May be of Historical Interest for Supercomputing Geeks & Groupies

This little pamphlet was given to me by a retiring coworker.

Cray Achieves Goal of a Unified, Processor Agnostic Architecture for Supercomputers

Cray has added Intel coprocessors and graphics accelerators from Nvidia to its XC30 supercomputers, in the latest step in its 7-year-old Adaptive Supercomputing initiative.

The initiative, introduced in 2006, was launched to create a unified supercomputer architecture that can support a range of processing technologies. The introduction of support for Intel’s Xeon Phi coprocessors and Nvidia’s Tesla K20X GPU accelerators fits in with that effort, according to Peg Williams, senior vice president of high performance computing systems at Cray.

link.

Cray to Offer Massively SMP Shared Memory Systems

Global supercomputer leader Cray Inc. CRAY -0.75% and ScaleMP(TM), a leading provider of virtualization solutions for high-end computing, today announced the availability of two new preconfigured solutions -- the Cray(R) CS300(TM) SMP system (a shared memory parallel system) and the Cray CS300 LMS system (a large memory system). Both products are optimized configurations of the Cray CS300 cluster supercomputer, and feature vSMP Foundation(TM) software from ScaleMP that is designed to deliver large-scale, high-performance virtual machines for complex modeling analysis and big data challenges.

Preconfigured and managed as one single system, the CS300 SMP and CS300 LMS solutions simplify system administration and programming, eliminating the need for separate cluster management software or having to update multiple operating systems and associated applications. With the CS300 LMS system, users can exploit the power of fast access to high speed memory in data analytics visualization for scientific research exploration, engineering design simulations, bioinformatics, weather simulations and more. Cray CS300 SMP users can leverage the processing speed and large memory in analytics and simulations in areas such as electronic design automation, mechanical computer-aided engineering, bioinformatics and financial analysis.

"We are very pleased to offer our customers these preconfigured CS300 systems that contain massive amounts of memory and large numbers of computer cores, and are also easy to deploy, manage and service," said Daniel Kim, senior vice president and general manager of cluster solutions at Cray. "Researchers and engineers in a wide array of scientific domains with needs for large scale analytics and complex simulations can now take advantage of these very powerful systems."

"Cray has always had a special relationship with the most demanding users, redefining the requirements for high-end systems. With this collaboration, Cray's new large memory and shared memory systems will allow a broader technical computing audience to benefit from the ability to address larger workloads and get faster results," said Shai Fultheim, CEO and founder of ScaleMP.

The CS300 SMP and LMS systems are based on air-cooled configurations of the Cray CS300 system, and feature hot-swappable, redundant high-efficiency power supplies for energy efficiency, reliability and serviceability. These systems are installed in a standard rack cabinet with optional 480V power distribution, and a choice of 208V or 277V three-phase power supplies. In addition, an option for chilled cooling rack, rear door heat exchangers keeps these densely populated configurations extra co

link.

Yes, We Were Hacked

Yep.  Our supercomputer, Franklin (affectionately called Cranky Franky, now decommissioned and replaced with Edison) was hacked.  Here's the story.  If I am given permission, I'll write up more: I need to tread very carefully here.  I will say there's a huge story which isn't in the Wired article. 

Cray to Install a $30 Million, 5 Petaflop (peak) XC30 Supercomputer at University of Edinburgh

Global supercomputer leader Cray Inc. (Nasdaq: CRAY) today announced the Company has been awarded a $30 million contract from the Engineering and Physical Sciences Research Council (EPSRC) to deliver a Cray XC30 supercomputer and a Cray Sonexion storage system to the University of Edinburgh in Scotland as part of the ARCHER project.

ARCHER is the next generation of a national high performance computing (HPC) facility in the UK, and is the follow-on to the High-End Computing Terascale Resource (HECToR) project. The new Cray XC30 supercomputer will provide nearly four times the scientific throughput of its predecessor, HECToR, which is a Cray XE6 supercomputer. It will also be an essential system for scientists in the UK, in particular those funded by EPSRC and the Natural Environment Research Council (NERC).

“One of our primary goals is to increase the ability of UK researchers to make valuable contributions to the solution of important, grand-challenge problems, and the Cray XC30 supercomputer will be a powerful tool in support of these efforts,” said Professor David Delpy, Chief Executive EPSRC. “The ARCHER project advances HECToR’s initiatives to assemble the computational resources necessary for breakthrough research in a broad range of disciplines, and we are pleased that Cray will continue to be a strategic partner for UK research.”

The ARCHER project is focused on an overall vision to build on the existing investments in national HPC facilities so that the UK is a recognized leader on the international scene for computational science and engineering. The project provisions high-end computing resources for use in a wide range of scientific and academic research in fields such as climate, oceanography, life sciences, aerospace, and many others.

“The HECToR and ARCHER programs are the premier supercomputing facilities in the UK and widely respected across Europe, and we are honored that EPSRC will once again provide researchers and scientists with a world-class supercomputer from Cray that is highly scalable, energy efficient and reliable,” said Dr. Ulla Thiel, Cray vice president, Europe. “We are very pleased to continue our collaboration and joint research initiatives with our partner institutions in the UK.”

[...]

The Cray scalable storage solution to be deployed at the University of Edinburgh includes the Cray Sonexion scale-out Lustre system. The solution includes nearly five petabytes of capacity and 100 gigabytes per-second of applications performance. Cray Sonexion vastly reduces deployment time and simplifies Lustre for petascale solutions. Cray Sonexion provides performance scalability from five gigabytes per-second to one terabyte per-second in a single file system – and performs optimally at scale. Management is simplified through component reduction, by over 50 percent for petascale systems.

This man...

The one kneeling while doing the install of NERSC's first Cray-1 supercomputer back in 1978 is downstairs installing NERSC's newest supercomputer, or rather the expansion to it, yes, a cray.  He's one of the longest serving Cray employees.  He's been at Cray for 36 years.  He did the install above after his first year at Cray.  

We're honored to have him back.

If you watch, you might see him.

Cray Nabs NCAR HPC Contract

Global supercomputer leader Cray Inc. (NASDAQ: CRAY) today announced that after an eight-year hiatus, the National Center for Atmospheric Research (NCAR) has become a Cray customer once again with an order for a Cray XT5m supercomputing system. NCAR purchased the world's first production Cray-1A supercomputer in 1976, and continued to perform scientific research on various Cray supercomputing systems for more than 25 years.

NCAR's new Cray XT5m midrange supercomputer, nicknamed "Lynx," will be installed in NCAR's Computational and Information Systems Laboratory (CISL) in Boulder, Colorado later this month. "Lynx" will be used primarily as a development and test platform for CISL and NCAR users who also have access to larger Cray XT supercomputers located at other institutions, including the petascale "Jaguar" and "Kraken" Cray XT5 systems housed at the Oak Ridge National Laboratory and the "Franklin" Cray XT4 system at the National Energy Research Scientific Computing Center.

Hey! Hopper's coming along too! ;)

Cray to Provide Petaflop System to LANL/NNSA Labs

Global supercomputer leader Cray Inc. (NASDAQ: CRAY) today announced that it has signed a sub-contract with Los Alamos National Security, LLC to provide the National Nuclear Security Administration (NNSA) with a next-generation Cray supercomputer. Currently valued at more than $45 million, the multi-year, multi-phase contract can be expanded if the NNSA exercises an option for a future upgrade. The new system will create a new supercomputing platform, named Cielo, for the Advanced Simulation and Computing program at the NNSA.

The Cielo platform will support all three of the NNSA national laboratories, which include Los Alamos National Laboratory, Sandia National Laboratories and Lawrence Livermore National Laboratory. The NNSA will use the new supercomputing system to ensure the safety, security and effectiveness of the United States' nuclear stockpile, and will run the NNSA's largest and most demanding modeling and simulation workload.

"Cielo is being acquired and deployed by the NNSA's New Mexico Alliance for Computing at Extreme Scales (ACES). This is a joint partnership between Los Alamos National Laboratory and Sandia National Laboratories. Both Los Alamos and Sandia have a long history with Cray, going back to the very beginning of the supercomputing era," said John Morrison, High Performance Computing Division Leader at Los Alamos. "With the Cielo platform, that history continues with the next generation of capability computing in support of the U.S. nuclear security enterprise."

"Cielo is the culmination of a two year partnership between Sandia and LANL on ACES," said Sudip Dosanjh, Sandia co-director for ACES. "We look forward to working with Cray to create an order of magnitude increase in capability for key NNSA national security applications. Cielo will target extremely large problems that require production, petascale supercomputing."

"The NNSA plays a critical role in protecting the safety and security of our country, and we are quite proud that some of the organization's most critical scientific research will be done on a Cray supercomputer," said Peter Ungaro, Cray president and CEO. "We have had a great partnership with the NNSA including the development of Red Storm -- a collaboration that enabled the launch of our first Cray XT3 supercomputer. We are honored to be able to continue this important partnership and are encouraged that the NNSA laboratories share in the excitement around our next-generation 'Baker' supercomputer."

The next-generation Cray supercomputer will be housed at the Strategic Computing Complex at the Los Alamos National Laboratory and is expected to be delivered in the second half of 2010. Code-named "Baker," Cray's new supercomputing system will feature a new interconnect chipset known as "Gemini" and enhanced system software that improves the performance, productivity and reliability of the system. Cray's planned "Baker" supercomputer builds on the Cray XT system architecture found in the world's fastest supercomputer and improves it in every key dimension.

no time to comment. I didn't even get a chance to work on the Medea post this weekend.

Cray Sweeps DoD HPC Contract Awards

Global supercomputer leader Cray Inc. (NASDAQ: CRAY) today announced it has won all three high performance computing (HPC) system awards by the Department of Defense (DoD) as part of its 2010 High Performance Computing Modernization Program (HPCMP). The contract, worth more than $45 million including multi-year services, is the largest DoD HPCMP system award to a single vendor in the history of the program.

Under the terms of the contract, Cray will provide three of its next generation supercomputing systems to top DoD Supercomputing Resource Centers located throughout the United States. These centers include the U.S. Air Force Research Laboratory (AFRL) located at the Wright-Patterson Air Force Base in Ohio; the Arctic Region Supercomputing Center (ARSC) in Fairbanks, Alaska; and the U.S. Army Engineer Research and Development Center (ERDC) in Vicksburg, Mississippi. The Cray supercomputers will be used to support basic and applied research, and product development and evaluation. These systems will help to protect the military through the development of new materials, fuels, armor and weapons systems, and assist long-term weather predictions to plan humanitarian and military operations throughout the world.

"To solve today's hardest research and development challenges, we have to employ leading-edge scalable software and the world's fastest supercomputers," said Mr. Cray Henry, director of the DoD's HPCMP. "We are excited to get the next generation Cray supercomputers working to solve our problems."

The three Cray supercomputers included in the 2010 DoD HPCMP procurement are expected to be delivered in the second half of this year, and will be the company's next generation supercomputing systems code-named "Baker." Featuring a new interconnect chipset known as "Gemini" and enhanced system software improving the performance, productivity and reliability of the system, Cray's planned "Baker" supercomputers build on the Cray XT system architecture of the world's fastest supercomputer and improve it in every dimension.

"We are incredibly honored that Cray supercomputers continue to play a prominent role in this important program," said Peter Ungaro, president and CEO of Cray. "As a result of this award, Cray will provide the DoD with the supercomputing tools necessary for its scientists and engineers to research, develop, test and evaluate breakthrough technologies that will strengthen our national security. This award not only reinforces our leadership position in high performance computing, it also speaks to the exciting potential of our upcoming 'Baker' system."

Cray previously received a DoD HPCMP award in 2008, and with the 2010 award, Cray will have supercomputing systems at five of the six DoD Supercomputing Resource Centers in the U.S.

Commentary another time.

Cray Gets KMA Contract for $40 Million

Global supercomputer leader Cray Inc. (NASDAQ: CRAY) today announced that it has signed a multi-year, multi-phase contract to provide the Korea Meteorological Administration (KMA) with a next-generation supercomputer. Valued at more than $40 million, the award is one of the largest in the area of operational numerical weather prediction in the world to-date, and will provide KMA with the necessary supercomputing, storage and data management technologies to reinforce its position as one of the world's premier weather forecasting and research centers.

600 teraflops and it's an XT5.

NERSC Awards Cray $50 Million Contract

Cray Inc. (NASDAQ: CRAY) today announced that the company has won the contract to install a next-generation supercomputer at the Department of Energy's (DOE) National Energy Research Scientific Computing Center (NERSC) located at the Lawrence Berkeley National Laboratory. The multi-year supercomputing contract includes delivery of a Cray XT5(TM) massively parallel processor supercomputer, which will be upgraded to a future-generation Cray supercomputer. When completed, the new Cray will deliver a peak performance of more than one petaflops (quadrillion mathematical calculations per second).

Like NERSC's current 355-teraflops Cray XT4(TM) system, nicknamed "Franklin," the new supercomputing system will help advance open science research in climate modeling, biology, environmental sciences, combustion, materials science, chemistry, geosciences, fusion energy, astrophysics, nuclear and high-energy physics, and other disciplines, along with scientific visualization of massive data sets. Click here for more information about the scientific work done on Franklin.

"As NERSC is the primary supercomputing center for DOE's Office of Science, making Cray's latest technology available to our users will accelerate innovation across a wide range of scientific disciplines, helping scientists tackle problems of vital importance to our nation's future," said Michael Strayer, Associate Director of DOE's Office of Advanced Scientific Computing Research.

According to NERSC Director Kathy Yelick, Cray was awarded the contract based on several factors, including performance and energy efficiency on a set of application benchmarks that capture the challenging workload of the 3,000 NERSC users.

"Because we serve such a large and scientifically diverse user community, it's critical that our systems deliver the best performance while running real-world applications -- especially as users scale their codes to run on tens of thousands of processor cores," Yelick said.

The new Cray system will provide many pioneering features, including the ability for users to customize the operating system for their own codes and to schedule jobs and access their data without logging in to the supercomputer. Yelick adds, "Cray's new cooling system and interconnect network technology mesh well with our research efforts into energy efficient computing and programming models."

"We are proud that NERSC chose Cray as its ongoing partner to provide its diverse and demanding users with advanced scientific computing capabilities," said Peter Ungaro, president and CEO of Cray. "Our partnership with NERSC expands beyond our supercomputers to working together to get the most efficient and effective use of the systems as possible -- a partnership that benefits all Cray customers around the globe. We are excited at the scientific achievements that NERSC's users have made on the Cray XT4 'Franklin' system, and we are looking forward to the advancements that will be made on both our Cray XT5 system as well as our future systems that are part of this contract."

Consisting of products and services, the multi-year contract is valued at over $50 million. The full system is expected to go into production in late 2010.

So, it's public. NERSC's love affair with Cray has been renewed. More than a petaflop! Woo! I have lots of details I can't share. Sorry.

We have another HPC machine that'll hit the floor relatively soon, too. Mums the word until then.

Cray Turns a Profit

Cray Inc. reported third-quarter net income of $5 million, or 15 cents per share, which compares with net income of $5.1 million, or 16 cents per share a year earlier.

Revenue for the Seattle supercomputer company (NASDAQ: CRAY) was $54.6 million compared with $55 million in 2007.

Analysts polled by Thomson Financial Network expected a loss of 5 cents per share and revenue of $54.3 million.

Yea!

We like our vendors to suffer...but not die.

*THIS*IS*NOT*A*SUPERCOMPUTER*

Let me state this is my opinion and only my opinion NERSC does not necessarily agree with me. Frankly, IDK what they think on the matter.

This is NOT a supercomputer.

Let me say that louder:

This is NOT a supercomputer.

64 cores is around 16 processors. Once this would have been a supercomputer, like when I first got involved back in the early 90s. Nowadays when the fastest machines have 38k+ cores, 64 is nada. STOP with the inane marketing, folks.

I really, really mean: JUST STOP! You're embarrassing yourselves.

HPC is a moving target and this, this "supercomputer" picture above is a server and nothing more.

Cray Wins Contract

Global supercomputer leader Cray Inc. (NASDAQ: CRAY) today announced that Cray UK Ltd., a wholly owned subsidiary of Cray, has signed contracts with UoE HPCX Ltd. and the UK's Engineering and Physical Sciences Research Council (EPSRC) to provide a massively parallel processing (MPP) hybrid supercomputer that will become one of the most powerful computing systems in the world. The multi-year contract, valued at approximately $85 million, includes hardware, maintenance and associated professional services.

[...]

The Cray supercomputer will be the first hybrid computer that is based on mature, proven technology. It will consist of Cray's XT4(TM) AMD Opteron(TM)-based MPP system enhanced with a tightly integrated next-generation Cray vector system code-named "BlackWidow." The design leverages Cray's Adaptive Supercomputing vision, which incorporates proven supercomputing processor technologies into a single hybrid system. The contract also includes a next-generation Cray MPP system, code-named "Baker," to be installed in 2009 that will further boost sustained application performance.

So Cray has sold their first nextgen vector system. To be sure, it's attached to a an XT4 basically. If you squint, it looks like an inverted T3D. This is a step to Cray's HPCS system.

New Toy Has Arrived.

They're installing it downstairs now. There's even a little Cray Lady.

Damn it's impressive how fast they're moving.