Today is the start of ISC09, and as usual, a new Top500 list kicks off the action. IBM’s Roadrunner hangs on to the top slot, with Cray’s XT5 Jaguar still there in second place. SGI‘s Pleiades system at NASA Ames is pushed down to fourth place, as the new Jugene Bluegene/P system enters at number 3.

The full list, with more details on each of the systems, can be found at http://top500.org/lists/2009/06

Silicon Graphics will be have a stand and be showcasing their wares up the upcoming International SuperComputing 09 conference in Hamburg, Germany, at the end of this month.

Along with showing off the 10,000 core ‘Molecule’ concept, Dr. Eng Lim Goh, SGI’s CTO and all round Very Clever Chap, will be giving a talk on scaling up HPC Architecture to handle multiple core CPUs.

“Scalable HPC Architecture for the Many-Core Era,” presented by Dr. Eng Lim Goh, senior vice president and chief technology officer at SGI.

Date: Thursday, June 25
Time: 10:30 a.m. and 3:15 p.m. CEST
Location: Hall 3, Congress Center Hamburg
Topic: The IT industry has moved from the era of regular clock speed improvements to the new many-core era. This is a fundamental change to which even consumer PC users and applications will need to adapt. This session will focus on the implications for productive high-performance and clustered computing architectures.

I mentioned in a previous post the press release covering the Silicon Graphics Altix installed at the Irish Centre for High-End Computing (ICHEC)

Well, SGI have put up a great video, showing the timelapse installation that took only 3 hours. The entire system was powered up in a day – madness!

Slide on over to http://www.sgi.com/company_info/newsroom/press_releases/2009/february/ichec_video.html and check it out.

SUNNYVALE, Calif., Feb. 10 /PRNewswire-FirstCall/ — Silicon Graphics, Inc. (SGI) (Nasdaq: SGIC) is in a category all by itself when it comes to massive globally shared main memory and globally addressable memory on SGI(R) Altix(R) 4700 systems. With numerous installations in the 4 TB range, and a number more in the 8 TB range, the groundbreaking scalability of SGI systems extends to 21 TB of globally addressable main memory at customer sites. This is over five times the size of memory that other vendors can offer today. The system is designed to accommodate 128 terabytes of globally shared memory under the control of a single instance of the Linux operating system! The system may also be partitioned among multiple instances of Linux and provide globally addressable shared memory among OS instances via SGI’s unique NUMAlink(R) interconnect technology.

More

Russia’s Federal Service for Hydrometeorology and Environmental Monitoring (Roshydromet) after a thorough evaluation and rigorous procurement procedure has deployed Silicon Graphics solutions to rapidly develop detailed models that enable more precise weather forecasts. The systems operate at 27 trillion operations per second, providing 10,000 times the computational power of Roshydromet’s previous Cray supercomputer. This enhanced capability has expanded both the forecast duration and the accuracy of these critical forecasts.

More