Who is on this subjectordered by the date of their first quote here
3 of 3 lanes rest on fewer than 5 quotes and are marked thin. Offsets are days from the middle first-quote date, 23 Dec 2024 — a date, and nothing else. It is not a claim about who reached a view first.
-
Bill Gurley
2
thin
23 Dec 2024 – 23 Dec 2024
+0 d
- Dylan Patel 2 thin 23 Dec 2024 – 30 Jun 2026 +0 d
-
Gavin Baker
3
thin
15 May 2026 – 15 May 2026
+508 d
The chronologysourced and dated, oldest first
-
-
Dylan Patel
Patel dubs Amazon's Trainium chip the Amazon Basics TPU due to less efficient but cost-effective design.
“Yeah. So so funnily enough, Amazon's chip is the Amazon I I call it the Amazon's basics TPU. Right?”
-
Bill Gurley
Patel says Amazon and Anthropic building 400,000 chip Trainium supercomputer proves they believe scaling continues.
“And and because of this, right, Amazon and Anthropic have decided to, you know, make a 400,000 Trainium server supercomputer. Right? 400,000 chips. Right? Going back to the whole scaling law is dead.”
-
Bill Gurley
Patel reveals Amazon and Anthropic are building a 400,000 Trainium chip supercomputer as evidence scaling continues.
“Amazon and Anthropic have decided to, you know, make a 400,000 Trainium server supercomputer. Right? 400,000 chips. Right? Going back to the whole scaling law is dead.”
-
-
-
“Tranium, by far. Tranium is going to be to 2026, especially in the second half of this year when Tranium three really ramps, as TPUs were to twenty twenty five.”
-
Gavin Baker
Baker argues Trainium is most underestimated because frontier mixture-of-expert models require switched scale-up networks for inference.
“And so Trainium is for sure the most underestimated, not only because of those design choices, but because the all of these frontier models are what are called mixture of expert models.”
-
Gavin Baker
Baker states only NVIDIA and Amazon Trainium have functioning switched scale-up networks for inference today.
“And the only two functioning switched scale up networks in the world today are the ones that power NVIDIA GPUs and Amazon's Trainiums.”
-
Dylan Patel
Patel reports Trainium rents for under $10 billion per gigawatt while GPUs cost $12-13 billion.
“Tranium sells at sub $10,000,000,000 per gigawatt rental rate to Anthropic and to OpenAI. GPUs, at least before the craziness of the last six months, usually went around 12 to $13,000,000,000 per gigawatt.”
-