of these two industry leaders will result in HPC systems that will help solve some of the world's most complex scientific, engineering and humanitarian challenges.
"We're excited at the potential of bringing together Intel's powerful silicon expertise and Cray's industry leadership in scalable HPC systems," said Peter Ungaro, president and CEO of Cray. "We pride ourselves in offering the most innovative supercomputing systems and our customers will now enjoy greater choice in processor technologies.
"This collaboration provides the HPC market segment with access to the best microprocessors the industry has to offer at any point in time, in the most advanced supercomputers in the world," Ungaro added. "This further strengthens Cray's industry-leading adaptive supercomputing vision as we move into the Cascade timeframe and beyond."
The two companies plan to explore future supercomputer component designs such as multi-core processing and advanced interconnects. As a result of this collaboration, Cray and Intel plan to develop a range of HPC systems and technologies over the next several years.
"Cray's commitment to Intel is a testament of our commitment to HPC and the strength of our hardware and software roadmap and many-core research," said Patrick Gelsinger, senior vice president and general manager of Intel's Digital Enterprise Group. "Throughout Cray's history, it has been an innovator in high-end HPC while Intel has pushed the boundaries of processor technology.
"The combination of this industry leadership and technical strength will allow HPC users to take advantage of future Xeon and other Intel processor technologies," Gelsinger said. "Together we will enable fundamental and historical problems of science and industry to be solved." |
|
| Back to top |
|
|
probtrader
Senior Poster
Joined: 22 Oct 2005
Posts: 117
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
| Posted: Tue May 06, 2008 11:51 pm Post subject: |
|
|
A new breed of supercomputers - as envisioned by researchers at UC Berkeley:
http://www.tgdaily.com/content/view/37298/113/
| Quote: | The researchers believe that 20 million Tensilica RISC processors would deliver at least 10 PFlops of sustained performance, while topping out at about 200 PFlops. The power consumption of such a system is estimated at about 4 Mega Watts and the construction and typical operation cost at about $75 million. A 200 PFlops system that is built on today’s common architecture could cost up to $1 billion and consume 200 Mega Watts – which is the equivalent of what a city with 100,000 people consumes.
In comparison, the currently fastest supercomputer tops out at 596 TFlops.
.....
According to the researchers, such a computer would make global climate change predictions more understandable and more credible. Climate models are created today largely by using historical data of rainfall, hurricanes, sea surface temperatures and carbon dioxide in the atmosphere. Accurate cloud simulations are much more complex, however, and well within the reach of current supercomputers. Past cloud models, the researcher claim, lack the details that could improve the accuracy of climate predictions: The required accuracy can only be provided by a system that can cope with 1 km-scale models that provide rich details not available in existing models.
To develop such a 1-km cloud model, the scientists said they will need a supercomputer that is 1000 times more powerful than what is available today, the researchers say. And the proposed 200 PFlops Tensilica system could put them into that range, at least in theory. |
|
|
| Back to top |
|
|
probtrader
Senior Poster
Joined: 22 Oct 2005
Posts: 117
|
| Posted: Thu May 08, 2008 3:16 pm Post subject: |
|
|
Simple business model: define performance-per-watt, rather than in absolute terms. Seems to work.
| Quote: | | Super Micro Computer today announced that the company has surpassed the $2 billion milestone in cumulative sales revenue since it was founded in 1993. Supermicro's annual growth rate during the last five fiscal years was several times faster than the server industry growth rate, as the company has consistently posted double-digit growth figures year after year |
.
http://www.supercomputingonline.com/article.php?sid=15572 |
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
| Posted: Mon May 12, 2008 9:12 pm Post subject: |
|
|
NASA set for a major supercomputer upgrade:
http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9084238
| Quote: | Thigpen declined to say what NASA is paying for the supercomputer or for the upgrades that will be needed to get it to 10 PFLOPS.
According to SGI, the system will have more than 20,800GB of memory, which is equal to the memory in average 10,000 desktop PCs. Also, NASA will be deploying a next-generation SGI InfiniteStorage InfiniBand disk solution, which is designed to store and manage 450 terabytes of data, an amount five times bigger than the entire print collection of the Library of Congress. |
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
| Posted: Mon Jun 09, 2008 11:15 am Post subject: |
|
|
IBM, in conjunction with Los Alamos, finally breaks the petaflop barrier:
http://www.nytimes.com/2008/06/09/technology/09petaflops.html?_r=1&hp&oref=slogin
| Quote: | The Roadrunner is based on a radical design that includes 12,960 chips that are an improved version of an I.B.M. Cell microprocessor, a parallel processing chip originally created for Sony’s PlayStation 3 video-game machine. The Sony chips are used as accelerators, or turbochargers, for portions of calculations.
The Roadrunner also includes a smaller number of more conventional Opteron processors, made by Advanced Micro Devices, which are already widely used in corporate servers.
“Roadrunner tells us about what will happen in the next decade,” said Horst Simon, associate laboratory director for computer science at the Lawrence Berkeley National Laboratory. “Technology is coming from the consumer electronics market and the innovation is happening first in terms of cellphones and embedded electronics.”
The innovations flowing from this generation of high-speed computers will most likely result from the way computer scientists manage the complexity of the system’s hardware.
Roadrunner, which consumes roughly three megawatts of power, or about the power required by a large suburban shopping center, requires three separate programming tools because it has three types of processors. Programmers have to figure out how to keep all of the 116,640 processor cores in the machine occupied simultaneously in order for it to run effectively. |
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
| Posted: Sat Jun 14, 2008 1:47 pm Post subject: |
|
|
Supercomputing at the petascale:
http://www.sciencedaily.com/releases/2008/06/080612140031.htm
| Quote: | On Saturday, Los Alamos researchers used PetaVision to model more than a billion visual neurons surpassing the scale of 1 quadrillion computations a second (a petaflop/s). On Monday scientists used PetaVision to reach a new computing performance record of 1.144 petaflop/s. The achievement throws open the door to eventually achieving human-like cognitive performance in electronic computers. PetaVision only requires single precision arithmetic, whereas the official LINPACK code used to officially verify Roadrunner's speed uses double precision arithmetic.
"Roadrunner ushers in a new era for science at Los Alamos National Laboratory," said Terry Wallace, associate director for Science, Technology and Engineering at Los Alamos. "Just a week after formal introduction of the machine to the world, we are already doing computational tasks that existed only in the realm of imagination a year ago."
Based on the results of PetaVision's inaugural trials, Los Alamos researchers believe they can study in real time the entire human visual cortex--arguably a human being's most important sensory apparatus.
The ability to achieve human levels of cognitive performance on a digital computer could lead to important insights and revolutionary technological applications. Such applications include "smart" cameras that can recognize danger or an autopilot system for automobiles that could take over for incapacitated drivers in complex situations such as navigating dense urban traffic. |
|
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
| Posted: Sat Jun 21, 2008 1:50 am Post subject: |
|
|
Stanford's Folding@Home project's current performance surpasses two petaflops. The recent jump (it was consistently under two petaflops just a week ago) is mostly due to the jump in GPUs' participation, which are immensely more powerful than CPUs in folding proteins. More importantly, note that the number of "active GPUs" are only 5,000 - suggesting that this could ramp up very quickly over the next 12 months. In fact, I would not be surprised if Folding@Home surpasses a sustained performance level of over 10 petaflops in 12 months time.
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=osstats |
|
| Back to top |
|
|
HenryTo
Site Admin
Joined: 06 Aug 2004
Posts: 7257
Location: Houston, Texas & Los Angeles, California
|
|
| Back to top |
|
|
Please log in to view without the ad banners |
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
FAQ
Search
Memberlist
Usergroups
Statistics
Register
Profile
Log in to check your private messages
Log in
