Visit additional Tabor Communication Publications
December 22, 2011
The supercomputing biz seems to have shaken off most of the after-effects of the global recession, with scads of new deployments large and small around the world. China, in particular, continued its big push into HPC, notching its first home-grown super. And Japan ushered in the era of 10-petaflop supercomputing this year with its world-beating K Computer. But, as always, not all the HPC news was rosy. Here are the top hits and misses for the year.
Hit: IBM Watson Ushers in New Era of Intelligent Machines
If you're an HPC enthusiast, the biggest feel-good story of the year has to be the debut of IBM's Watson supercomputer, which trounced trivia-champs Ken Jennings and Brad Rutter in a nationally televised Jeopardy tournament in February. The Watson software not only showed an uncanny ability to decipher obscure game clues and spit out correct answers at lightning speed, but also to use a rather sophisticated betting strategy to maximize its winnings. Watson's impressive Jeopardy performance has already landed the system a paying gig in the health care industry.
Miss: IBM Forsakes NCSA, Blue Waters
IBM was also at the center of the most embarrassing story of the year, namely the abandonment of its $200 million NSF-funded contract to build the 10-petaflop "Blue Waters" supercomputer for the NCSA. IBM had already begun deploying Power7 server hardware, when they came to the realization that it wasn't going to be feasible for them to complete the project under the time and technical constraints imposed by the project. The contract termination in August forced NCSA into a quick search for a new vendor that could sub for IBM. So in November...
Hit: Cray Snatches Up Blue Waters Deal
In what must have been an especially sweet win for Cray, the supercomputer maker came to the rescue of NCSA, promising a Blue Waters supercomputer on schedule (fall 2012) and within budget. The IBM replacement will be based on Cray's XE6 and GPU-equipped XK6 hardware and incorporate NVIDIA's upcoming "Kepler" processors. It is expected to top out at 11.5 peak petaflops and deliver a sustain petaflop for big science applications. The win came on the heels of Cray's October announcement that it would be building the Titan supercomputer to Oak Ridge National Lab -- another multi-petaflop GPU-accelerated machine, also scheduled for installation in 2012.
Hit and Miss: Storage Vendor Compression
The fast-growing HPC storage business got churned up in 2011, with two major acquisitions. In March, NetApp announced that it would acquire LSI's Engenio storage business for $480 million. And in September, Hitachi Data Systems announced that it bought Blue Arc Corporation for an undisclosed sum. Both deals give two prominent storage vendors a prominent play in the HPC biz, but at the risk of subsuming companies that were singularly focused on just high performance computing. The NetApp deal has already paid some dividends, with the company notching a deal for a 55-petabyte storage system for the Sequoia supercomputer at Lawrence Livermore National Laboratory.
Hit: Fujitsu Roars Back Into Supercomputing
Japanese computer maker Fujitsu captured the TOP500 crown in June with its SPARC CPU-based "K Computer" deployed at RIKEN. The 8-petaflop Linpack mark bested the 2 petaflop Tianhe-1A from China, the previous number one system. By November, the fully tricked-out K machine topped 10 petaflops on Linpack. Later in the year, the company commercialized this same K Computer technology (adding the next-generation SPARC64 IXfx CPU) in its new PRIMEHPC FX10 offering, which the company intends to sell internationally (The first system was deployed at the University of Tokyo.) Fujitsu also inked a £15m deal with HPC Wales to deliver a 190-teraflop HPC grid for the Welsh, representing the company's largest ever European deployment in the HPC arena.
Miss: Microsoft Loses HPC Mojo
With rumors swirling most of the year that high performance computing at Microsoft was being ditched, the company revealed that they were reorganizing their HPC organization. Kyril Faenov, who led the Technical Computing Group, was moved to an advisory role, with Ryan Waite, the GM for the company's High Performance Computing group, more or less running the show. The HPC group, though, was folded into the Server and Cloud Division, where the Azure Cloud platform gets most of the attention. The company maintains it will continue development of its Windows HPC Server offering for standalone clusters, but most of the emphasis now seems to be on making the Azure cloud the common denominator for all its enterprise software. The HPC community seems mostly unimpressed.
Hit: China Races Ahead
Despite being dethroned from the number one spot on the TOP500, China continues to wow the HPC community. It currently has two supers in the top 10 and 74 total systems in the top 500, more than any other country except for the US, with 263. In 2002, China claimed just 5 such systems. In addition, China is quickly building a domestic HPC server industry, with Dawning (now Sugon), Inspur, Lenovo, Sunway, and PowerLeader supplying about half the nation's top 100 HPC machines. China also deployed its first native supercomputer this year, the Sunway BlueLight MPP, which sports domestically developed ShenWei SW1600 processors. With plans to have at least 17 petascale supercomputing centers within 5 years, China is on track to become the planet's second HPC superpower.
Hit: Quantum Computing Goes Commercial
In May, D-Wave Systems sold the world's first quantum computer. The buyer was Lockheed Martin Corporation, who did not disclose how they intend to use the machine. The system, named D-Wave One, employs a 128-qubit chip, called Rainier, and uses superconducting technology to generate "adiabatic quantum computing" (that some claim is not true quantum computing). The cost of the system was not disclosed, but undoubtedly this is one of those cases in which if you have to ask, you probably can't afford it.
Miss: SGI Loses Its CEO
This month, server maker SGI announced that CEO Mark Barrenechea had resigned, and will be leaving the company at the end of the year. Barrenechea, who was recently named Best Large Company CEO by the San Francisco Business Times, has been at the helm of SGI since it merged with Rackable in 2009, and has managed to increase sales and narrow losses substantially over that period. The company now appears to be on the most encouraging financial trajectory it has enjoyed in several years. Filling his shoes at SGI is likely to be something of a challenge.
Hit: Intel Nabs Big Win for Manycore MIC Accelerator
Although Intel has yet to produce its first commercial Many Integrated Core (MIC) part, aka Knights Corner, the chip maker has already chalked up a significant win. In September, the Texas Advanced Computing Center (TACC) announced it will team with Dell and Intel to build a 10-petaflop supercomputer that will derive most of its flops from the Knights Corner parts, specifically 8 petaflops from the MIC coprocessors, and the other 2 petaflops from Sandy Bridge Xeon CPUs. The system, named "Stampede," is being funded by the NSF to the tune of $27.5 million, and is slated for boot-up in late 2012. At SC11 in November, Intel previewed early silicon of Knights Corner, demonstrating one teraflop of double-precision goodness. While the upcoming Kepler GPU from NVIDIA is likely to match the MIC chip, flops for flops, the competition between the two accelerators is going to be one of the more interesting battles to watch in 2012 and beyond.
Jun 19, 2013 |
Supercomputer architectures have evolved considerably over the last 20 years, particularly in the number of processors that are linked together. One aspect of HPC architecture that hasn't changed is the MPI programming model.
Jun 18, 2013 |
The world's largest supercomputers, like Tianhe-2, are great at traditional, compute-intensive HPC workloads, such as simulating atomic decay or modeling tornados. But data-intensive applications--such as mining big data sets for connections--is a different sort of workload, and runs best on a different sort of computer.
Jun 18, 2013 |
Researchers are finding innovative uses for Gordon, the 285 teraflop supercomputer housed at the San Diego Supercomputer Center (SDSC) that has a unique Flash-based storage system. Since going online, researchers have put the incredibly fast I/O to use on a wide variety of workloads, ranging from chemistry to political science.
Jun 17, 2013 |
The advent of low-power mobile processors and cloud delivery models is changing the economics of computing. But just as an economy car is good at different things than a full size truck, an HPC workload still has certain computing demands that neither the fastest smartphone nor the most elastic cloud cluster can fulfill.
Jun 14, 2013 |
For all the progress we've made in IT over the last 50 years, there's one area of life that has steadfastly eluded the grasp of computers: understanding human language. Now, researchers at the Texas Advanced Computing Center (TACC) are utilizing a Hadoop cluster on its Longhorn supercomputer to move the state of the art of language processing a little bit further.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?
Join our webinar to learn how IT managers can migrate to a more resilient, flexible and scalable solution that grows with the data center. Mellanox VMS is future-proof, efficient and brings significant CAPEX and OPEX savings. The VMS is available today.