An update is currently rolling out for Canonical’s Ubuntu which supports the NVIDIA DGX-2 AI system, the world’s first 2 petaFLOPS system that combines 16 interconnected GPUs into a single computing unit.
The NVIDIA DGX-2 AI system already supports the Red Hat Enterprise Linux operating system, but with this latest update, users will be able to opt for Ubuntu. Canonical, the developer of Ubuntu, has often touted the operating system’s abilities when it comes to AI and machine learning applications, especially thanks to its support for Kubernetes. This update makes it an even better choice for those wishing to deploy AI functionality.
“Ubuntu is the preferred AI and ML platform for developers and the No. 1 operating system for Kubernetes deployments on-premises and in the public cloud. This collaboration with NVIDIA enables enterprises to enhance their developers’ productivity and incorporate AI more quickly through development stages to production,” said Stephan Fabel, Director of Product at Canonical.
“The combination of DGX-2 and Ubuntu helps organisations to realise the vast potential of AI, allowing them to develop and deploy models at scale via the world’s most powerful AI system.”
For NVIDIA’s part, it wants the DGX-2 AI system to be as widely available as possible. The company claims that the system is capable of handling extremely demanding AI applications, and blows traditional processing power out of the water. In fact, according to the company, its DGX-2 AI system deliver 195x faster deep learning training when compared to a traditional CPU – unsurprising considering its use of GPUs, which are widely regarded as the best tools for AI applications.
Support for NVIDIA’s DGX-2 AI system is available as part of the Ubuntu 18.04 LTS update, which is rolling out now.