Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
Supercomputing

University of Texas Announces Fastest Academic Supercomputer In the World (utexas.edu) 31

Posted by EditorDavid from the think-fast dept.
On Tuesday the University of Texas at Texas launched the fastest supercomputer at any academic facility in the world.

The computer -- named "Frontera" -- is also the fifth most-powerful supercomputer on earth. Slashdot reader aarondubrow quotes their announcement: The Texas Advanced Computing Center (TACC) at The University of Texas is also home to Stampede2, the second fastest supercomputer at any American university. The launch of Frontera solidifies UT Austin among the world's academic leaders in this realm...

Joined by representatives from the National Science Foundation (NSF) -- which funded the system with a $60 million award -- UT Austin, and technology partners Dell Technologies, Intel, Mellanox Technologies, DataDirect Networks, NVIDIA, IBM, CoolIT and Green Revolution Cooling, TACC inaugurated a new era of academic supercomputing with a resource that will help the nation's top researchers explore science at the largest scale and make the next generation of discoveries.

"Scientific challenges demand computing and data at the largest and most complex scales possible. That's what Frontera is all about," said Jim Kurose, assistant director for Computer and Information Science and Engineering at NSF. "Frontera's leadership-class computing capability will support the most computationally challenging science applications that U.S. scientists are working on today."

Frontera has been supporting science applications since June and has already enabled more than three dozen teams to conduct research on a range of topics from black hole physics to climate modeling to drug design, employing simulation, data analysis, and artificial intelligence at a scale not previously possible.
Here's more technical details from the announcement about just how fast this supercomputer really is.

From the Texas Advanced Computer Center's official announcement:

First announced in August 2018, Frontera was built in early 2019, and earned the #5 spot on the twice-annual TOP500 list in June, achieving 23.5 PetaFLOPS (23.5 thousand million million floating-point operations per second) on the high-performance LINPACK benchmark, a measure of the system's computing power...

Frontera combines Dell EMC PowerEdge servers with 8,008 compute nodes, each of which contains two 2nd generation Intel Xeon scalable ("Cascade Lake") processors, totaling more than 16,000 processors and nearly half a million cores, connected by a 200 gigabit per second HDR Mellanox InfiniBand high-speed network.

The system incorporates innovative flash storage from DataDirect Networks and novel cooling systems from CoolIT, Cooltera, and Green Revolution Cooling (GRC), and employs several emerging technologies at unprecedented scale, including high-powered, high-clock rate versions of the latest Intel Xeon processors, Intel Deep Learning Boost, Intel Optane memory, and several kinds of liquid cooling.

In August, Frontera added two new subsystems — using technologies from NVIDIA, IBM and Green Revolution Cooling (GRC) — which provide 11 PetaFLOPS of additional single precision performance and to explore alternate computational architectures for the future. A 360 NVIDIA Quadro RTX 5000 GPU (graphics processing unit) system submerged in liquid coolant racks developed by GRC will explore more efficient ways to cool future systems, as well as explore single-precision optimized computing. An IBM POWER9-hosted system with 448 NVIDIA V100 GPUs will provide additional performance, and provide the top-performing GPU architecture with tight coupling to the processor and memory subsystems.

These additional systems will accelerate artificial intelligence, machine learning, and molecular dynamics research for Frontera researchers in areas ranging from cancer treatment to biophysics.
This discussion has been archived. No new comments can be posted.

University of Texas Announces Fastest Academic Supercomputer In the World More

University of Texas Announces Fastest Academic Supercomputer In the World

Comments Filter:
  • I had $100 on it happening at the University of Texas at Connecticut.

    • color me surprised, I thought the Texas Advanced Computing Center was in Arkansas, like one of the two Taxarkansas was

  • The most bleeding edge, speed-binned parts from OEMs, a complete water cooling system for the whole thing, humidity controlled, and all nodes connected by ultra-high throughput fiber.

    Exceptional
    -
    Processing
    Excessively
    Expensive
    Network

  • They'll be able to deep fake more campus cuties than ever.

  • An IBM POWER9-hosted system with 448 NVIDIA V100 GPUs will provide additional performance, and provide the top-performing GPU architecture with tight coupling to the processor and memory subsystems.

    Hmm . . . IBM POWER folks are located in Austin, Texas . . . NVIDIA too . . . IBM just bought Red Hat . . . is NVIDIA the next on their shopping list . . . ?

    When IBM bought Red Hat it was a serious blow to their z/OS mainframe folks, who have always tried to whack down IBM's AIX Unix and IBM's PC OS/2. They succeeded with OS/2. Apparently AIX is still alive.

    It's not too long ago that I read an IBM Redbook that claimed that z/OS was IBM's "Flagship Operating System."

    Well, now it is Redhat Linux.

    And now

    • Nonsense.

      how would Red Hat Linux be a blow to z/os market? Red Hat Linux can't run software written for Z/VM, Z/VSE nor Z/TPF. The big corporations that spent decades building the mainframe software aren't going to rewrite it all.

      Running Linux on System Z has been around since 1999, nothing new there. Redhat and SLES have made distro for it for years, and Ubuntu for last four years. Five more zero cost distros run on it too. This changes nothing in the mainframe world.

      • how would Red Hat Linux be a blow to z/os market?

        Z system is great, because it is a dedicated zero time down transaction processor. Although IBM tried this in the late 80's, it's not for spreadsheets and text processing. That is done better by PCs. Is anyone here old enough to have used IBM mainframe PROFS or OfficeVision.?

        My condolences.

        Red Hat Linux can't run software written for Z/VM, Z/VSE nor Z/TPF. The big corporations that spent decades building the mainframe software aren't going to rewrite it all.

        Of course they won't rewrite anything. A guy who works at a big insurance company in Germany told me that they don't even have the source for some stuff any more!

        They will let it die out. And then rewrite it for th

        • A guy who works at a big insurance company in Germany told me that they don't even have the source for some stuff any more!

          https://playingintheworldgame.... [wordpress.com]

        • The banking, trading, insurance and big city clients I've had with mainframes had their source code and add/modify it with in-house staff. COBOL, REXX, Java EE, mainframe assembler, C++ continuously developed over decades. On the whole, nothing will die out and there is no reason to port to anything else.

          You do realize IBM and others offers mainframes in cloud? Not seeing any reason RedHat could be a threat to monster code bases with tens of millions of lines for thousands of apps when there is no need t

  • Comment removed based on user account deletion

  • Nvidia, Mellanox. Old tech. Second rate.

  • It can repeat "I hate Trump!" and "We love socialism!" over a trillion times a second!

  • I checked the prices [servethehome.com] and did the math (2^14 processor * 28 cores = 458752 cores) and for the processors alone, it would be more than $160M (2^14 * $10,009).

    This means that in order to afford the rest of the machine and the chips that Intel didn't just give them a discount, they basically sold them at or even below cost.

    Cascade Lake will also be the first Intel microarchitecture to support fixing Spectre and Meltdown security vulnerabilities via hardware changes.

    I ponder if the hardware fixes are actual fixes or just shitty patch jobs around the issues, akin the the microcode patches.

    • . . . and that's why they used Cascade Lake instead of Rome.

    • Re: (Score:2)

      by Junta ( 36770 )

      Incidentally, if you dig into *any* top ~15 class supercomputer, you'll see one or several vendors take a bath to get the marketing win. This is actually one reason for organizations to pool their resources into a single bid, as the marketing leverage will buy a much cheaper solution in aggregate.

      Ditto for Amazon/Microsoft/Google, who pay *far* below market prices due to the tantalizing volumes.

      Sort of similar to how Wal-Mart bullies their suppliers of retail goods.

  • The UT TACC blurb on this project breathlessly states "major advancement in high-speed quantum computation". Absolutely dead wrong, and inexcusable coming from a major research university, even it is from the goofily named "Pickle Research Center" campus. "Quantum computation" has a specific meaning, and you do not do it with Xeons. The actual client investigator's quote "It's a great machine, especially for quantum mechanics applications" a sentence later, is the credible statement.
  • The record was achieved using 1809 and when they upgraded to 1903, they dropped to the 10th position.

Slashdot Top Deals

You know you've landed gear-up when it takes full power to taxi.

Close