Will Google sign a verifiable long-term TPU deal with an outside customer? — Contested · resolves 2026-10-31

The debate: our existing view holds that whether Google's TPUs take share hinges on supplying outside customers, not on price-performance, and on September 7 SemiAnalysis published a reading that TPUs deliver up to 50% better performance per dollar.

Why it matters: cloud providers' in-house chips have mainly served their own inference, so a verifiable long-term contract with a named outside customer would move a non-Nvidia option from internal use into the open market.

Who is on each side: we bet that externalization is the key and that strong price-performance numbers alone do not threaten share; the opposite reading is that a price-performance lead by itself moves share.

Our existing view is that whether TPUs take share hinges on externalization, not on price-performance. This card is a yes-or-no question, not a bet on either side: by 2026-10-31, will Google sign a verifiable long-term TPU contract with a named outside customer, with capacity or term that can be checked? If so, the externalization switch has started to turn, the load-bearing leg of our view holds, and the next question is whether share follows; if not, the switch has not moved, and however good the price-performance readings are, they are not evidence of a threat to share.

Sources still being linked up — the node(s) behind this call don't have a clickable original link resolved yet. Logged for follow-up.

Only Google or Alphabet's official press releases and SEC filings, and the customer's own official press releases and statutory filings, count. Any of these naming an outside customer's long-term TPU contract with capacity or term before the due date answers yes; no such announcement or filing answers no. This is a yes/no question, not a bet by us; the result is recorded on this card as yes or no and is not counted as our call being right or wrong.

What would overturn it

If verifiable readings of share movement (third-party market measurements) appear while no outside long-term contract can be found, attributing the bottleneck to externalization is wrong and the result has to be recorded as driven by price-performance.