HPC might rises in the Far East
Like the space race. But with petaflops
As spring turns into summer, we get – like clockwork – a new Top 500 list. While there’s plenty of analysis yet to be done, what’s getting lots of press is how the Chinese National Supercomputing Center has captured the number two slot on the list with their 1.27 petaflop (sustained), 4,640 GPU monster box. This system is noteworthy not only from a performance standpoint, but also because it relies so heavily on NVIDIA GPUs, further confirming a trend toward hybrid CPU/GPU computing.
Much of the attention will be focused on the big move that China Inc. has made on the list. There are two China-based systems in the top ten, and the Chinese own 24 systems in the Top 500. In total performance, this puts China behind only the US as a supercomputing powerhouse. There is speculation that by this time next year, China is going to be rolling out a new, all-China-designed supercomputer that might be the fastest in the world.
This is somewhat reminiscent of the space race, when the USSR and the US vied for dominance in the heavens. Reinforcing this thought are various conversations I’ve had with HPC gurus who increasingly see countries regarding HPC capacity as a national asset that can be used to improve economic competitiveness through improved material science research and product design. Essentially, HPC can be a national resource to help domestic industries dominate world markets. Of course, it’s also pretty good for doing economic research and weapons design too.
Is the specter of a committed communist government taking the lead in supercomputing something new to be scared of? Yeah, it might be. I (along with a lot of other people) believe that innovation springs from strong, basic research – and that strong, basic research is enabled by plentiful processing resources. ®