Look, I mean, the 8th-generation TPU lineup — a training chip at three times Ironwood's processing power, an inference chip at 80% better performance per dollar — sits in a multi-architecture offering we frame as the industry's widest: Axion now hosts TPUs replacing x86, and we'll be among the first with NVIDIA Vera Rubin. Broadcom remains our TPU partner. The shift: delivering TPUs to select customers for their own data centers — Thinking Machines Lab, Hudson River Trading, Boston Dynamics. Revenue there is a small percent in 2026, majority in 2027, lumpy with ship timing. Very early innings.