New Nvidia ‘supercomputer’ could reduce data centre footprint

Nvidia CEO Jenson Huang making a presentation on stage

Nvidia has been a big name in graphics processing units for many years, although in the past, GPUs were mainly of interest to people who played memory-hungry computer games. 

Gamers are still a big market for Nvidia, but lately, it’s seen much more demand for its GPUs from cryptocurrency miners.

Cryptocurrency mining, like computer games, requires computers with massive processing power, and Nvidia GPUs are particularly sought-after.

Now, Nvidia has used some of its GPUs to produce what is described as a “supercomputer in a box”, called the “DGX-2” system.

If widely adopted, Nvidia’s new server could reduce the global data centre footprint.

DataCentreKnowledge.com says Nvidia’s mini-supercomputer “can replace hundreds of CPU nodes”.

Data Centre Knowledge reports that the DGX-2 is priced at $399,000 and that Nvidia describes the machine as “the world’s largest GPU”.

Nvidia itself says the DGX-2 is particularly suitable for deep learning applications, which is a branch of artificial intelligence, like machine learning, only deeper.

The company says the DGX-2 is the first single server capable of delivering two petaflops of computational power.

It also claims that the DGX-2 has the deep learning processing power of 300 servers occupying 15 racks of data center space, while being 60 times smaller and 18 times more power-efficient.

Jensen Huang, Nvidia founder and CEO, says: “The extraordinary advances of deep learning only hint at what is still to come.

“Many of these advances stand on Nvidia’s deep learning platform, which has quickly become the world’s standard.

“We are dramatically enhancing our platform’s performance at a pace far exceeding Moore’s law, enabling breakthroughs that will help revolutionise healthcare, transportation, science exploration and countless other areas.”