Subha

math + cs @stanford

[Subha @subha_v1]

At @hackwithtrees, we built a project that retrofits self-driving into ANY car, powered by the Jetson Thor. Won 1st place on @nvidia’s edge AI track! 🥳

Went from organizing last year to competing this year and it honestly hit different 🥹 love this community 💚

Links below!

[Subha @subha_v1]

@Leik0w0 @kennethnym lol which harness do u use

[Subha @subha_v1]

@sophiecliffa8 @ScalingIntelLab @kkaitlyn111 Thanks for your input Soph! Chief artist is a #real #position at this lab. We use them to generate a controllable corpus of images with supervision. This is used to #fine-tune our #VLM, resulting in a model with improved compositional visual understanding in #anime.

[Subha @subha_v1]

im excited to join @ScalingIntelLab as a full time chief artist! 📈

[Subha @subha_v1]

@vincentzed_cuda inference larping lmao

[Subha @subha_v1]

@alexhuang06 @ScalingIntelLab @kkaitlyn111

[Subha @subha_v1]

@kkaitlyn111 @ScalingIntelLab

[Subha @subha_v1]

@isaacsun07 @ScalingIntelLab @kkaitlyn111

[Subha @subha_v1]

@kkaitlyn111 @VivianCaiIAm @alexhuang06 I concur, @alexhuang06 is one of the brightest minds I know

[Ajay Anubolu @kvcache_]

excited to join @togethercompute this fall as an inference research intern!

looking forward to improving efficient inference of open models

[Subha @subha_v1]

@stuart_sul wow congrats! are there any plans to port over these ideas to amd gpus w/ hipkittens? i've been working on this recently, although the communication-computation overlap methods may be quite different 😅

[Cursor @cursor_ai]

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.

[AI at AMD @AIatAMD]

Meet AMD Skills: A growing catalog of agent skills that helps AI coding assistants work more effectively with AMD hardware and software.

Compatible with coding agents including Cursor, Claude Code, OpenAI Codex, and Gemini CLI, AMD Skills provides the knowledge, scripts, and best practices developers need to build faster on AMD.

Why try it?
→ Install in seconds with the Skills CLI
→ Learn AMD best practices and workflows
→ Build AI applications from client to cloud
→ Pick only the skills you need

Get started: npx skills add amd/skills

Explore the project, install a skill, and let us know what you'd like to see next: https://t.co/hb7XqpLDYu

[GPU MODE @GPU_MODE]

Kinda wild that our community got together and collectively made Kimi inference on MI355X faster than B200 across all settings. Congrats to the winners team RadeonFlow and thanks again to AMD for working so closely with our community

[Subha @subha_v1]

@diyasdrafts yayy hbd too😃

[Subha @subha_v1]

@gpusteve it’s goated at kernel engineering esp compared to opus or sol😼

[Subha @subha_v1]

@joshuaykoo lebron👍

[Subha @subha_v1]

@gpusteve kittens!

[Subha @subha_v1]

@thsottiaux

[Subha @subha_v1]

@Tagrtagr yayy cu there :D

[Subha @subha_v1]

@joshuaykoo yay congrats josh