In the new episode I'm breaking down the engineering behind Colossus 2 - the world's largest AI data center: t.co/kTYDfgOL1y

I met the world's first synthetic human: t.co/Z3d7Ltyfm4

New episode is just dropped - don't miss this one😎 t.co/ybC4dLPlAw

💡Dojo’s Story Announced at AI Day in 2021, entered production in 2023. Over ~4 years, Tesla poured several billion dollars into it. They did some impressive stuff along the way. Most notably — innovating in packaging, cooling and SW.

⁉️Why it didn’t work out Building a datacenter-class AI supercomputer from scratch is one of the most capital- and talent-intensive bets in tech. The engineers capable of pulling this off are rare — and easily poached.

More and more of the progress is now coming from system-level engineering: advanced packaging, stacking memory on top of logic, and especially interconnect innovation.

Here is my breakdown of @elonmusk Terafab: t.co/JEvJi6Ipeq

New video covering @IBM new tech: youtu.be/6msVefMxPCA

Moore's Law gave us more transistors for less money. @Huawei 's Tau New Scaling Law gives more performance through stacking and tighter timing. But unlike Moore's Law, every gain makes chips more expensive.

There was also an interesting keynote from @ASMLcompany about future lithography systems and larger reticle sizes. You can really feel that the industry is shifting from “make the transistor smaller” to “how do we architect the system differently”.

@A_deLaLoyere @Tesla @SpaceX @xai thank you!

@gregtaniguchi thank you!

@abnabdullah5 Thank you!