[REVISION HISTORY]

Alphabet/Tesla AI spending and EU AI gigafactory race

Updated 49 times since CLSTR started tracking revisions of this situation.

What changed

2026-09-05 11:21 UTC → 2026-09-09 05:32 UTC · added ~~removed~~

Alphabet is expanding its footprint in artificial intelligence and cloud computing. Following reports of Q2 2026 revenue growth driven by Google Cloud, the company has begun directly monetizing its Tensor Processing Units (TPUs) by shipping these custom chips to third-party data centers to compete with Nvidia. This move coincides with an 82% year-over-year revenue increase in its cloud computing division. To support this ecosystem, Alphabet maintains a partnership with Broadcom through 2031 and has entered an expanded custom-silicon relationship with Marvell Technology for TPU components. Additionally, Berkshire Hathaway has significantly increased its position in Alphabet, while other institutional investors, such as Elyxium Wealth LLC, have also adjusted their holdings. In its autonomous driving division, Waymo has received approval to operate thousands of robotaxis in Nevada and has expanded service areas in Los Angeles, Phoenix, and San Francisco. Waymo is also developing custom AI chips to reduce reliance on external providers such as Nvidia and AMD. Tesla reported Q2 revenue of $28.2 bn, a 26% YoY increase, but adjusted EPS fell to $0.33 and operating margin slipped to 1.4%. The firm signaled AI-related capex exceeding $25 bn for the year. The European Commission is advancing its InvestAI strategy to establish up to seven AI gigafactories to reduce dependence on U.S. and Chinese infrastructure. To date, 18 of the 27 EU member states have committed to the project. ~~Within this framework,~~ Poland and the Czech Republic have entered a formal partnership to develop these EU-funded facilities, aiming to establish a leading computing hub in Central Europe. Through a memorandum of understanding signed at the CEE AI Summit in Prague, Warsaw has committed 400 million zlotys (approximately 93 million euros) to support infrastructure intended to house at least 75,000 AI accelerators. The two nations currently prefer a model of two interconnected facilities.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.