Feeds

Gawker - Texas's supercomputing Ranger

60,000 cores of justice

Beginner's guide to SSL certificates

We first met Ranger in Oct. of 2006. Back then, Sun Microsystems, AMD and the Texas Advanced Computer Center (TACC) speculated that they might be on the way to producing the fastest supercomputer in the world.

Thanks to AMD's delays shipping a four-core Opteron chip, cabling issues and general InfiniBand woes the top dog dream failed to occur. Instead, Ranger's official unveiling took place this week with the supercomputer standing as the world's fastest "open" system. Fair enough.

Ranger exemplifies what can result from Sun's so-called Constellation designs. With the help of co-founder Andy Bechtolsheim and others, Sun has crafted servers, storage and especially networking gear that can scale - and how. For example, the Ranger system runs on 15,744 of AMD's quad-core "Barcelona" versions of Opteron. It also uses Sun's x4500 'sterver' (hybrid server and storage box) to reach 1.7 petabytes of data capacity. And then there's the Magnum switch, which offers up 3,456 ports of InfiniBand goodness and total bandwidth of 110Tb/s.

The system's peak performance comes in at 504 teraflops, which trails the IBM Blue Gene system sitting at Lawrence Livermore National Lab (LLNL). The IBM system enjoyed an upgrade while TACC sorted out the Opteron issues, bringing it to 106,496 PowerPC-based nodes and a peak performance of 596 teraflops.

Grabbing the top slot may not seem like a big deal, but it is. IBM receives for a mountainous volume of extra press because of Blue Gene's status. Even more impressive, IBM has managed to maintain its kingly position for years, which is tough to do in the ever-upgrading world of supercomputers.

But where LLNL's machine ultimately falls under control of the US Department of Energy, Ranger makes it way to researchers of various stripes.

Built via a $59m award, Ranger will reside on the National Science Foundation's TeraGrid, which pulls together numerous computing centers to provide researchers with lots of horsepower. In addition, the system arrived thanks to the efforts of TACC, the University of Texas's Institute for Computational and Engineering Sciences (ICES), Arizona State University and Cornell University.

"Ranger offers more than six times the performance of the previous largest system for open science research," the parties said.

And in perhaps the oddest calculation comparison of all time, they added, "The boost in performance offered by Ranger relative to the previously largest open science machine is comparable to reducing the flight time from New York to London to just one hour."

That's just fine, but how many gibbon months will this knock off deforestation?

A solid 90 per cent of Ranger power will go to the TeraGrid, while TACC will suck down ten per cent for its own good and for the good of research projects across the great state of Texas.

Sun and AMD, of course, receive their fair share of glory for putting Ranger together as well. But all is not perfect for AMD.

The company made a special effort to get those Opterons to TACC, meaning that it had to send in the buggy first batch of chips. A temporary fix for these chips does cause some performance degradation, which is not what you want in a supercomputer.

Sun's server chief John Fowler, however, has characterized the performance issues as minimal. Nonetheless, AMD will be digging into its own pockets to pay for an upgrade to fixed Opterons in the coming months, as we understand it. ®

Security for virtualized datacentres

More from The Register

next story
It's Big, it's Blue... it's simply FABLESS! IBM's chip-free future
Or why the reversal of globalisation ain't gonna 'appen
'Hmm, why CAN'T I run a water pipe through that rack of media servers?'
Leaving Las Vegas for Armenia kludging and Dubai dune bashing
Facebook slurps 'paste sites' for STOLEN passwords, sprinkles on hash and salt
Zuck's ad empire DOESN'T see details in plain text. Phew!
CAGE MATCH: Microsoft, Dell open co-located bit barns in Oz
Whole new species of XaaS spawning in the antipodes
Microsoft and Dell’s cloud in a box: Instant Azure for the data centre
A less painful way to run Microsoft’s private cloud
AWS pulls desktop-as-a-service from the PC
Support for PCoIP protocol means zero clients can run cloudy desktops
prev story

Whitepapers

Choosing cloud Backup services
Demystify how you can address your data protection needs in your small- to medium-sized business and select the best online backup service to meet your needs.
Forging a new future with identity relationship management
Learn about ForgeRock's next generation IRM platform and how it is designed to empower CEOS's and enterprises to engage with consumers.
Security for virtualized datacentres
Legacy security solutions are inefficient due to the architectural differences between physical and virtual environments.
Reg Reader Research: SaaS based Email and Office Productivity Tools
Read this Reg reader report which provides advice and guidance for SMBs towards the use of SaaS based email and Office productivity tools.
Storage capacity and performance optimization at Mizuno USA
Mizuno USA turn to Tegile storage technology to solve both their SAN and backup issues.