Under the joint Collaboration of Oak Ridge, Argonne, and Lawrence Livermore (CORAL) initiative, the U.S. Department of Energy (DOE) announced a $200 million investment to deliver a next-generation supercomputer, known as Aurora, to the Argonne Leadership Computing Facility (ALCF). When commissioned in 2018, this supercomputer will be open to all scientific users – drawing America’s top researchers to Argonne National Laboratory. Additionally, Under Secretary Orr announced $10 million for a high-performance computing R and D program, DesignForward, led by DOE’s Office of Science and National Nuclear Security Administration (NNSA).
CORAL was established to leverage supercomputers that will be five to seven times more powerful than today’s top supercomputers and help the nation accelerate to next-generation exascale computing. DOE earlier announced a $325 million investment to build state-of-the-art supercomputers at its Oak Ridge and Lawrence Livermore laboratories. Aurora should have 180 peak petaflops.
Intel will work with Cray Inc. as the system integrator sub-contracted to provide its industry-leading scalable system expertise together with its proven supercomputing technology and HPC software stack. Aurora will be based on a next-generation Cray supercomputer, code-named “Shasta,” a follow-on to the Cray® XC™ series.
The US government has refused to let Intel help China update the world’s biggest supercomputer. The Tianhe-2 uses 80,000 Intel Xeon chips to generate a computational capacity of more than 33 petaflops. This year the Chinese machine was due to undergo a series of upgrades to boost its number-crunching abilities past 110 petaflops. The upgrades would depend largely on new Intel Xeon chips. The chipmaker informed US authorities of its involvement with the upgrade programme and was told to apply for an export licence.
In a notice published online the US Department of Commerce said it refused Intel’s application to export the chips for Tianhe-2 and three other Chinese supercomputers because the machines were being used for “nuclear explosive activities”.
China is now believed to be accelerating its own home-grown chipmaking efforts to boost the power of the four supercomputers and complete the upgrade programme.
The website VR World, which was the first to focus the ban, said in a report that the U.S. government is doing something like “cutting your nose to spite your face.”
“NUDT and Tianhe may be the losers for now, but only short term. They will simply speed up their” supercomputer chip plan, the report said. “Intel comes out the big loser from this … Then comes Uncle Sam himself: they lost even that little bit of influence on the high end China HPC (high performance computing).”
Several Chinese supercomputer sites were expected to order some 250+ PFLOPS of compute in the next few years (around 500,000 top-end Broadwell-EP Xeon E5v4 processors, or approximately $1 billion high margin list price).
Indigenous China high end processor architectures can now push the government to gradually remove any dependence on the US. This means just one thing: an AMD or Intel x86 processor technology is increasingly becoming errata non grata. Should the Chinese government react in force, it will give the Chinese vendors the blank check support to go all the way a developing their Alpha, POWER and MIPS processors for both the government and the mainstream commercial use.
You may think they are not up to the mark, but remember how fast British ARM architecture became the dominant processing architecture in the world.
China has the best MIPS, key Alpha and other chip architects.
The USA has chosen to motivate China to adopt and make those once competitive with Intel architectures and fully revive them and merge them with GPGPU and other systems.
China is now motivated to toss tens of billions more to make sure they get their hundred petaFLOP and ExaFLOP supercomputers running sooner and to build the domestic semiconductor industry.
China has tried to spur semiconductor development in the past, but remains far behind foreign competitors in both design and manufacturing technology. This time, though, the government’s funding is more generous and its tactics more sophisticated, analysts say. Since the fall of 2013, China has announced plans to funnel more than 160 billion yuan ($25.6 billion) into a variety of chip-related initiatives, with further investment expected.
China still gets more than 90% of semiconductors from abroad, Mr. Jones of IBS estimates, but the percentage of homegrown chips—which stood at 4.5% in 2010—reached 8.6% in 2014.
For Spreadtrum, the government assistance has made a difference, says Mr. Li. The Nasdaq-listed company was acquired at the end of 2013 by a Chinese state-owned firm, Tsinghua Unigroup. Since then, government grants have helped Spreadtrum increase its staff by a third last year to 4,000, with plans to hire an additional 1,500 this year, he said.
The company has also been able to accelerate its chip development. Though Spreadtrum still lags years behind leading U.S. firms, Mr. Li says his company will be able this year to match a key selling point of companies like Qualcomm: the ability to handle speedy long-term evolution—or LTE—cellular connections on a new multifunction “system-on-a-chip.”
Previously, only one Chinese chip firm had LTE capability, a Huawei Technologies Co. subsidiary named Hisilicon that makes chips almost exclusively for Huawei-branded devices.
Foreign giants in the roughly $350 billion global semiconductor business seem to have little to fear soon from Chinese competition. Chinese chip design companies only held 8% of the global market in 2014, according to research firm IC Insights, and this was mainly low-end chips.
SOURCES- Wall Street Journal, VR World, EWeek, Energy.gov