Intel expands Google tie-up
Intel and Google have broadened a multi‑year partnership to co‑develop AI data‑centre infrastructure and custom chips, with a renewed emphasis on Intel Xeon processors and new infrastructure processing units. (simplywall.st) The deal highlights a push to compete on orchestration, cost and manageability as well as raw accelerator performance.
On April 9, 2026 Intel and Google announced an expanded multiyear partnership to co‑develop AI data‑centre infrastructure, committing to future Intel Xeon processors and custom infrastructure processing units. (intel.com) Infrastructure Processing Units are purpose‑built chips that offload networking, storage and security tasks from central processing units so servers can dedicate more cycles to AI training and inference. (intel.com) Google said it will deploy multiple generations of Intel Xeon processors in Google Cloud, including Intel’s Xeon 6 family used on C4 and N4 virtual machine instances for training coordination, inference and general workloads. (intel.com) The companies said they will expand co‑development of custom, application‑specific IPUs to improve utilization, energy efficiency and performance at hyperscale. (businesswire.com) Google and Intel framed the agreement as a push to lower total cost of ownership and simplify orchestration and manageability for large AI deployments, not just to increase raw accelerator throughput. (intel.com) The pact arrives as hyperscalers confront system‑level bottlenecks and as Intel races competitors such as Nvidia, AMD and emerging Arm server designs for server‑CPU and infrastructure roles. (siliconangle.com) Intel and Google first announced joint IPU work in October 2021 with the Mount Evans project and later brought co‑designed IPUs into Google Cloud C3 instances, laying the groundwork for today’s ASIC‑focused expansion. (datacenterdynamics.com) “Scaling AI requires more than accelerators,” Intel CEO Lip‑Bu Tan said, and Google’s Amin Vahdat added that “CPUs and infrastructure acceleration remain a cornerstone” as both firms begin wider deployment across Google Cloud’s fleet. (intel.com)
