I love to hate on google, but I suspect this is strategic enough that they wont kill it.

Like graviton at AWS its as much of a negotiation tool as it is a technical solution, letting them push harder with NVIDIA on pricing because they have a backup option.

Google has done stuff primarily for negotiation purposes (e.g. POWER9 chips) but TPU ain't one. It's not a backup option or presumed "inferior solution" to NVIDIA. Their entire ecosystem is TPU-first.

Was Gemini 2.5 trained on TPUs thought? I seem to be struggling to find that information. Wouldn't they want to mention it in every press release?

Pretty sure the answer is yes. I have no direct knowledge of the matter for Gemini 2.5, but in general TPUs were widely used for training at Google. Even Apple used them to train their Apple Intelligence models. It’s not some esoteric thing to train on TPU; I would consider using GPU for that inside Google esoteric.

P.S. I found an on-the-record statement re Gemini 1.0 on TPU:

"We trained Gemini 1.0 at scale on our AI-optimized infrastructure using Google’s in-house designed Tensor Processing Units (TPUs) v4 and v5e. And we designed it to be our most reliable and scalable model to train, and our most efficient to serve."

Google's been doing custom ML accelerators for 10 years now, and (depending on how much you're willing to stretch the definition) has been doing them in consumer hardware for soon to be five years (the Google Tensor chips in pixel phones).