Wow, AMD better than Intel (it's official)

It's time, once again, for the Top 500 list of speedy supercomputers. AMD is crowing that the first four are powered by its chips. In IT Blogwatch, bloggers dissect the list and marvel at the big numbers.

By Richi Jennings. November 16, 2009.


Top500 logo
Your humble blogwatcher selected these bloggy morsels for your enjoyment. Not to mention jingle-jingle...

    Stephen Lawson gets his units mixed up:

The list of the Top 500 supercomputers, set to be released on Monday during the SC09 supercomputing conference in Portland, Oregon, is compiled twice a year and is now in its 34th installment. The total capacity of the systems on the new list is 27.6 petaflops, up from 22.6 petaflops on the previous list in June. ...  A Cray supercomputer ... has regained the title of the world's most powerful. ... The upgraded Jaguar supercomputer at Oak Ridge, in Tennessee, now boasts a speed of 1.759 petaflops per second from its 224,162 cores.


China entered the Top 10 with a hybrid Intel-AMD system. ... achieved 563 teraflops per second for the No. 5 ranking. It uses Intel Xeon processors with Advanced Micro Devices GPUs (graphics processing units) as accelerators. Each node of the 71,680-core system has two Xeons attached to two AMD GPUs. ... Tianhe-1 was built by the National University of Defense Technology.

Timothy Prickett-Morgan marvels:

You would never know the world was battling recession. Supercomputer centers keep chucking out old tech and rolling in new tech. ... The single biggest transition in the list is the move to quad-core - and in some notable cases, six-core - processors inside supercomputing systems. And most of the machines on the list now run Linux with x64 processors.


The differences are substantial enough for specific workloads to keep a whole bunch of HPC suppliers busy at a time in the market when you think there might be two suppliers left. ... AMD is still pretty pumped. ... After trying to outrun IBM's hybrid "Roadrunner" Opteron-Cell massively parallel blade server ... for the past several lists, the "Jaguar" all-Opteron XT5 box made by Cray and installed at Oak Ridge National Laboratory has pounced upon Roadrunner and pulled out some flesh and spit out some feathers.

Sebastian Pop notices a notable absence:

A notable fact is that Intel-powered machines weren't even runner-ups. The best-performing Intel-powered conglomerate only managed to score number five. ... One can only guess what kind of reaction this development stirred in Intel directors just days after having to sign an uneasy truce with their rival.


Roadrunner is also powered by AMD CPUs, along with the Kraken (which achieved third place) and Jugene (fourth place). ... Powering the best four supercomputers is an obvious indicator of just how much faster AMD central processing units are (or can be) compared with Intel's, and why the latter couldn't, so far, eliminate the former from the IT market.

Aharon Etengoff talked to the victor:

[AMD] Spokesperson John Frueheh told TG Daily that his company had achieved a "strong presence" in the HPC arena, with both AMD CPUs and ATI GPUs powering a number of top supercomputing systems. ... [He] added that AMD was planning to debut a line of 8- and 12-core x86 processors during the first quarter of 2010.


Codenamed "Magny-Cours," [they] will be capable of delivering up to a 100 percent increase in compute core density and multi-threading capability. In addition, the processors will feature four channels of DDR-3 memory, as well as new power management and virtualization features. AMD is also slated to introduce its "Bulldozer" processor architecture in 2011. The chip's modular design is expected to increase concurrency, simultaneous command capabilities and multi-threading capacity with up to 16 dedicated cores.

But Miha Ahronovitz asks, "So what?":

The business model this list promotes brought the bankruptcy of SGI, Thinking Machines, Cray Research, SiCortex and many who designed supercomputers based on one criteria: to pass a LINPACK test. This was originally introduced in 1979, 30 years ago. It tests the floating point and little more. LINPACK tells nothing of how easy is to solve complex problems with a given supercomputer.


We should take actual applications used by enterprises and test the fastest supercomputers running them. We can have for example a TOP 500 for E-OLTP (Extreme on line transaction processing) computers designed to process more than 500,000 transactions per second, now that all banking and credit card processing and stock exchange need these types of volumes.

So what's your take?
Get involved: leave a comment.

And finally...

Richi Jennings, your humble blogwatcher
  Richi Jennings is an independent analyst/consultant, specializing in blogging, email, and security. A cross-functional IT geek since 1985, he is also an analyst at Ferris Research. You can follow him as @richi on Twitter, or richij on FriendFeed, pretend to be richij's friend on Facebook, or just use good old email:

Don't miss out on IT Blogwatch:

Copyright © 2009 IDG Communications, Inc.

Shop Tech Products at Amazon