Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


Archives for August 2008

China Launches Homegrown Supercomputer

China has announced the launch of its latest homegrown supercomputer today.  The new machine, based on homegrown server standards, it said to make 230Tflops peak and a usable compute mark of 160Tflops.  Dawning 5000, as its named, was developed by a cooperation between the Chinese Academy of Sciences [CAS] and Dawning Information Industry Co. Ltd. […]

NVISION 2008: Conference WrapUp

I’m finally home from my first journey to the NVIDIA NVISION conference.  Frankly, I have mixed reviews.  First and foremost, to those readers at NVIDIA.  Please ask Jen-Hsun Huang to hire an outside conference management firm.  The logistics for the conference were far from smooth.  However, since this is your first conference NVIDIA, I’ll give […]

High Performance Computing Ph.D. Fellowship Applications due 9/8

I’m told the deadline on these is hard: absolutely no exceptions. The ACM/IEEE Computer Society High Performance Computing (HPC) Ph.D. Fellowship Program is now accepting nominations for its second annual competition at https://submissions.supercomputing.org. The deadline for submissions is Friday, September 8, 2008. ACM/IEEE-CS HPC Ph.D. Fellowships are awarded with a certificate and a stipend of […]

Cray workshop in Edinburgh

My pal and regular reader Rich Hickey (hi, Rich!) dropped me an email to let me know about an upcoming Cray technical workshop in Scotland. Here are the details Please join us at the 5th Cray Technical Workshop Europe, hosted by EPCC at the University of Edinburgh in Scotland during 24-26 September 2008. This workshop […]

Fujitsu supercomputer donation to Computer History Museum

Here’s a shorty announced yesterday by Fujitsu Fujitsu has donated to the Computer History Museum several components based on technologies developed for the Numerical Wind Tunnel supercomputer, which was ranked No. 1 when it debuted in the Top 500 Supercomputer Sites list in November 1993. The museum will accept the gift at a dedication ceremony […]

Cray's new liquid cooling system

Today Cray announced a new liquid cooling approach that promises more efficient cooling. The system is called ECOphlex (PHase-change Liquid EXchange), and will ship in new XT5s from the company later this year. Here’s an interesting quote from the release The ECOphlex technology is designed to be “room air neutral,” meaning that the temperature of […]

Star-Trib: "NCAR layoffs shouldn't affect Cheyenne supercomputer"

From Joan Barron at the Star-Tribune comes the following news about continued hard times at NCAR: CHEYENNE n [sic] Budget problems at the National Center for Atmospheric Research in Boulder, Colo., should not affect the proposed $60 million Cheyenne supercomputer project, officials said this week. NCAR is facing a $10M budget shortfall this year according […]

British NHS supercomputer program under review

ComputerWorld UK had an article last week about increasing pain in the British government’s troubled IT program in the National Health Service. Some high points: The Conservative party has commissioned the British Computer Society to review NHS IT policy including the £12.7 billion National Programme for IT, which they claim is “crashing down”. The review […]

"World's first supercomputer outside US"

Here’s a great headline from the Economic Times: Canada to get world’s first supercomputer outside the US This refers to the big IBM being built at U of Toronto. At first I thought this was a copy edit problem, but upon further reading I think it’s just a language problem Further, it will also be […]

Hundreds of thousands of threads? Yes, with Erlang

Multicore.info with the pointer to a blog post by Bartosz Milewski on some of the problems with threaded programming in the mainstream languages Thread model based on heavy-duty OS threads and mutexes has its limitations. You can ask server writers, or Google for “thread per connection” to convince yourself. Servers use thread pools exactly because […]