A look at IBM's Sequoia - the world's fastest supercomputer (photos)

A look at IBM's Sequoia - the world's fastest supercomputer (photos)

Summary: IBM's Sequoia supercomputer retakes the crown for the fastest supercomputer in the world.

SHARE:
TOPICS: IBM, CXO
1

 |  Image 3 of 12

  • Thumbnail 1
  • Thumbnail 2
  • Thumbnail 3
  • Thumbnail 4
  • Thumbnail 5
  • Thumbnail 6
  • Thumbnail 7
  • Thumbnail 8
  • Thumbnail 9
  • Thumbnail 10
  • Thumbnail 11
  • Thumbnail 12
  • Sequoia was built by IBM at the Lawrence Livermore National Laboratory for the National Nuclear Security Administration.

    IBM placed 42.6 percent off the Top 500 Supercomputers. HP was second with 138 supercomputers for a 27.6 percent, and Cray was third with 26 supercomputers - 5.2 percent of those on the list.

    Intel dominated the chip list with 70 percent using it Xeon processor. In celebration the chipmaker has released the production schedule for their Xeon family Many Integrated Cores Architecture (MIC) co-processor called Xeon Pi. It can reach 1 TeraFLOP performance from a single PCIe card.

    Here's David Chernicoff's take on the new processor.

    Credit: Lawrence Livermore National Laboratory

  • How was this list determined?
    The group used the Linpack Benchmark which is a "measure of a computer’s floating-point rate of execution. It is determined by running a computer program that solves a dense system of linear equations. Over the years the characteristics of the benchmark has changed a bit. In fact, there are three benchmarks included in the Linpack Benchmark report." Here's a list of FAQ's about the Linpack Benchmark.

    This photo shows the 96 racks of the Sequoia system as they are being  installed - while integration is ongoing.

  • A member from the IBM install team works on the ASC Sequoia computer. The supercomputer is used for nuclear research including how to prolong the life of older nuclear weapons.

Topics: IBM, CXO

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Related Stories

Talkback

1 comment
Log in or register to join the discussion
  • 10 Tf from a single core

    not part of a company so to speak, but I am deep in to Hardware Engineering, Quantum Mechanics and Electronics with Qualifications in all 4. I have a design that makes use of Molecular Computing and Electronics and will allow for speeds of up to 10.71 Terra flops per second. That is faster then Nvidea's best Tesla Card. the core size itself is no bigger then the size of a 555 chip W 1.7cm, L 2.42cm, D 0.3cm

    jonathonmirza@hotmail.co.uk

    contact me if intrested
    yojojo12