Given the punishing stock performance of large cap software, is it realistic to think the large caps and big spenders to take a breather, in a somewhat coordinated outlook - I mean isn’t that really the only way to boost up the stock and restore morale at this point?
What’s interesting to see is a) all three had recently announced progress on their own inference chip development and capabilities (MSFT is moving toward Maia 200 for Azure inference; AMZN is already the most production-mature with Trainium powering real workloads. META is iterating fast on MTIA with newer generations aimed at recommendation and GenAI inference) and b) all are considering cheaper and smaller models for deployments (Amazon Nova
, LLama MoE)