July 30, 2003

IBM plans world's most powerful Linux supercomputer

Japanese research lab places order for supercomputer cluster

TOKYO -- A Japanese national research laboratory has placed an order with IBM for a supercomputer cluster that, when completed, is expected to be the most powerful Linux-based computer in the world.

The order, from Japan's National Institute for Advanced Industrial Science and Technology (AIST), was announced by the company on Wednesday as it simultaneously launched the eServer 325 system on which the cluster will be largely based. The eServer 325 is a 1U rack mount system that includes two Advanced Micro Devices. Opteron processors of either model 240, 242, or 246, said IBM in a statement.

The supercomputer ordered by AIST will be built around 1,058 of these eServer 325 systems, to make a total of 2,116 Opteron 246 processors, and an additional number of Intel servers that include a total of 520 of the company's third-generation Itanium 2 processor, also known by its code name Madison.

The Opteron systems will collectively deliver a theoretical peak performance of 8.5 trillion calculations per second while the Itanium 2 systems will add 2.7 trillion calculations per second to that for a total theoretical peak performance for the entire cluster of 11.2 trillion calculations per second.

That would rank it just above the current most powerful Linux supercomputer, a cluster based on Intel's Xeon processor and run by Lawrence Livermore National Laboratory (LLNL) in the U.S. That machine has a theoretical peak performance of 11.1 trillion calculations per second, according to the latest version of the Top 500 supercomputer ranking.

Based on that ranking, the new machine would mean Japan is home to two out of the three most powerful computers in the world. The current most powerful machine, the NEC-built Earth Simulator of the Japan Marine Science and Technology Center, has a theoretical peak performance of 41.0 trillion calculations per second while that of the second-fastest machine, Los Alamos National Laboratory's ASCI Q, is 20.5 trillion calculations per second.

The eServer 325 can run either the Linux or Windows operating systems and the supercomputer ordered by AIST will run SuSE Linux Enterprise Server 8. IBM said it expects to deliver the cluster to AIST in March, 2004. AIST will link the machine with others as part of a supercomputer grid that will be used in research of grid technology, life sciences bioinformatics and nanotechnology, IBM said.

General availability of the eServer 325 is expected in October this year and IBM said prices for the computer start at $2,919. The computers can also be accessed through IBM's on-demand service where users pay for processing power based on capacity and duration.

IBM's announcement is the second piece of good news for AMD and its Opteron processor within the last two weeks. The processor, which can handle both 32-bit and 64-bit applications, was launched in April this year.

China's Dawning Information Industry Co. announced plans last week to build a supercomputer based on AMD's Opteron processor. The Dawning 4000A will include more than 2,000 Opteron processors, with a total of 2TB of RAM and 30TB of hard-disk space and is expected to deliver performance of around 10 trillion calculations per second. The Beijing-based company has an order for the machine but has not disclosed the name of the buyer or when the computer will be put into service.

Opteron processors were also chosen for a supercomputer which is likely to displace the AIST machine as the most powerful Linux supercomputer. Cray is currently constructing a Linux-based supercomputer called Red Storm that is expected to deliver a peak performance of 40 trillion calculations per second when it is delivered in late 2004. Linux developer SuSE is also working with Cray on that machine.

 

Close

On Twitter now

Platforms

Powered by Twitter

On Twitter now

White Paper

D2D Virtual Tape Library Replication Primer

This whitepaper explains the terminology and concepts behind Data Replication technologies and establishes some sizing rules through worked examples. Learn the new paradigm in disaster tolerance—protect data anywhere.

Download now »

White Paper

An Alternative to Virtualization for Datacenter Cost Savings

Server virtualization is a popular option for dealing with mounting datacenter costs. Another equally promising approach is the use of an Application Delivery Controller. Citrix NetScaler provides a low-cost way for organizations to reduce their server count and accrue cost savings from a reduction in space, cooling, power and personnel.

Download now »

White Paper

Why Your Firewall, VPN, and IEEE 802.11i Aren't Enough to Protect Your Network

The emergence of WLANs has created a new breed of security threats to enterprise networks.

Included in HP ProCurve WLAN solutions is security technology that alleviates threats from WLANs through:
* Monitoring wireless activity inside and out of the enterprise
* Classifying WLAN transmissions into harmful and harmless
* Preventing transmissions that pose a security threat to the enterprise network
* Locating participating devices for physical remediation

Download now »

White Paper

Bringing the Edge to the Data Center

Effectively address data protection challenges, implementing solutions that help store and protect business–critical data while cutting costs and improving efficiency and reliability.

Download now »

Sign up to receive Platforms Resource Alerts

Subscribe to the Today's Headlines: First Look Newsletter

Find out what will be news for the day, with our first-thing-in-the-morning briefing.

©1994-2009 Infoworld, Inc.