AMD EPYC Processors Power Google Cloud TPU 8tx for Confidential AI Computing

AMD EPYC Secures Role in Google's Confidential AI Push

Google Cloud unveiled its TPU 8tx system on Monday, November 9, 2026, powered by AMD's 5th Gen EPYC processors in a strategic partnership that brings hardware-enforced confidential computing to cloud-based AI infrastructure for the first time1

1

. The collaboration pairs AMD EPYC CPU performance, AMD Infinity Guard security capabilities, and x86 software compatibility with Google Cloud TPU AI acceleration to help enterprises deploy inference, fine-tuning, and reinforcement learning workloads2

2

. This expanded AMD and Google collaboration addresses a critical requirement for production AI: protecting sensitive data while it remains in active use by processing systems.

Why Google Chose AMD Over Its Own Silicon

The processor selection stands out because Google already designs and uses its own Axion processors for host infrastructure. When Google introduced TPU 8t and TPU 8i systems in April 2026, both ran on Axion chips, replacing the x86 processors used in earlier generations3

3

. TPU 8tx is built on the same silicon as TPU 8t, yet Google specifically chose AMD's fifth-generation EPYC processors for this confidential computing variant. Google built the system for regulated enterprises and governments running frontier or proprietary models, noting that highly regulated customers are requesting specialized encryption capabilities for unique workloads3

3

. AMD's advantage lies in its established security track record within Google's infrastructure—AMD SEV technologies already help protect workloads across Google Cloud's Confidential VMs, Confidential GKE, Confidential Space, Confidential Managed Service for Apache Spark, and Confidential Dataflow4

4

.

Hardware-Level Data Protection for Sensitive AI Workloads

In TPU 8tx systems, 5th Gen AMD EPYC processors manage workload orchestration, data preparation, and high-throughput communication with TPU accelerators5

5

. The system utilizes AMD Secure Encrypted Virtualization (SEV) capabilities, which extend confidential computing protection beyond the CPU to connected accelerators, helping protect sensitive data as it moves among components2

2

. The architecture verifies an accelerator's identity and security configuration before granting direct access to protected memory, ensuring customer prompts, model weights, and intermediate data remain encrypted in CPU memory and in transit among the host and accelerators, with hardware isolation during accelerator execution4

4

. Mark Lohmeyer, vice president and general manager for AI and Computing Infrastructure at Google Cloud, explained that teams moving AI agents into production need confidence that their data and intellectual property stay private, even from the underlying infrastructure, noting that customers get silicon-level data protection alongside the low-latency performance needed for modern reasoning models2

2

.

AMD's Dual Revenue Path in AI Infrastructure

AMD competes with Nvidia in the AI chip market through its Instinct GPUs, but the company also sells processors that run AI systems built by other companies, creating a dual revenue opportunity1

1

. Dan McNamara, senior vice president and general manager of Compute and Enterprise AI at AMD, stated that the most capable AI systems need a CPU foundation that can match the demands of the accelerators they host, emphasizing that AMD has built leadership capabilities in performance, memory bandwidth, and I/O that keep accelerated systems productive4

4

. AMD EPYC processors serve as the host CPU of choice for more than 10 leading hyperscalers and frontier AI labs, spanning GPUs and custom AI accelerator ASICs2

2

. This breadth gives customers a proven CPU platform across accelerator choices, supported by the established x86 software ecosystem, meaning AMD benefits even when customers select Nvidia chips or Google's own technology over its competing GPU products1

1

.

Market Response and Future Implications

AMD shares climbed 13.26% on Monday, November 9, 2026, closing at $508.58, recovering from a volatile week that saw a 17.26% drop on November 5 followed by an 11.04% rebound on November 63

3

. The rally was amplified by Oppenheimer analyst Rick Schafer upgrading AMD to Outperform from Perform with a $600 price target1

1

. As of November 6, 2026, 53 analysts tracked by StockAnalysis rated AMD a consensus Strong Buy with an average price target of $658.533

3

. Bloomberg Intelligence estimates Google's TPU-related annual sales could grow from $12 billion in 2026 to more than $135 billion by 2028, though most new TPU systems use Google's own Axion CPUs, meaning AMD won't necessarily benefit from all that growth1

1

. Google expects the new TPU 8tx systems to become available in 20271

1

. The development signals that data privacy and compliance are becoming critical factors in AI chip selection, potentially favoring suppliers with established security records that take years to build3

3

.