Intel says multiple hyperscalers are sampling Habana’s accelerator for ML training.

The startup says its plug-and-play design can pack 1MW of data center inside 320 square feet.

Other Nvidia announcements at SC19 include a GPU-accelerated Arm server design and software that makes GPU clusters faster.

AI Anywhere models crunch 8 petaFLOPs and 13 petaFLOPs inside self-sufficient racks, thanks to proprietary cooling tech.

Alibaba Cloud joins the race for the cloud platform with the best AI processor.

InfoSight on SimpliVity uses predictive analytics to catch data center issues before they balloon.

One of the key engineers behind the AI chips in Google data centers now has his own AI chip company.

Cloud artificial intelligence presents some challenges that other cloud-based offerings do not.

Ascend’s Structured Data Lake uses AI to intelligently manage data storage across the data life cycle.

The Nvidia GPU-powered G4 compute instance was built to help organizations handle machine learning workloads in the cloud more efficiently.