Nvidia focuses on GPUs with Chinese researchers

Graphics specialist Nvidia has teamed up with the Beijing Genomics Institute to develop research applications that make use of GPUs.

The link-up, announced on Thursday, will yield the creation of a Joint Innovation Lab to tweak bioinformatics applications for GPU acceleration.

"Through our combined efforts, we can advance life science research and improve response time for virus outbreaks and health issues," Nvidia wrote in a blog on Thursday.

The team from the Beijing Genomics Institute (BGI) has already ported two programs to Nvidia GPUs. According to Nvidia, tests showed the GPUs yielded a major performance increase.

GPUs are structured differently to CPUs. A typical GPU is composed of many small processors for computing relatively simple tasks in parallel. GPUs tend to be better at graphical tasks than CPUs, though they can also be used for rendering, simulation and other activities of interest to scientific research applications.

China's most powerful computer, the Tianhe-1a, relies on over 7,000 Nvidia Tesla GPUs and 14,000 Xeon CPUs to make its computations. According to the Top500 rankings, the Tianhe-1a is the second most powerful supercomputer in the world, behind Japan's K Computer.