Intel expands Google partnership
Intel and Google have broadened a multi‑year partnership to co‑develop AI datacentre infrastructure and custom chips, with wider use of Intel Xeon processors and infrastructure processing units across Google’s AI cloud. The collaboration emphasises orchestration and server plumbing rather than only accelerator performance. (simplywall.st)
Artificial intelligence data centers still need the chips that move data, schedule jobs, and keep servers talking to each other, not just the chips that do the math. Intel and Google said on April 9 they are expanding a multiyear partnership around that layer of cloud infrastructure. (intel.com) The deal keeps Intel Xeon processors in Google Cloud across artificial intelligence, inference, and general-purpose workloads, and extends joint work on custom infrastructure processing units, or IPUs. Intel said Google Cloud already uses Xeon 6 processors in its C4 and N4 instances. (intel.com) A central processing unit is the traffic manager in a server, while an infrastructure processing unit is a helper chip that takes over networking, storage, and security jobs. Intel said those IPUs are designed to free host processors for other work and make performance more predictable at hyperscale. (intel.com) Google and Intel did not disclose financial terms or a timeline for the expanded agreement. CNBC reported Google committed to use multiple generations of Intel chips in its artificial intelligence data centers. (cnbc.com) The announcement lands in a market where most attention has gone to Nvidia graphics processing units, the chips that train and run many large artificial intelligence models. Intel and Google are instead emphasizing the rest of the system: orchestration, data movement, and general-purpose compute that sits around those accelerators. (cnbc.com) (intel.com) That focus reflects how cloud operators build artificial intelligence clusters in practice. Intel said CPUs handle orchestration, data processing, and system-level performance, while Google’s Amin Vahdat said CPUs and infrastructure acceleration remain a cornerstone of systems used for training orchestration, inference, and deployment. (intel.com) Google has also been broadening its own chip strategy rather than relying on one supplier. CNBC noted Google has long used Intel processors, dating back to its early server buildout, even as the company has also developed in-house silicon such as Tensor Processing Units and Arm-based Axion central processing units. (cnbc.com) For Intel, the agreement offers a fresh cloud design win at a time when the company is trying to show its processors still matter in artificial intelligence spending. Intel shares rose nearly 5% on April 9, while Alphabet closed only marginally higher, according to CNBC. (cnbc.com) The immediate result is not a new chatbot chip or a consumer product launch. It is a bet that the less visible parts of an artificial intelligence data center — the traffic control, storage handling, and server coordination — will decide how efficiently those systems scale. (intel.com)
