AMD EPYC CPUs Power Google Cloud's First TPU With Hardware-Enforced Confidential Computing

AMD (AMD) said on Monday that its 5th Gen EPYC server processors power the host infrastructure for Google Cloud TPU 8tx, the first cloud Tensor Processing Unit system to deliver hardware-enforced confidential computing.

The announcement, made with Alphabet (GOOGL) unit Google Cloud, expands a collaboration that pairs AMD EPYC CPU performance and Infinity Guard security features with Google Cloud's TPU AI acceleration. AMD said the combination helps enterprises run inference, fine-tuning and reinforcement learning workloads while protecting sensitive data.

Dan McNamara, senior vice president and general manager of Compute and Enterprise AI at AMD, said the most capable AI systems need a CPU foundation that can match the demands of the accelerators they host, and that AMD EPYC has built leadership in performance, memory bandwidth and input/output across generations, backed by engineering work with leading cloud providers.

Mark Lohmeyer, vice president and general manager for AI and Computing Infrastructure at Google Cloud, said the pairing brings hardware-enforced confidential computing to cloud TPUs for the first time, giving customers silicon-level data protection alongside the low-latency performance needed for modern reasoning models.

In TPU 8tx systems, the 5th Gen EPYC processors manage workload orchestration, data preparation and high-throughput communication with the TPU accelerators. The system uses AMD Secure Encrypted Virtualization capabilities, which together with the TPU extend confidential computing protection beyond the CPU to connected accelerators, keeping customer prompts, model weights and intermediate data encrypted in CPU memory and in transit.

AMD said its EPYC processors are the host CPU of choice across more than 10 leading hyperscalers and frontier AI labs, spanning systems built with AMD Instinct GPUs, Nvidia GPUs and custom accelerator silicon. The company also said EPYC underpins Google Cloud's broader confidential computing portfolio, including Confidential VMs, Confidential GKE and Confidential Dataflow, with TPU 8tx extending that work to confidential AI acceleration.

The companies did not disclose financial terms or a deployment timeline for TPU 8tx.