Federico Cassano
@ellev3n11
understanding the universe @spacexai - trained composer @cursor_ai
354 Following 7.8K Followers
Compute is the bottleneck
@elonmusk People don’t understand how much compute the world will need
would be cool to visit mars sometime
Cursor is now part of @SpaceX.
Today, we have officially closed our acquisition. We will join the @SpaceXAI team to help make Grok the world's most useful AI and improve Grok Build, Grok Bot, Grok API, Cursor, and more.
SpaceX has built some of the most inspiring and impressive technology in the world, and we’re grateful for the opportunity to become part of such a special company. Onwards.
Show more
model factory is coming together
Introducing Grok 4.6.
It delivers frontier intelligence and is a significant improvement over Grok 4.5 at the same price.
We are integrating Mixture-of-kittens into Nvidia NeMo-Automodel. 7%-50% e2e bf16 training speed gain on gb200 from ep4 to ep64 settings. mxfp8 integration coming soon!
Checkout our PR for experiment details:
And Zhihu tutorial:
Show more
most cracked intern
We’ve open-sourced the MoE megakernel we use to train models on NVL72s. A few of my favorite details: bf16 and mxfp8 support, pull-based dispatch, configurable overlap granularity, full determinism, and no CPU-GPU syncs. Read more in our blog and code!
Show more
a kernel adequate for le chaton fat
We're open-sourcing Mixture-of-Kittens (MoK), our MoE training megakernel for NVL72s.
It fuses all Mixture-of-Experts communication and computation into a single, fully deterministic kernel, and runs up to 2.37x faster than the strongest public baselines.
Show more
just 10 dollars shy of 420
it's just interpolating text guys
hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final
((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)
Show more
Excited to release Grok 4.5 with @SpaceXAI.
It's an Opus-class model that's fast and low cost. It's a significant step up over any model we've developed so far, including Composer 2.5, and has become the daily driver for many on our team.
First of many releases. More soon.
Show more
Personal update: starting a new team chasing longer-term experimental bets. If you want to explore the next scaling paradigm with serious compute behind you and an exceptionally talent-dense team, let's talk.
Show more
proud to call the greatest country on earth my home. happy america day!
cuda frameworks that JIT and autotune are so painful. i don't want to have magic happen during the start of training. i want to precompile my kernels, test them once, and be happy.
Personal update, I'm starting a new role at Cursor!
I'm moving into ML, working on training Composer. I'll be researching how to improve model behavior and personality.
guys they sent me to the ranch
they are clearing my kv cache tomorrow
