Google has reportedly chosen Intel's contract-manufacturing arm, Intel Foundry, to produce more than three million of its in-house AI chips, known as TPUs (tensor processing units), with production targeted for 2028. The news was first reported by The Information and picked up widely, including by CNBC, which framed it as Google expanding its partnership with Intel for AI chips.
According to TweakTown and games.gg, the order comes as TSMC — the Taiwanese giant that makes most of the world's advanced chips — faces a capacity crunch, pushing Google to seek additional manufacturing. Yahoo Finance reported the deal represents Intel Foundry's largest external manufacturing commitment to date. Wccftech reported that Google will use Intel's EMIB advanced packaging technology for its next-generation TPU.
The story is also about a possible turnaround for Intel, which has struggled to win major foundry customers. Several outlets, including Seeking Alpha and the-decoder.com, reported that both Google and Nvidia are exploring Intel as a backup chipmaker to TSMC, with Nvidia said to be testing Intel's 18A manufacturing process for a future chip codenamed Feynman.
Investors reacted fast. According to The Globe and Mail, Intel shares surged on June 8 following the report, with Gotrade citing an 11% jump and 24/7 Wall St. citing a 12% gain, helped along by a separate Hitachi pact. Bloomberg, via the Mercury News, noted it was Intel's biggest single-day rise in a month. Some of those gains later wobbled amid a broader chip selloff, per TechStock².
Why it matters: if confirmed, the deal would mark a rare validation of Intel's costly bet on becoming a contract chipmaker — and signal that the world's reliance on a single Taiwanese supplier for cutting-edge AI silicon is beginning to crack.