x
Loading
 Loading
Hello, Guest | Login | Register

Extreme Linux

In the past several months, a good number of corporate, government, academic, and research institutions — Pixar, the Lawrence Livermore National Lab (LLNL), the Centers for Disease Control (CDC), Shell E&P, and others — have announced the installation of substantial, high performance Linux computing clusters. In the case of LLNL, for example, the largest of its three new clusters (built by Linux Networx) is composed of 252 Pentium 4 processors, capable of a theoretical peak of 857 gigaflops, making it one of the fastest clusters ever built. In the case of Pixar, a 1,024-processor blade cluster (using 2.8 GHz Xeons) from RackSaver is replacing the company’s existing Sparc-based render farm.

In the past several months, a good number of corporate, government, academic, and research institutions — Pixar, the Lawrence Livermore National Lab (LLNL), the Centers for Disease Control (CDC), Shell E&P, and others — have announced the installation of substantial, high performance Linux computing clusters. In the case of LLNL, for example, the largest of its three new clusters (built by Linux Networx) is composed of 252 Pentium 4 processors, capable of a theoretical peak of 857 gigaflops, making it one of the fastest clusters ever built. In the case of Pixar, a 1,024-processor blade cluster (using 2.8 GHz Xeons) from RackSaver is replacing the company’s existing Sparc-based render farm.

Four things are startling about these groups’ adoptions of commodity clusters. The first two surprises are the scale of the machines: the machines are enormous and simultaneously diminuitive. The machines have a mind-boggling number of fast CPUs, yet can fit in a series of 45U racks.

The next surprise is cost: clusters provide the biggest bang for your computing buck. Built from off-the-shelf parts, clusters leverage the pace of Moore’s Law and the dwindling prices of CPUs and memory.

And the last, albeit pleasant surprise, is that Linux is the operating system powering these clusters. Flexible, scalable, and perhaps most important, customizeable, Linux has earned its “street cred” in high performance computing. And as you’ll read in Dr. Thomas Sterling’s personal account of the history of Beowulf clusters, the symbiosis between Linux and Beowulf…

Please log in to view this content.

Not Yet a Member?

Register with LinuxMagazine.com and get free access to the entire archive, including:

  • Hands-on Content
  • White Papers
  • Community Features
  • And more.
Already a Member?
Log in!
Username

Password

Remember me

Forgotten your password?
Forgotten your username?
Read More
  1. Scheduling HPC In The Cloud
  2. GP-GPUs: OpenCL Is Ready For The Heavy Lifting
  3. HPC Madness: March Is More Cores Month
  4. HPC Turn-Offs: Power Control
  5. The Cost to Play: CUDA Programming
Follow Linux Magazine
Rackspace