Faraz Shahsavan

Kernel Engineer at @Baseten | Ex CentML (Acq. by NVIDIA) | Ex Neural Magic (Acq. by Red Hat) | Research at UofT, EPFL, Uni. of Tehran, IPM Institute of Research

[Brian Li @BrianLi23]

https://t.co/ZyhZUElPGk

[Baseten @baseten]

We're building in Canada!

We opened two new offices in Toronto and Montreal, and we're growing fast. If you’re excited about building high-performance AI infrastructure for the world's most ambitious companies, we'd love to meet you!

[Faraz Shahsavan @Faraz9877]

@waterloo_intern Kernels team JACKED!!!

[ali @waterloo_intern]

https://t.co/qe2HpEQZPd

[Faraz Shahsavan @Faraz9877]

@waterloo_intern There's no faster way of doing something than not doing it. INF speedup.

[Faraz Shahsavan @Faraz9877]

@anaskar @baseten @philipkiely Philip is amazing!

[Faraz Shahsavan @Faraz9877]

Check out the new work of the kernels team, led by @YikaiZhu98!

[Yikai Zhu @YikaiZhu98]

https://t.co/rNXhTtf79D

[Faraz Shahsavan @Faraz9877]

@YikaiZhu98 Nice stuff!!

[Faraz Shahsavan @Faraz9877]

@OmerShlomovits We're faster because of timestep distillation

[Faraz Shahsavan @Faraz9877]

Fast image generation has become critical for production AI products, where latency directly affects user experience, throughput, and cost.

Proud to share we at Baseten reached a new milestone in optimizing image generation serving for two frontier models, FLUX.2-dev and Qwen-Image, on NVIDIA Blackwell and Hopper GPUs! 🚀

Key results:
1️⃣ FLUX.2-dev:
* 2.3× faster on B200
* 1.9× faster on H100
2️⃣ Qwen-Image:
* 1.57× faster on B200 with FP4
* 1.18× faster on B200 with FP8
* 1.08× faster on H100 with FP8

These gains come from deep-stack optimizations. Read the article for more details:

[Faraz Shahsavan @Faraz9877]

https://t.co/ngsumzG6jb

[Faraz Shahsavan @Faraz9877]

https://t.co/ngsumzG6jb