According to Bitcoin Watch, the entire Bitcoin network has reached one exaFLOPS, which may not mean anything to most people from the face of it, but compare that statistic to the world’s most powerful supercomputers and it’s quite jaw-dropping. One exaFLOP is over eight times faster than the combined speed of the top 500 supercomputers in the world.

FLOPS is a term that you may have seen most in graphics card specs, and it stands for “Floating-point Operations Per Second,” or how many math problems it can do in one second. An exaFLOP can calculate one quintillion math problems per second. Compare that to the most powerful supercomputer in the world currently (IBM’s Sequoia, pictured below), which only does a measly 16.3 petaFLOPS (0.0163 exaFLOPS), and you really have something.

However, some people may be quick to point out that Bitcoin mining technically doesn’t operate using FLOPS, but rather integer calculations, so the figures are converted to FLOPS for a conversion that most people can understand more. Since the conversion process is a bit weird, it’s led to some experts calling foul on the mining figures. In any case, though, the conversion comes out to one Bitcoin mining hash equaling 12.7 thousand FLOPS.

What is Bitcoin anyway? It simplest of terms, it’s a currency that only lives online, and it’s been quite controversial ever since its inception. There’s no central bank to monitor the currency, and it uses peer-to-peer networking for transactions, with no middleman to pass along the currency (like a bank). Furthermore, there’s no serial number attached to Bitcoins, so they’re untraceable, making it a good medium for illegal transaction, but also a good thing for privacy advocates. In the end, Bitcoin is still very young, and there are very few places that accept it, but PayPal is thinking about hopping on board, so it could be a matter of just a couple years before we see it take off, but it’ll most likely be a fight in order for Bitcoin to gain ground.

VIA: Gizmodo

SOURCE: Bitcoin Watch

Big Red II is a next-generation Cray XK supercomputer that operates on GPU-enabled nodes and standard CPU compute nodes. There are 344 CPU nodes which each use two 16-core AMD Abu Dhabi processors. It has 676 GPU nodes each using one 16-core AMD Interlagos and one NVIDIA Kepler K20. The supercomputer has 21,824 processor cores total, 43,648GB of RAM, and 180TB of storage.

Indiana University says that the new supercomputer will be used to “enable vital new research to be done and breakthroughs in fields” such as medicine, physics, fine arts, global climate research, astronomy, and much more. The supercomputer should also help the university attract big research grants and boost Indiana’s economy. Big Red I helped IU accumulate $253 million in grant funding, so Big Red II is expected to bring in much more.

What Indiana University means by Big Red II being a dedicated university resource is that it will only be used by the university, for the university. It will be used to benefit the university in many academic areas. It will also help fund the entire state, and will be “without any constraints from an outside funding agency.” Many professors, staff members and researchers are excited to use Big Red II. D. Craig Brater, the Dean of Indiana University’s School of Medicine, says,

“Having been involved in the evolution of IU’s advanced computing environment since 2000, I have seen how advanced computing has become more critical to medical research and innovation, and watched as the IU computational resources have been deployed in ways that are more and more valuable to IU medical research. Big Red II will be a critical and strategic aid to accelerating new medical breakthroughs and enabling research that will improve human health.”

[via Ars Technica]

The computer was installed at the Los Alamos National Laboratory where it has been in use for the last five years. Yesterday, the laboratory officially decommissioned Roadrunner. The supercomputer has 12,960 IBM PowerXCell 8i processors and 6480 AMD Opteron dual-core processors.

Those processors shared 114 TB of memory and about 1.09 million TB of storage. The supercomputer isn’t being completely dismantled, researchers will continue to utilize the machine and its impressive power for various experiments. These experiments will include things such as determining the methods for compressing operating system memory and optimizing data routing.

With Roadrunner being decommissioned from research duties, scientists and other researchers can now use the computer for projects that couldn’t have been done while the supercomputer was being used for research projects. The computer is housed in 6000 ft.² of space and cost $125 million to build. Roadrunner may not be fast enough for the scientists and researchers at Los Alamos, but the computer is still incredibly fast and sits at number 22 on the list of the world’s most powerful supercomputers. The computer gulps power needing 2345 kW when running full tilt. Modern supercomputers need significantly less power to achieve significantly more performance.

[via PCMag]

Santa Clara University researcher Ed Karrels ended up breaking the world record for computing digits of Pi to eight quadrillion places to the right of the decimal point. Karrels used graphics cards to do the work rather than CPUs, and he spread the work across three different computers: one with four NVIDIA GTX 690 cards, one with two NVIDIA GTX 680 cards, and 24 computers at the Santa Clara University Design Center with one NVIDIA GTX 570 card each.

The calculation took 35 days to complete, from December 19 to January 22, beating out the previous held by a team at Yahoo, who used 1,000 CPU-only computers, which took 23 days to compute Pi to two-quadrillion places, just a quarter of what Karrels’s setup achieved. After the 35-day run, Karrels conducted a second run to double-check the math, which took just 26 days using newer versions of his programming tools.

Karrels will speak at the GPU Technology Conference in San Jose, California next Tuesday, where he’ll be explaining the math behind the Pi calculation achievement, as well as the programming tricks he used, as well as the logistics of conducting supercomputing tasks on a budget.

NVIDIA says that the new GTX Titan graphics card is “powered by the fastest GPU on the planet,” which we certainly can’t refute at this point. The graphics card itself is huge, measuring in at 10.5-inch long, and it’s capable of pushing 4,500 Gigaflops, which is quite impressive if we do say so ourselves.

However, the GTX Titan falls just a tad short of NVIDIA’s current top-tier offering, the GTX 690, as far as raw specs and computing power are concerned, but efficiency is where the Titan really shines. The GTX Titan features over a thousand more CUDA cores than the GTX 690, but it requires less power, as well as generates less heat and runs quieter overall.

As far as availability goes, NVIDIA The Titan GPU will be available starting on February 25 from various partners, including ASUS, eVGA, Gigabyte, and MSI, at a price of around $1,000, which certainly isn’t going to want you to make an impulse purchase, but if you’re looking for supercomputer-like speeds with your gaming rig, this card may be well worth it.

In a NVIDIA Blog post, the company explains that Eurora managed to reach “3,150 megaflops per watt of sustained performance,” which just so happens to beat the efficiency of the highest-ranked supercomputer on the Green500 list by 26%. That isn’t bad at all, and it’s thanks to Eurotech’s Aurora hot water cooling system.

By using hot water cooling, the Eurora doesn’t need to be cooled by air conditioning, obviously saving on energy costs. NVIDIA describes the benefit of using hot water by pointing out that it can be “re-purposed to heat buildings or drive absorption chillers, and then returned back to the supercomputer at a cooler temperature.” The Eurora is equipped with 64 compute nodes, which are each made up of two Intel Xeon E5-series CPUs and two NVIDIA Tesla K20 GPU accelerators. Since the water cooling system allows Eurotech to save space in the Eurora, NVIDIA says it’s able to fit 256 CPUs and GPUs into a single rack.

While the water cooling system helped the Eurora meet its efficiency goals, it didn’t do all the work along. The Telsa K20 GPUs are also quite efficient too, with NVIDIA pointing out that they’re four times more efficient than x86 CPUs. The Eurora made its way to the Cineca supercomputer facility in Bologna, Italy this week, with NVIDIA and Eurotech predicting that it will save 2.5 million kilowatt hours and eliminate 1,500 tons of CO2 emissions over the course of its 5-year life. That, ladies and gentlemen, is one efficient machine.

Rensselaer will receive the Watson system thanks to a grant that allows the university to invest more resources to research and development of big data, analytics, and cognitive computing. However, in return, IBM is asking the university to send its findings their way so that they can improve Watson even more.

Rensselaer’s private Watson supercomputer will have 15 terabytes of storage, which is actually more than even the Jeopardy! version had. Plus, the room that Watson will be stored in will allow 20 people at a time to work inside, including faculty, graduate students, and a few undergraduate students.

So what will the supercomputer be used for at the university? Artificial intelligence researchers at Rensselaer want to improve Watson’s mathematical ability and help it figure out the meaning of newer words. They also want to improve the computer’s ability to analyze all of the images, videos, and emails floating around on the internet, something that will prove to be no easy task for the folks at the university.

Transcendence will be put together by the team behind the incredibly-popular film Inception, which starred Leonardo DiCaprio. Christopher Nolan will be the executive producer, and his long-time friend and partner-in-crime, Oscar-winner Wally Pfister, will be directing the new movie, which will be his first feature film. Jack Paglen signed on to write the screenplay.

According to TheWrap, who talked with the movie’s producer, Andrew Kosove, Transcendence will essentially be about a scientist whose brain is uploaded into a supercomputer with the aim of creating the world’s first machine that can think for itself. Kosove also confirmed that singularity, nanotechnology, and artificial intelligence concepts are important factors in the movie.

The script mentions three scientists: Max, and a husband-and-wife couple named Will and Evelyn. Johnny Depp will play Will, while the other two roles have yet to be cast. In the script, the trio of scientists have been developing a programming code for the world’s first fully self-aware computer. Eventually, an anti-technology terrorists assassinate Will, and Evelyn uploads his brain into the prototype computer that the team has been working on.

Kosove says that there will most likely be a lot of script changes from now until the start of filming, and it’s not yet said when the movie will release, but IMDb lists a 2014 release date. So far, all other details have been layered in secrecy, but just like Kosove said, there will most likely be a lot of changes to the movie, since the production is still in its early stages.

[via TheWrap]

Image via Flickr

It’s a big day for NVIDIA with the Tesla K20 architecture being reintroduced in its final form powered by CUDA – also known as “the world’s most pervasive parallel programming model.” NVIDIA backs this claim up with 8,000 institutions with CUDA developers, 1,500,000 CUDA downloads, and a massive 395,000,000 GPUs shipped with CUDA built in. With 629 university courses being taught on CUDA across 62 countries, it’s safe to say that it’s here for some time to come.

The K20 family also makes with the undeniably next-level powerful performance on scientific applications – this being exactly why the Titan supercomputer uses the architecture for the massive bulk of its processes. The 2011 Gordon Bell Winner for computational simulation was 3.1 Petaflops (3.08 Petaflops on K Computer) with NVIDIA’s new effort bringing on 10+ Petaflops here in 2012.

Both the Tesla K20 and the Tesla K20X work with a single GK110 Kepler GPU with your favorite features – Dynamic Parallelism and Hyper-Q! These units have more than one teraflop peak double precision performance and deliver 10 times the performance of a single CPU – this claim by NVIDIA being based on the following: “Ws-lsMs performance comparison between single E5-2687W @ 3.10GHz vs single Tesla K20X. Tesla K20X > 650 gigaflops.”

There’s also a Tesla K10 model out there, you should know, with memory size of 8GB per board and just SMX inside instead of the addition of Dynamic Parallelism and Hyper-Q, which the K20 and K20X have. The K10 (again, having been on the market now for some time,) has a peak double precision floating point performance of 0.19 teraflops and is made for servers only – it’s peak single precision floating point performance, on the other hand, is 4.58 teraflops. The K20 rings in 1.17 teraflops and 3.52 teraflops for double and single precision floating point performance respectively. The K20X nabs 1.31 teraflops and 3.95 teraflops.

The K20 has 5GB memory size per board while the K20X has 6GB, and both devices have just the one GK110 GPU while the K10 has two GK104 units inside. The K20 units are made for massive beastly tasks like financial computing, computational chemistry and physics, and satellite imaging. The K10 on the other hand is made for seismic, image, signal processing, and video analytics.

The NVIDIA Tesla K20 family of GPU accelerators is ready for action this week – shipping today and available for order from your favorite computer store. NVIDIA is working with Appro, ASUS, Cray, Eurotech, Fujitsu, HP, IBM, Quanta Computer, SGI, Supermicro, T-Platforms, Tyan, and NVIDIA reseller partners as well – you’ll have no shortage of choices on your hands. Grab a K20 as fast as you can!

The performance record that this beast now holds is a 17.59 petaflop mark as measured by the benchmark system known as Linpack – a system that measures all manner of devices all the way down to smartphones (with GPUs packed inside as well.) The Titan makes this massive stride into the future with the Tesla K20X accelerator, the “flagship of NVIDIA’s accelerated computing line.” NVIDIA notes that this new solution provides “the highest computing performance ever available in a single processor.”

NVIDIA’s claims are backed with two more benchmark results: 3.95 teraflops singleprecision and 1.31 teraflops double-precision peak floating point performance – beastly. Those come from a setup as follows: CPU results: Dual socket E5-2687w, 3.10 GHz, GPU results: Dual socket E5-2687w + 2 Tesla K20X GPUs. NVIDIA also notes that the family of processors being used here also includes the K20 (without the X) which has busted out 3.52 teraflops of single precision and 1.17 teraflops of double-precision peak performance.

The Tesla K20X and K20 GPU accelerators have brought on more than 30 petaflops of performance over the past 30 days – that’s big. It’s so big, in fact, that it’s equivalent to the computational performance of the top 10 fastest supercomputers from 2011 combined.

In addition to being the fastest, the Tesla K20X GPU accelerator has been revealed to be three times more energy efficient than previous generation GPU accelerators – so says NVIDIA. The Titan has achieved 2,142.77 megaflops of performance per watt, this surpassing the previous most energy-efficient supercomputer on the planet as well – this according to the official Green500 list.

Have a peek at the timeline below to get more information on Titan as well as the K20 family of GPUs from NVIDIA – it’s big time computing action for all!

This project is a next-level teaming of the Cray XK7, the “most scalable supercomputer” on the planet, and the NVIDIA Tesla K20 GPU, aka the “world’s fastest accelerator.” This combination beings on CUDA and Open ACC programming and new features that expand programmability far beyond what’s been available before, and with the NVIDIA GPU units being used now, they’re working with 3x higher performance per watt. This means one whole heck of a lot less power consumed for the same tasks as were being performed before.

This supercomputer is currently in the acceptance process for a series of scientific applications. The program that surrounds this unit is made to expand the access groups have to supercomputing, judging each application for a program individually and giving them time based on the percentages allotted to each of the following: Plasma, Nuclear, Materials, Engineering, Earth Science, Computer Science, Chemical Science, Biology, and Astrophysics. This is all done through the US Department of Energy’s INCITE: Innovative and Novel Computational Impact on Theory and Experiment.

This program has seen a record number of proposals, with demand being approximately three times larger than they’re actually able to supply. Have a peek at the gallery below to see a few examples of what these applicants are proposing:

This also brings the Jaguar – again, now called Titan – up to a whole new specifications set. The compute notes remain the same at 18,688, but the Login and I/O nodes go up from 256 all the way to 512. The memory per node was at 16Gb and is now at 32GB + 6GB. Number of Opteron cores jumps from 224,256 to 299,008, and the total system memory was at 300TB and is now at 710TB. With the addition of 18,688 NVIDIA K20 Kepler accelerators, this beast’s former peak performance at 2.3 Petaflops is dwarfed by its current peak at 20+ Petaflops.

That research will include simulating physical systems, such as weather patterns, or progressions in energy, climate change, efficient engines, materials, and other fields. However, unlike most supercomputers, where access is jealously guarded, Titan takes a more open approach to access.

Researchers from schools and universities, government labs, and private industry can access Titan – by arrangement, of course – to crunch their own data. Final testing is still underway by the laboratory and Cray, and the supercomputer’s first year will be dominated by work on the Department of Energy’s Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program.

“The improvements in simulation fidelity will accelerate progress in a wide range of research areas such as alternative energy and energy efficiency, the identification and development of novel and useful materials and the opportunity for more advanced climate projections” James Hack, director of ORNL’s National Center for Computational Sciences, said of the new machine.

In total, there are 299,008 CPU cores, sixteen to each of 18,688 nodes; each node also has an NVIDIA Tesla K20 graphics accelerator. The cores are used to guide the simulations, while the GPUs are relied upon to do the actual data crunching; altogether, it’s more than 10x faster and 5x more power efficient than the Jaguar supercomputer Titan replaces.

In fact, Titan can simulate 1-5 years per day of computing time, whereas Jaguar took a day to work through around three months worth of data. ORNL says it’s the equivalent of “the world’s 7 billion people being able to carry out 3 million calculations per second.”

Xeon Phi uses the same  22nm, 3-D tri-gate transistors as Ivy Bridge, and comes as a PCIe card with more than 50 cores and at least 8GB of GDDR5 memory. It has 512b wide SIMD support – allowing for multiple elements to be worked on in a single instruction – and Intel claims it’s good for more than 1 TeraFLOPs per node.

As well as working with x86 programming models, the Xeon Phi coprocessor will also show up as “an HPC-optimized, highly-parallel, separate compute node that runs its own Linux-based operating system independent of the host OS.” Intel says that in doing so it wil support more flexible cluster-solution, such as those that might be incompatible with existing GPU-accelerated systems.

Xeon Phi will first show up in Stampede, a 10 Petaflop HPC Linux cluster expected to be operational by the start of 2013. Stampede, a project by the Texas Advanced Computing Center (TACC) with the University of Texas at Austin, will pair Xeon E5 CPUs – contributing around two Teraflops – with the remaining eight Teraflops delivered by Xeon Phi coprocessors.

Intel says 44 companies have already signed up to use Xeon Phi, including Bull, Cray, Dell, HP, IBM, Inspur, SGI and NEC. It has already promised to hit Exascale performance by 2018.

As well as running faster, Sequoia also consumes less power than the Fujitsu. The IBM system demands 7.9 megawatts in use, more than a third less than the K Computer.

“Sequoia will provide a more complete understanding of weapons performance, notably hydrodynamics and properties of materials at extreme pressures and temperatures” Bob Meisner, NNSA director of the ASC program said of the new supercomputer. ”In particular, the system will enable suites of highly resolved uncertainty quantification calculations to support the effort to extend the life of aging weapons systems; what we call a life extension program (LEP).”

The system’s huge crunching power will allow it to quickly sift through incredible numbers of different models, trying to figure out which is most stable and attempt to cut out some of the uncertainty involved in nuclear devices.

For the most part water-cooled, Sequoia’s 96 racks consist of 98,304 compute nodes, 1.6 million cores and 1.6 petabytes of memory. A US company last held the top spot in supercomputing performance two years ago, though was beaten by a Chinese rival.

The supercomputer pictured here is the Cray Jaguar and it is able to perform 1 million billion operations per second. The gigantic supercomputer takes up 5000 square feet and has been in use at the Oak Ridge National Laboratory for years. Supercomputers by the end of the decade will be as powerful as 50 million laptop computers. Scientists and researchers say these exaflop supercomputers will usher in a new era of scientific discovery.

Exaflop supercomputers will be 1000 times faster than the fastest supercomputers today and will be able to perform 1 billion billion operations every second. The first machines capable of this sort of performance are expected to go online around 2020. The fastest supercomputer today is the Japan K computer and it is capable of 10 petaflops of performance. The machine uses 88,128 computer processors. Future exaflop machines are expected to have somewhere between 1 million and 100 million processors.

[via CNN]

An ExaFLOP/s capable computer – a quintillion computer operations per second – would be a hundred times faster than current supercomputers, Intel points out, and demand each possible source of latency and delay to be ironed out. To do this, rather than using switched networks as in ethernet connections, InfiniBand relies on switched fabric network topology, spreading traffic over a number of links simultaneously. There’s more on InfiniBand tech at Wikipedia.

QLogic says that shedding its InfiniBand business will allow it to concentrate on its other networking and interconnect specialisms, including “converged networking, enterprise Ethernet, and storage area networking products.” Intel has extended job offers to “a significant number” of QLogic employees from the acquired division.

The latest chip from Intel is a 1-teraflop beast that has over 50-cores. Intel is testing the processor inside a computer at the SC11 supercomputer conference in Seattle. The little chip that is pushing out so much performance is the size of a matchbook and called Knights Ferry. The architecture of the chip is called many integrated core or MIC.

To put a bit of perspective on the teraflop number, that means that the chip is capable of a trillion floating-point operations per second. The photo above is of a slide from the presentation that sums things up nicely. The Knights Bridge chip needs a single PCIe slot to hit 1TF. Back in 1997 to hit 1TF performance took 9,298 Pentium II Xeon processors and 72 cabinets. How’s that for progress?

[via Seattle Times]

The slightly more specific goal of the Blue Waters project is to deploy a supercomputer that’s capable of working on a “diverse range of real-world science and engineering applications” at a sustained performance of one petaflop. Scientists and engineers across the USA will benefit from the eventual products of this initiative, and the whole project is supported by the National Science Foundation at the University of Illinois. NVIDIA chief technology officer of Tesla GPUs Steve Scott had the following to say on the situation:

NCSA has seized this opportunity to make Blue Waters into an even more amazing scientific computing instrument than originally planned. The performance and wide access of Blue Waters will enable the scientific community to accelerate the race for better science.” – Scott

Cray system’s general purpose CPUs will work side by side with NVIDIA’s Tesla-brand GPUs and will accelerate compute-intensive applications across the board. This newest Blue Waters system will be an ultra-powerful hybrid supercomputer consisting of more than 235 Cray XE6 cabinets. In addition, there will be more than 30 cabinets of a future version of what’s recently been announced as the Crazy XK6 supercomputer, these containing the next-generation Tesla GPUs again based on the “Kepler” architecture. Can’t wait to grab a setup for my underground beat laboratory!

This project is known as the EU Mont-Blanc Project and will otherwise take part in developing a portfolio of exascale applications that work well on the technology now being developed with NVIDIA. Sound familiar? The leader of this Mont-Blanc project mister Alex Remirez, notes that while current systems need to give the “lion’s share” of energy to the CPUs, 40 percent or more, that is, “the Mont-Blanc architecture will rely on energy-efficient compute accelerators and ARM processors used in embedded and mobile devices to achieve a four- to 10-times increase in energy-efficiency by 2014.” Exciting!

NVIDIA expanded upon what it’d do to work with developers by noting that they’d be releasing a new hardware and software development kit for its ARM-based initiatives around the globe. This kit will of course contain a brand new quad-core NVIDIA Tegra 3 ARM CPU complete with a “discrete” NVIDIA GPU, all of this available inside the first half of 2012. This kit’s hardware is being developed by SECO and will be supported by the NVIDIA CUDA parallel programming toolkit. Get in on the action!

The machine can be scaled up to a 1,024 rack configuration to reach the 23.2 petaflop performance level. Fujitsu is aiming the machine at things like drug development, disaster prevention and mitigation and other research that demands high performance from the computer. The processors inside the machine are SPARC64 IXfx units.

Each of the processors has 16 cores and uses water cooling to maximize the performance per watt with standalone performance levels of 236.5 gigaflops and a performance per watt of over 2 gigaflops. Each node of the supercomputer can pack in 32GB or 64GB of RAM with memory bandwidth of 85GB/s. The supercomputers will start shipping in January with a goal of 50 systems sold over the next three years.

The new supercomputer is installed at the National Supercomputer Center in Jinan. According to reports, the machine is capable of about 1,000 trillion calculations per second landing it in the top 20 fastest globally. Inside the machine, it has 8,700 ShenWei SW1600 microprocessors. The entire machine was designed and built in China.

The theoretical performance of the BlueLight would put it at about 74% of the computing power of the fastest supercomputer in the US. That fastest American machine is the Jaguar supercomputer that is the third most powerful supercomputer in the world. Jaguar is in use at the Department of Energy facility at the Oak Ridge National Laboratory and was built by Cray.

[via Hot Hardware]

Technology pundits such as Rick Doherty (of Envisioneering Group) are thus far seeming to focus on the fact that IBM will be returning the money they’ve recieved from the project up to this point while the NCSA will be returning all IBM equipment, while both groups have promised to work on future petascale computing projects together.

“It’s not often you hear of contract money being returned, especially with government contracts. IBM was putting a lot of resources into the contract [but] It wasn’t a profitable direction for IBM.” – Rick Doherty

If the project’s end-product computer had indeed been built, it would have been one of the top most powerful supercomuters in the world according to top500.org. The University of Illinois and the National Science Foundation funded the project, planning then to allow researchers to use the computer for all manner of advanced simulation, data-crunching, and modeling.

In addition to mounting costs and new complexities in the project, the fact that new techniques with potential for less cost and less complexity in their execution have been cropping up since 2008 may have something to do with the abandonment of the project as it existed up until two days ago.

[via PCworld; Image credit: kosheahan]

In fact, they’re both correct: Tianhe-1A combines 14,396 Intel Xeon processors with 7,168 NVIDIA Tesla M2050 GPUs, and manages 2.57 petaflops (quadrillions of calculations per second). Neither company is particularly keen to give the other much credit; Intel refers to NVIDIA’s input only as the vague “accompanied by accelerator cards”, while NVIDIA pointedly highlights the fact that to match Tianhe-1A’s performance solely using CPUs would require around 50,000 of the chips and double the floor space.

Press Release:

Intel Powers World’s Fastest Supercomputer

NEWS HIGHLIGHTS
Tianhe-1A supercomputer, featuring more than 14,000 Intel® Xeon® 5600 series processors, is No. 1 supercomputer on 36th edition of Top500 list.
Intel powers nearly 80 percent of systems on latest Top500 list, including four of the top six.
Intel powers more than 90 percent of new systems on the Top500 list.
Intel demonstrates applications powered by Intel® Many Integrated Core (MIC) architecture.

NEW ORLEANS, Nov. 14, 2010 – Intel Corporation today announced that its Intel® Xeon® 5600 series processors, announced earlier this year, are at the heart of the world’s most powerful supercomputer, the Tianhe-1A. Located at the National Supercomputing Center in Tianjin, China, Tianhe-1A contains 14,396 Intel processors accompanied by accelerator cards, and has demonstrated groundbreaking performance of 2.57 petaflops (quadrillions of calculations per second).

In partnership with Inspur, a computer manufacturer in China, Intel worked closely with the National Supercomputing Center and its technology partners to achieve this groundbreaking performance.

The 36th edition of the TOP500 list of supercomputers, as announced at SC10, Nov. 13-19 in New Orleans, shows that nearly 80 percent of the world’s top 500 systems have Intel processors inside. Such machines are increasingly featured in computers designed for geophysics, financial calculations and scientific research focusing on mainstream applications such as improving the safety of football players and enhancing medical imaging. According to the list, Intel chips now power four of the top six systems – in addition to the No. 1 system. Xeon® 5600 series processors are a key building block in the No. 3 system (Shenzhen), and the newly listed No. 4 system at the Tokyo Institute of Technology. The New Intel® Xeon® Processor 7500 series is featured in the Bull Supernode system at CEA, newly listed at No. 6. Overall, 398 new systems on the list feature Intel processors. According to the list, Intel processors are in 90 percent of the systems newly listed in 2010.

“Our Xeon processor roadmap continues to deliver hugely powerful supercomputers that are helping solve mankind’s greatest challenges,” said Rajeeb Hazra, general manager of Intel’s High Performance Computing organization. “Securing the top position on the Top500 is a source of great pride for Intel, and demonstrates the tremendous leaps in performance and versatility that our processors are delivering across a range of compelling workloads.”

Additional Top500 Success
In addition to the Tianjin supercomputer, 17,296 Intel chips power the No. 6 system on the list. The CEA system from Bull features the largest shared memory system built around the Xeon® 7500 series processor, achieving performance in excess of one petaflop.

Another notable supercomputer hails from the Tokyo Institute of Technology. Featuring Xeon® 5600 series processors within an NEC/HP system, this No. 4-ranked supercomputer achieved 2.4 petaflops.

The semi-annual TOP500 list of supercomputers is the work of Hans Meuer of the University of Mannheim, Erich Strohmaier and Horst Simon of the U.S. Department of Energy’s National Energy Research Scientific Computing Center, and Jack Dongarra of the University of Tennessee. The complete report is available at www.top500.org.

Intel® Many Integrated Core (MIC) Demonstrations
During SC10, Intel conducted demonstrations showcasing the real-world capabilities of the recently announced Intel® MIC architecture. These included using Intel® MIC architecture as a co-processor running financial derivative Monte Carlo demonstrations that boasted twice the performance of those conducted with prior generation technologies. The Monte Carlo application for Intel® MIC was generated using standard C++ code with an Intel® MIC-enabled version of the Intel® Parallel Studio XE 2011 software development tools, demonstrating how applications for standard Intel CPUs can scale to future Intel® MIC products.

Intel also showcased breakthrough compressed medical imaging developed with Mayo Clinic on “Knights Ferry,” the first Intel MIC design and development kits. This demonstration used compressed signals to rapidly create high-quality images, reducing the time a patient has to spend having an MRI.

That’s because the NVIDIA GPUs can each be set to work doing the task of several equivalent CPUs.  In fact, 7,168 Tesla M2050 GPUs (along with 14,336 CPUs) have been harnessed, which its creators, the National University of Defense Technology, reckon would normally require over 50,000 regular CPUs and double the floor space.  It’s also a more power-efficient setup, consuming “only” 4.04 megawatts – a third the consumption of a regular machine.

As for what the Chinese plan to do with Tianhe-1A, they’re going to run it as an open access system for large scale scientific computations.

Press Release:

NVIDIA Tesla GPUs Power World’s Fastest Supercomputer

Half the Size, Lower Power and 50% Faster Than World’s Top Supercomputer

The Tianhe-1A Supercomputer, located at National Supercomputer Center, Tianjin

SANTA CLARA, CA — (Marketwire) — 10/28/2010 — Tianhe-1A, a new supercomputer revealed today at HPC 2010 China, has set a new performance record of 2.507 petaflops, as measured by the LINPACK benchmark, making it the fastest system in China and in the world today(1).

Tianhe-1A epitomizes modern heterogeneous computing by coupling massively parallel GPUs with multi-core CPUs, enabling significant achievements in performance, size and power. The system uses 7,168 NVIDIA® Tesla™ M2050 GPUs and 14,336 CPUs; it would require more than 50,000 CPUs and twice as much floor space to deliver the same performance using CPUs alone.

More importantly, a 2.507 petaflop system built entirely with CPUs would consume more than 12 megawatts. Thanks to the use of GPUs in a heterogeneous computing environment, Tianhe-1A consumes only 4.04 megawatts, making it 3 times more power efficient — the difference in power consumption is enough to provide electricity to over 5000 homes for a year.

Tianhe-1A was designed by the National University of Defense Technology (NUDT) in China. The system is housed at National Supercomputer Center in Tianjin and is already fully operational.

“The performance and efficiency of Tianhe-1A was simply not possible without GPUs,” said Guangming Liu, chief of National Supercomputer Center in Tianjin. “The scientific research that is now possible with a system of this scale is almost without limits; we could not be more pleased with the results.”

The Tianhe-1A supercomputer will be operated as an open access system to use for large scale scientific computations.

“GPUs are redefining high performance computing,” said Jen-Hsun Huang, president and CEO of NVIDIA. “With the Tianhe-1A, GPUs now power two of the top three fastest computers in the world today. These GPU supercomputers are essential tools for scientists looking to turbocharge their rate of discovery.”

NVIDIA Tesla GPUs, based on the CUDA™ parallel computing architecture, are designed specifically for high performance computing (HPC) environments and deliver transformative performance increases across a wide range of HPC fields, including drug discovery, hurricane and tsunami modeling, cancer research, car design, even studying the formation of galaxies.

For more information on NVIDIA Tesla high performance GPU computing products, go here

About NVIDIA
NVIDIA (NASDAQ: NVDA) awakened the world to the power of computer graphics when it invented the GPU in 1999. Since then, it has consistently set new standards in visual computing with breathtaking, interactive graphics available on devices ranging from tablets and portable media players to notebooks and workstations. NVIDIA’s expertise in programmable GPUs has led to breakthroughs in parallel processing which make supercomputing inexpensive and widely accessible. The Company holds more than 1,600 patents worldwide, including ones covering designs and insights that are essential to modern computing. For more information, see www.nvidia.com.

The Akara 2010 managed to defeat the top women’s shogi player, Ichiyo Shimizu, in a matter of six hours, in 86 moves. The match was an event that took place at the University of Tokyo, and according to Japan’s national broadcaster NHK, Akara 2010 “aggressively pursued Shimizu from the beginning,” ultimately resulting in its success over the human player. This is the first time that a computer has been able to defeat a human player at shogi, which is said to be more complex than the western version.

The Japan Shogi Association is currently analyzing the data of the decisive game. Their goal? To determine whether or not they will allow the Akara 2010 to go up against a male professional shogi player. Shimizu made an interesting comment at the end of the match, saying that “It made no eccentric moves, and from partway through it felt like I was playing against a human.” She went on to add that she hopes that computers and humans can become stronger, through friendly competition. We’ll have to wait until the second match between Akara 2010 and a human player to find out if it was just a fluke, or if the computer is able to consistently defeat humans in the shogi game.

[via New Scientist]

The new system is being called Koronis and has shared memory system, ultrafast disk storage, and high-end visualization capabilities. Koronis features 1152 cores and the shared memory architecture can access all 3.1TB of memory directly.

The supercomputer will have attached high-performance workstations for visualization to handle biomed data and more. The university states that since the machine was purchased with NIH grant money only those working on projects funded by NIH can use the machine.

The machine is scalable from 32 to 2048 cores with architectural provisioning for up to 262,144 cores. The machine supports up to 16TB of global shared memory in a single system image. The supercomputer also uses the SGI high speed 15Gb per second NUMAlink 5 interconnect and MPI Offload Engine.

The system is designed to run Novell SUSE or Red Hat Linux. The Altix UV comes in two models including the Altix UV 1000 designed for scalability in a cabinet with up to 256 sockets and 16TB of memory in four racks. The Altix UV 100 is a standard rack mount 3U system with up to 96 sockets and 6TB of memory in two racks.

Together that’s a total of 960 graphics cores, which ASUS are billing as a more cost effective and energy-efficient alternative to traditional high-performance workstations.  Still, the PSU is rated for 1,100W and – while ready to ship now – it’s not an impulse purchase.  ASUS haven’t revealed definitive figures, but they do say the ECS 1000 “has a cost structure in software and hardware of US$14,519 over five years.”

RAM in the desktop supercomputer amounts to 24GB of DDR3-1333MHz memory, while the hard-drive is a 500GB SATA-II unit.  The Tesla C1060 cards themselves are double-width PCI and each have 4GB of memory; they use parallel computing between the hundreds of cores each squeezes in to break down larger tasks and churn through them more quickly.  ASUS worked with NVIDIA and the National Chao Tung University in Taiwan to develop the ECS 1000.

[via Slashdot]

The Octane III is an office ready machine that uses normal wall outlets and measures one foot x two feet and is whisper quiet. The low noise output of a high performance computer is a big deal for anyone who has worked in the office with a large server with lots of fans inside.

The machine permits up to 80 computing cores and nearly 1TB of memory. Configurations are available using NVIDIA GP-GPUs and up to ten dual-socket quad-core Xeon 5500 nodes. The machine can also be had in configurations with Xeon 3400 nodes of 19 single socket quad-core CPUs and 19 single socket dual-core Atom processor nodes. Pricing for the Octane III starts at $7,995.