8 Comments

Wonderful article Richard!

Expand full comment

Thank you.

Expand full comment

Smart stuff. Appreciate your explaining. Notable, that AWS margins expanded over the past few years, even with all of the $ spent with NVDA, and scaling up overall AI intiatiives.

Recognize Nvidia integration with CUDA, still hard to see NVDA being able to maintain the same margins as Microsoft, a company that has barely any manufacturing costs.

Expand full comment

Very good analysis. Spending more computing time to process to serve a request is the next logical step now that we have reach an upper limit in data set. It makes total sense. Inference chips specialization may lead to a whole new innovation. I also share a similar portfolio with Amazon Google TSM. Whether google is weaken by this is a concern. The commodization of LLM model weskens openAI which is good but it also lower the cost of being served by LLM thus strenghten the case of using LLM versus search. What do you think?

Expand full comment

The faster the progress of LLMs, the faster the cannibalization of Google Search will be, so yes, the speed up in availability (costs) of serving these models is a negative for the Search business, but good for their cloud business, waymo, etc.

Expand full comment

Very interesting read, thank you for sharing!

Expand full comment

Excellent writing Richard.

Expand full comment

thanks Nick!

Expand full comment