Arrowz

Fascinating - $GOOG TPU has the best architecture for AI compute, and “there aren’t enough TPUs,” $AVGO $ANT “Today I would argue that probably the best architecture is the Google TPU architecture, because it is literally thought of as a single system. And I think an NVIDIA-style architecture is migrating in the direction of where the TPU architecture is, but it still has a few pieces that you can kind of swap out. And then you have everybody else — like, you know, I'm going to take a Cerebras chip and throw an Arista switch on top of it, and all that good stuff. At least in today's world, those systems have inefficiencies because they're not constructed as a system — they're more of a patchwork. I think the industry will evolve, but today, people that can think of the whole thing as a single system and optimize the architecture through that have a competitive advantage. If you look at how Google does this, with these toroidal pass-through optical MEMS switches — it's literally like the whole data center is the minimum unit. But my sense is that's where NVIDIA wants to go also. Actually, the two dominant providers are all going to that systems approach, versus — I mean, even the smaller startups are heading that direction. Right now Jensen's king of the hill, incredibly dominant, right? Google TPU very successful, currently mostly sold by Google — it's expected to change. But not enough TPUs.”