__tensorcore__'s profile picture. MLIR, CUTLASS,Tensor Core arch @NVIDIA. Mechanic @hpcgarage. Exercise of any 1st amendment rights are for none other than myself.

Vijay

@__tensorcore__

MLIR, CUTLASS,Tensor Core arch @NVIDIA. Mechanic @hpcgarage. Exercise of any 1st amendment rights are for none other than myself.

مثبتة

🚨🔥 CUTLASS 4.0 is released 🔥🚨 pip install nvidia-cutlass-dsl 4.0 marks a major shift for CUTLASS: towards native GPU programming in Python slidehelloworld.png docs.nvidia.com/cutlass/media/…

__tensorcore__'s tweet image. 🚨🔥 CUTLASS 4.0 is released 🔥🚨

pip install nvidia-cutlass-dsl

4.0 marks a major shift for CUTLASS: towards native GPU programming in Python

slidehelloworld.png

docs.nvidia.com/cutlass/media/…

Vijay أعاد

Today we are releasing our first public beta of Nsight Python! The goal is to simplify the life of a Python developer by proving a pythonic way to analyze your kernel code! Check it out, provide feedback! Nsight Python — nsight-python docs.nvidia.com/nsight-python/


Vijay أعاد

BREAKING: The CUDA moat has just expanded again! PyTorch Compile/Inductor can now target NVIDIA Python CuTeDSL in addition to Triton. This enables 2x faster FlexAttention compared to Triton implementations. We explain below 👇 As we explained in our April 2025 AMD 2.0 piece,…

SemiAnalysis_'s tweet image. BREAKING: The CUDA moat has just expanded again! PyTorch Compile/Inductor can now target NVIDIA Python CuTeDSL in addition to Triton. This enables 2x faster FlexAttention compared to Triton implementations. We explain below 👇

As we explained in our April 2025 AMD 2.0 piece,…

Vijay أعاد

Quick life update. Moved to California to work at NVIDA. Oh I have so much to learn

KuterDinel's tweet image. Quick life update. Moved to California to work at NVIDA. Oh I have so much to learn

Vijay أعاد

Excited to share what friends and I have been working on at @Standard_Kernel We've raised from General Catalyst (@generalcatalyst), Felicis (@felicis), and a group of exceptional angels. We have some great H100 BF16 kernels in pure CUDA+PTX, featuring: - Matmul 102%-105% perf…

anneouyang's tweet image. Excited to share what friends and I have been working on at @Standard_Kernel 

We've raised from General Catalyst (@generalcatalyst), Felicis (@felicis), and a group of exceptional angels. 

We have some great H100 BF16 kernels in pure CUDA+PTX, featuring:
- Matmul 102%-105% perf…

Vijay أعاد

Today Thinking Machines Lab is launching our research blog, Connectionism. Our first blog post is “Defeating Nondeterminism in LLM Inference” We believe that science is better when shared. Connectionism will cover topics as varied as our research is: from kernel numerics to…

thinkymachines's tweet image. Today Thinking Machines Lab is launching our research blog, Connectionism. Our first blog post is “Defeating Nondeterminism in LLM Inference”

We believe that science is better when shared. Connectionism will cover topics as varied as our research is: from kernel numerics to…

“TogetherAI’s chief scientist @tri_dao announced Flash Attention v4 … uses CUTLASS CuTe Python DSL” As always, thanks for being the tip of the spear and pushing us along too 💚

TogetherAI's Chief Scientist @tri_dao announced Flash Attention v4 at HotChips Conference which is up to 22% faster than the attention kernel implementation from NVIDIA's cuDNN library. Tri Dao was able to achieve this 2 key algorithmic changes. Firstly, it uses a new online…

SemiAnalysis_'s tweet image. TogetherAI's Chief Scientist @tri_dao announced Flash Attention v4 at HotChips Conference which is up to 22% faster than the attention kernel implementation from NVIDIA's cuDNN library. Tri Dao was able to achieve this 2 key algorithmic changes. Firstly, it uses a new online…


Vijay أعاد

Using CUTLASS CuTe-DSL, TogetherAI's Chief Scientist @tri_dao announced that he has written kernels that is 50% faster than NVIDIA's latest cuBLAS 13.0 library for small K reduction dim  shapes on Blackwell during today's hotchip conference.  His kernels beats cuBLAS by using 2…

SemiAnalysis_'s tweet image. Using CUTLASS CuTe-DSL, TogetherAI's Chief Scientist @tri_dao announced that he has written kernels that is 50% faster than NVIDIA's latest cuBLAS 13.0 library for small K reduction dim  shapes on Blackwell during today's hotchip conference.  

His kernels beats cuBLAS by using 2…

Vijay أعاد

Cute-DSL is basically perfect (for me). thank you nvidia and cutlass team. i no longer need to wait for long compile times because i underspecified a template param. i hope everyone involved gets an extra chicken nugget in their happy meal


Vijay أعاد

On Sep 6 in NYC, this won't be your typical hackathon where you do your own thing in a corner and then present at the of the day. You'll deploy real models to the market, trades will happen, chaos should be expected. The fastest model is great but time to market matters more.

marksaroufim's tweet image. On Sep 6 in NYC, this won't be your typical hackathon where you do your own thing in a corner and then present at the of the day. You'll deploy real models to the market, trades will happen, chaos should be expected. The fastest model is great but time to market matters more.

Vijay أعاد

ariXv gpu kernel researcher be like: • liquid nitrogen cooling their benchmark GPU • overclock their H200 to 1000W "Custom Thermal Solution CTS" • nvidia-smi boost-slider --vboost 1 • nvidia-smi -i 0 --lock-gpu-clocks=1830,1830 • use specially binned GPUs where the number…

SemiAnalysis_'s tweet image. ariXv gpu kernel researcher be like:
• liquid nitrogen cooling their benchmark GPU
• overclock their H200 to 1000W "Custom Thermal Solution CTS" 
• nvidia-smi boost-slider --vboost 1
• nvidia-smi -i 0 --lock-gpu-clocks=1830,1830
• use specially binned GPUs where the number…

Vijay أعاد

Part 2: developer.nvidia.com/blog/cutlass-3… Covers the design of CUTLASS 3.x itself and how it builds a 2 layer GPU microkernel abstraction using CuTe as the foundation.


CUTLASS 4.1 is now available, which adds support for ARM systems (GB200) and block scaled MMAs

🚨🔥 CUTLASS 4.0 is released 🔥🚨 pip install nvidia-cutlass-dsl 4.0 marks a major shift for CUTLASS: towards native GPU programming in Python slidehelloworld.png docs.nvidia.com/cutlass/media/…

__tensorcore__'s tweet image. 🚨🔥 CUTLASS 4.0 is released 🔥🚨

pip install nvidia-cutlass-dsl

4.0 marks a major shift for CUTLASS: towards native GPU programming in Python

slidehelloworld.png

docs.nvidia.com/cutlass/media/…


Vijay أعاد

Hierarchical layout is super elegant. Feels like the right abstraction for high performance GPU kernels. FlashAttention 2 actually started bc we wanted to rewrite FA1 in CuTe


Vijay أعاد

CuTe is such an elegant library that we stopped working on our own system and wholeheartedly adopted CUTLASS for vLLM in the beginning of 2024. I can happily report that was a very wise investment! Vijay and co should be so proud of the many strong OSS projects built on top 🥳


This is what the internet was made for 🥹

presenting: big jeff's trainium hell



Vijay أعاد

Cosmos-Predict2 meets NATTEN. We just released variants of Cosmos-Predict2 where we replace most self attentions with neighborhood attention, bringing up to 2.6X end-to-end speedup, with minimal effect on quality! github.com/nvidia-cosmos/… (1/5)


Vijay أعاد

Getting mem-bound kernels to speed-of-light isn't a dark art, it's just about getting the a couple of details right. We wrote a tutorial on how to do this, with code you can directly use. Thanks to the new CuTe-DSL, we can hit speed-of-light without a single line of CUDA C++.

🦆🚀QuACK🦆🚀: new SOL mem-bound kernel library without a single line of CUDA C++ all straight in Python thanks to CuTe-DSL. On H100 with 3TB/s, it performs 33%-50% faster than highly optimized libraries like PyTorch's torch.compile and Liger. 🤯 With @tedzadouri and @tri_dao

WentaoGuo7's tweet image. 🦆🚀QuACK🦆🚀: new SOL mem-bound kernel library without a single line of CUDA C++ all straight in Python thanks to CuTe-DSL. On H100 with 3TB/s, it performs 33%-50% faster than highly optimized libraries like PyTorch's torch.compile and Liger. 🤯 

With @tedzadouri and @tri_dao


Vijay أعاد

🦆🚀QuACK🦆🚀: new SOL mem-bound kernel library without a single line of CUDA C++ all straight in Python thanks to CuTe-DSL. On H100 with 3TB/s, it performs 33%-50% faster than highly optimized libraries like PyTorch's torch.compile and Liger. 🤯 With @tedzadouri and @tri_dao

WentaoGuo7's tweet image. 🦆🚀QuACK🦆🚀: new SOL mem-bound kernel library without a single line of CUDA C++ all straight in Python thanks to CuTe-DSL. On H100 with 3TB/s, it performs 33%-50% faster than highly optimized libraries like PyTorch's torch.compile and Liger. 🤯 

With @tedzadouri and @tri_dao

Loading...

Something went wrong.


Something went wrong.