Piotr Mazurek (in SF 🇺🇸 at last 🤗)(@tugot17):the reason you will have to care about heterogenous compute is that as we get closer to RSI you will be just priced out of accessing the leading edge chips by the labs; there is no upper cap on the demand here

Piotr Mazurek (in SF 🇺🇸 at last 🤗)

@tugot17

inteligence involution @liquidai; from first principles ex @Aleph__Alpha (acq. by @cohere)

参加 September 2021

966 フォロー中 4.7K ファン

Piotr Mazurek (in SF 🇺🇸 at last 🤗)@tugot17

2026.09.25 16:45

the reason you will have to care about heterogenous compute is that as we get closer to RSI you will be just priced out of accessing the leading edge chips by the labs; there is no upper cap on the demand here

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex

2026.09.24 13:26

$1B makes sense. 70% on training is logical > DeepSeek is using NVIDIA gaming GPUs to run inference on smaller models this is news to me but tbh is an obvious move. Imagine how much work you could speed up just with Jev-style small models, or a custom-made Qwen 35B… I want it