China beefs up its HPC training

Nvidia also opens up the Cuda platform

China can use its power of government control to bring major changes quickly, and it is moving to expand parallel programming training to help its supercomputing efforts -- and possibly its outsourcing industry, too.

The Chinese Ministry of Education said Tuesday that it will offer Nvidia's Cuda (Compute Unified Device Architecture) at 200 universities, and will be training up to 20,000 students annually on it.

Cuda is a parallel computing architecture that allows developers, writing in common languages, to adapt their program for parallel environments.

It was developed by Nvidia. But the company, in a separate announcement made at a conference in China Tuesday, said it was opening up the Cuda platform by releasing the compiler source code.

Sumit Gupta, a senior marketing manager at Nvidia, said China's decision to expand Cuda training is a recognition "of China's continued investment in parallel programming and high performance computing," as well as the use of GPUs as an easy way to teach parallel programming. That's due in part to the wide availability of GPUs on everything from laptops to servers, he said.

Cuda training is now available in three or four universities today in China, but is taught in nearly 500 universities worldwide, said Gupta. Release of the source code will allow support for more programming languages, and even alternative processor architectures, such as AMD.

Cuda supports C, C++ and Fortran.

Nathan Brookwood, principal analyst at Insight 64, said the parallel programming training is needed.

"Most people don't get trained in how to write software that uses those kinds of parallel approaches, and, to a certain extent, the tools to write those parallel approaches have been relatively obscure," said Brookwood.

Brookwood said it's to Nvidia's credit that it has delivered tools, such as Cuda, "that let people express their algorithms in a more parallel manner and therefore have a better fit with GPUs."

"There are an awful lot of very smart people in China and now they are going to be educated in how to take advantage of these tools," said Brookwood.

Having this training will also help China develop its outsourcing industry , said Brookwood.

In announcing the plan to expand Cuda training in China, Li Maoguo, division director of the higher education bureau in the Chinese Ministry of Education, said in a statement that "in the era of knowledge economy, technology education is critical in building students' competitiveness in a global environment," he said.

Patrick Thibodeau covers cloud computing and enterprise applications, outsourcing, government IT policies, data centers and IT workforce issues for Computerworld. Follow Patrick on Twitter at @DCgov or subscribe to Patrick's RSS feed . His e-mail address is pthibodeau@computerworld.com .

Read more about mainframes and supercomputers in Computerworld's Mainframes and Supercomputers Topic Center.

Join the newsletter!

Or

Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

Tags processorsnvidiaComponentsApp DevelopmentMainframes and Supercomputers

More about AMDInsight 64NvidiaTopic

Show Comments
[]