from the leveling-the-playing-field dept.
The Platform reports that CPU export restrictions to Chinese supercomputing centers may have backfired. Tianhe-2 has remained the world's top supercomputer for the last five iterations of the TOP500 list using a heterogeneous architecture that mixes Intel's Xeon and Xeon Phi chips. Tianhe-2 will likely be upgraded to Tianhe-2A within the next year (rather than by the end of 2015 as originally planned), nearly doubling its peak performance from 54.9 petaflops to around 100 petaflops, while barely raising peak power usage. However, instead of using a new Intel Xeon Phi chip, a homegrown "China Accelerator" and novel architecture will be used.
A few details about the accelerator are known:
Unlike other [digital signal processor (DSP)] efforts that were aimed at snapping into supercomputing systems, this one is not a 32-bit part, but is capable of supporting 64-bit and further, it can also support both single (as others do) and double-precision. As seen below, the performance for both single and double precision is worth remarking upon (around 2.4 single, 4.8 double teraflops for one card) in a rather tiny power envelope. It will support high bandwidth memory as well as PCIe 3.0. In other words, it gives GPUs and Xeon Phi a run for the money—but the big question has far less to do with hardware capability and more to do with how the team at NUDT will be able to build out the required software stack to support applications that can gobble millions of cores on what is already by far the most core-dense machine on the planet.
According to VR World and HPCwire, the U.S. government has blacklisted "high technology" shipments to the National Supercomputing Center Changsha (NSCC-CS), National Supercomputing Center Guangzhou (NSCC-GZ), National Supercomputing Center Tianjin (NSCC-TJ), and the National University of Defense Technology (NUDT) in China. This effectively means that these major supercomputing facilities can no longer purchase Intel Xeon chips. Tianhe-2, the world's fastest supercomputer since June 2013 according to Top500, is located at the NUDT in Guangzhou and uses a total of 32,000 Intel Xeon and 48,000 Xeon Phi chips.
The main claim of the Bureau of Industry and Security's End-User Review Committee (ERC) is that NUDT, which used US-manufactured parts to produce the Tianhe-1A and Tianhe-2 supercomputers located at the National Supercomputing Centers in Changsha, Guangzhou, and Tianjin, is believed to be engaged in activities related to nuclear explosives.
The U.S. also uses supercomputers for nuclear weapons research.
The news coincides with the Intel Developer Forum 2015 in Shenzhen, China, at which the company announced new Braswell, SoFIA, and Cherry Trail chips, among other products. VR World speculates that the move could cost Intel $1 billion on lost Broadwell-EP Xeon E5v4 sales and accelerate the development of homegrown Chinese processors.
The TOP500 List of the world's fastest supercomputers for June 2015 has been released. China's Tianhe-2 remains the leader with 33.86 petaflops on the LINPACK benchmark. It has topped the list since June 2013. The only new supercomputer in the top 10 is the Shaheen II in Saudi Arabia, a 5.536 PFlop/s Cray XC40 system using 196,608 Intel Xeon E5-2698v3 cores.
The Platform has an analysis of the results. Although performance growth is slowing, pre-exascale supercomputers (100+ petaflops) can be expected within the next two to three years. The U.S. Department of Energy's Aurora supercomputer will deliver 180 petaflops of performance in 2018. Around the same time, the Summit supercomputer is expected to reach 150-300 petaflops while Sierra will reach 100+ petaflops. ~1 exaflop supercomputers are expected to appear around 2018-2022.
The June 2015 Green500 list ranking supercomputers by megaflops per watt will be available sometime later in the month. Here is the November 2014 Green500 list. The Piz Daint supercomputer appears within the top 10 on both lists.
In the name of national security, China is restricting export of certain drones and computers:
From August 15, manufacturers of certain powerful drones and computers will have to give technical details to the authorities to obtain a licence prior to export, Xinhua news agency says.
[...] In the first five months of 2015, China exported some 160,000 civilian drones, a jump of 70 per cent year-on-year, worth more than $120 million, the official China Daily newspaper reported in July.
[...] The tightening of regulations comes two weeks after an incident in disputed Kashmir in which the Pakistani army claimed to have shot down an Indian "spy drone", reportedly Chinese-made.
China is also likely tightening controls on exports of powerful computers as it looks to maintain its edge in the global supercomputer battle long dominated by US-Japanese rivalry.
Starting August 15th, drone and supercomputer manufacturers will have to present technical details to the authorities in order to get a license to export.
New information has emerged about China's exascale plans, which are a part of China's 13th five-year plan for 2016-2020. Despite U.S.-imposed export restrictions on processors, two 100 petaflops systems will be launched sometime during 2016, possibly as soon as the 2016 International Supercomputing Conference in June. One of these systems will be an upgrade to Tianhe-2, and both may utilize homegrown accelerators.
At least one exascale prototype system will be built prior to a 1 exaflops system:
The exascale prototype will be about 512 nodes, offering 5-10 teraflops-per-node, 10-20 Gflops/watt, point to point bandwidth greater than 200 Gbps. MPI latency should be less than 1.5 µs, said [Beihang University Professor Depei] Qian. Development will also include system software and three typical applications that will be used to verify effectiveness. From there, work will begin on an efficient computing node and a scheme for high-performance processor/accelerator design.
"Based on those key technology developments, we will finally build the exascale system," said Qian. "Our goal is not so ambitious – it is to have exaflops in peak. We are looking for a LINPACK efficiency of greater than 60 percent. Memory is rather limited, about 10 petabytes, with exabyte levels of storage. We don't think we can reach the 20 megawatts system goal in less than five years so our goal is about 35 megawatts for the system; that means 30 Gflops/watt energy efficiency. The expected interconnect performance is greater than 500 Gbps."
Zhaoxin, a joint venture between Via Technologies and the Chinese government, this week for the first time displayed its upcoming x86-compatible CPU, the KaiXian KX-6000. The SoC features eight cores running at 3 GHz and increases performance over its predecessor by at least 50%.
The KaiXian KX-6000 is a successor to the KX-5000 CPU launched earlier this year. Both chips integrate eight-core x86-64 cores with 8 MB of L2 cache, a DirectX 11.1-capable iGPU with an up-to-date display controller, a dual-channel DDR4-3200 memory controller, contemporary I/O interfaces (PCIe, SATA, USB, etc), and so on. The key differences between the KaiXian KX-5000 and the KaiXian KX-6000 are frequencies and manufacturing technology: the former is produced using TSMC's 28 nm fabrication process and runs at up to 2 GHz, whereas the latter is made using TSMC's 16 nm technology and operates at up to 3 GHz. Zhaoxin claims that the Kaixian KX-6000 offers compute performance comparable to that of Intel's 7th Generation Core i5 processor, which is a quad-core non-Hyper-Threaded CPU. Obviously, performance claims like that have to be verified, yet a 50% performance bump over the direct predecessor already seems beefy enough.
Related: Russia Plans to Dump Some American CPUs for Homegrown Technology
Russian Homegrown Elbrus-4C CPU Released
U.S. Export Restrictions Lead to Chinese Homegrown Supercomputing Chips
Linux-Based, MIPS-Powered Russian All-in-One PC Launched
China Dominates TOP500 List, Leads With New 93 Petaflops Supercomputer
Chinese Company Produces Chips Closely Based on AMD's Zen Microarchitecture