IBM Cloud got an EPYC power up this week with the announcement that the cloud provider was adding AMD’s second-generation server chips to its portfolio.

At the heart of the new offering is AMD’s 96-core EPYC 7642 processors, which boast a maximum clock speed of 3.3 GHz and support for up to four terabytes of memory per socket.

The processors will power IBM Cloud’s latest bare-metal offerings, which the cloud provider claims are well suited for data analytics, electronic design automation, artificial intelligence (AI), virtualized and other containerized workloads. The servers are available in IBM data centers in North America, Europe, and the Asia-Pacific region, and can be ordered via the IBM cloud portal, API, or command-line interface (CLI) on a monthly "pay-as-you-use" model.

AMD has celebrated several cloud wins since launching its second-generation EPYC chips in mid-2019.

Google Cloud Platform (GCP) launched its largest general-purpose virtual machines (VMs) based on AMD’s EPYC chips back in February. The N2D VMs can be configured with up to 224 virtual CPUs (vCPUs), and are targeted at general-purpose and high-performance computing (HPC) workloads requiring high-memory bandwidth.

The chips have received wide praise from analysts, including Patrick Moorhead, president and principal analyst at Moor Insights and Strategy, who called AMD’s latest server chips a “big step forward.”

But while IBM represents a win for AMD’s growing data center efforts, the chipmaker has faced increased competition from long-time rivals. Intel slashed the pricing of its second-generation Scalable chips earlier this year, and Arm-based offerings from the likes of Marvell and Ampere have steadily won cloud contracts from Microsoft Azure and Oracle Cloud respectively.