Hawking's big-bang team harness SGI super power

A precise science demands big hardware


Cambridge University cosmologists working with physicist Stephen Hawking are getting their first real taste of supercomputing power as they upgrade to Silicon Graphics' Altix UV parallel supers.

The UK Computational Cosmology Consortium, which was established in 1997 by Hawking to probe the structure of the universe in the immediate wake of the hot Big Bang, now has over 30 researchers from ten different universities around the country.

COSMOS, as the consortium and its systems are both known, has been an SGI customer since day one, starting out with a 32-processor Origin 2000 parallel system back in 1997 when that was still pretty cool iron.

The operation upgraded through the Origin MIPS/Irix line and eventually moved to the Linux-Itanium Altix 3700 systems from SGI in 2003. The last upgrade that COSMOS did back in September 2007 was to a 152-core Altix 4700 shared memory system using 1.6GHz, dual-core Itanium 2 processors and equipped with 456GB of shared memory and 25TB of storage.

This is an embarrassingly small supercomputer for such a distinguished physicist and one that should have, along with his long-time partner, Roger Penrose, received a Nobel Prize in Physics.

Like Einstein, Hawking has done his best work using the wetware that nature gave him, but concedes that at this point, cosmological research needs a less blunt instrument than the human brain.

"Recent progress towards a complete understanding of the universe has been impressive, but many puzzles remain," Hawking said in a statement accompanying the Altix UV news. "Cosmology is now a precise science, and we need supercomputers to calculate what our theories of the early universe predict and test them against observations of the present universe."

To that end, COSMOS is getting a three-rack Altix UV 1000 machine, code-named UltraViolet by SGI, that has just shy of 1,000 cores and a little over 2TB of shared memory. The idea is to simulate the conditions at the very beginning of the Big Bang, including the background radiation in the microwave part of the spectrum and the initial structure of the hot, dense universe, and then see how fast-forwarded simulations compare to the actual universe we observe. You can see the scope of the Hawking team's research here.

This Altix UV machine being installed at COSMOS will span three racks, including room for storage. COSMOS and SGI did not divulge the teraflops rating on this machine, but a fully loaded Altix UV 1000 with 2,048 cores - that's 256 of the eight-core Xeon 7500 processors from Intel - is rated at 18.5 teraflops, with a maximum of 16TB of shared memory.

This basic Altix UV processing building block takes up four racks by itself. These four-rack blade modules can be interlinked to create petaflops-scale parallel supers, but not with a single shared memory space spanning all of the blades.

If you do the math, it looks like COSMOS is getting an Altix UV machine that should be rated at nine teraflops. The prior Altix 4700 machine was rated at 937 gigaflops doing double-precision floating point math - something you can get out of two Nvidia or ATI graphics cards these days. While the new box is not exactly a speed demon, with several petaflops-class machines already installed around the world - that's a lot more oomph than COSMOS had.

COSMOS could have created a custom cluster using a mix of x64 and GPU technologies, as many labs are starting to do as they bump up against thermal, budget, and space limitations in their data centers.

But by staying with the Altix UV all that COSMOS has to do is port its code from Itanium to x64 processors and recompile it on the new machine. The Altix UV architecture does allow for GPUs to be plugged into server blades if that helps boost performance for particular workloads.

SGI says that as the COSMOS team needs more oomph, they plan to add more Xeon 7500 processors at this point. ®


Other stories you might like

  • India reveals home-grown server that won't worry the leading edge

    And a National Blockchain Strategy that calls for gov to host BaaS

    India's government has revealed a home-grown server design that is unlikely to threaten the pacesetters of high tech, but (it hopes) will attract domestic buyers and manufacturers and help to kickstart the nation's hardware industry.

    The "Rudra" design is a two-socket server that can run Intel's Cascade Lake Xeons. The machines are offered in 1U or 2U form factors, each at half-width. A pair of GPUs can be equipped, as can DDR4 RAM.

    Cascade Lake emerged in 2019 and has since been superseded by the Ice Lake architecture launched in April 2021. Indian authorities know Rudra is off the pace, and said a new design capable of supporting four GPUs is already in the works with a reveal planned for June 2022.

    Continue reading
  • Prisons transcribe private phone calls with inmates using speech-to-text AI

    Plus: A drug designed by machine learning algorithms to treat liver disease reaches human clinical trials and more

    In brief Prisons around the US are installing AI speech-to-text models to automatically transcribe conversations with inmates during their phone calls.

    A series of contracts and emails from eight different states revealed how Verus, an AI application developed by LEO Technologies and based on a speech-to-text system offered by Amazon, was used to eavesdrop on prisoners’ phone calls.

    In a sales pitch, LEO’s CEO James Sexton told officials working for a jail in Cook County, Illinois, that one of its customers in Calhoun County, Alabama, uses the software to protect prisons from getting sued, according to an investigation by the Thomson Reuters Foundation.

    Continue reading
  • Battlefield 2042: Please don't be the death knell of the franchise, please don't be the death knell of the franchise

    Another terrible launch, but DICE is already working on improvements

    The RPG Greetings, traveller, and welcome back to The Register Plays Games, our monthly gaming column. Since the last edition on New World, we hit level cap and the "endgame". Around this time, item duping exploits became rife and every attempt Amazon Games made to fix it just broke something else. The post-level 60 "watermark" system for gear drops is also infuriating and tedious, but not something we were able to address in the column. So bear these things in mind if you were ever tempted. On that note, it's time to look at another newly released shit show – Battlefield 2042.

    I wanted to love Battlefield 2042, I really did. After the bum note of the first-person shooter (FPS) franchise's return to Second World War theatres with Battlefield V (2018), I stupidly assumed the next entry from EA-owned Swedish developer DICE would be a return to form. I was wrong.

    The multiplayer military FPS market is dominated by two forces: Activision's Call of Duty (COD) series and EA's Battlefield. Fans of each franchise are loyal to the point of zealotry with little crossover between player bases. Here's where I stand: COD jumped the shark with Modern Warfare 2 in 2009. It's flip-flopped from WW2 to present-day combat and back again, tried sci-fi, and even the Battle Royale trend with the free-to-play Call of Duty: Warzone (2020), which has been thoroughly ruined by hackers and developer inaction.

    Continue reading

Biting the hand that feeds IT © 1998–2021