Labs daub Silicon Valley to accelerate computing

183 views Leave a comment

Scalable computing clusters - like a Tundra clusters being procured from Penguin Computing - are designed to be connected like Legos to emanate some-more absolute systems.

Scalable computing clusters – like a Tundra clusters being procured from Penguin Computing – are designed to be connected like Legos to emanate some-more absolute systems.

The National Nuclear Security Administration’s (NNSA’s) Lawrence Livermore National Laboratory currently announced a awarding of a subcontract to Penguin Computing — a heading developer of high-performance Linux cluster computing systems formed in Silicon Valley — to accelerate computing for save stewardship during a 3 inhabitant confidence laboratories.


Under a terms of a contract, awarded by Lawrence Livermore for NNSA’s tri-lab Advanced Simulation and Computing (ASC) program, Penguin Computing will accept $39 million to yield some-more than 7 Petaflops (quadrillion floating operations per second) of “capacity” computing capability to Los Alamos, Sandia and Lawrence Livermore inhabitant laboratories. These commodity record systems (CTS) are designed to run a vast series of “jobs” concurrently on a singular system.

“These computing clusters will yield indispensable computing ability for NNSA’s day-to-day work during a 3 labs handling a nation’s chief deterrent,” pronounced Doug Wade, conduct of NNSA’s ASC program. “This tri-lab bid will assistance revoke costs, boost operational efficiencies and promote collaborations that advantage a nation’s security, support academia and allege a record that promotes American mercantile competitiveness.”

This plan allows NNSA’s some-more absolute “advanced record system” (ATS) supercomputers to be dedicated to a largest and many formidable calculations vicious to save stewardship. High opening computing (HPC) is a cornerstone of NNSA’s Stockpile Stewardship Program to safeguard a safety, confidence and trustworthiness of a nation’s aging chief halt but testing. The ASC module combines a computing imagination and resources during a 3 labs to allege a computing capabilities that capacitate a scholarship indispensable to accommodate a hurdles of save stewardship.

NNSA’s new ability computing systems, called a Commodity Technology Systems – 1 (CTS-1), will be NNSA’s third corner buying of this form and reinstate those procured in 2011, that are now impending retirement. This tri-lab buying indication reduces costs by economies of scale formed on standardised hardware and module environments during a 3 labs.

Under terms of a contract, computing clusters built of “scalable units” (SUs) will be delivered to any of a laboratories between Apr 2016 and Sep 2018. Each scalable section represents approximately 200 teraflops of computing power. These SUs are designed to be connected, most like Legos, to emanate some-more absolute systems. Scalable units will be divided among a 3 labs with any configuring a SUs into clusters according to goal needs.

Advances in computational technology, enabled in partial by NNSA’s ASC program, have brought down a cost of HPC systems from approximately $100 million per teraflop (trillion floating indicate operations per second) in 1995 to reduction than $5,000 per teraflop today, a cause of 20,000. Each unbroken era of computing complement has supposing larger computing appetite and appetite efficiency.

CTS-1 clusters will support NNSA’s Life Extension Program and investigations into technical issues associated to aging weapons systems, efforts vicious to ensuring a reserve confidence and trustworthiness of a chief weapons in a save as they age good over their dictated deployment life.

Source: LLNL