Statistics for language Cuda

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

📚Modern CUDA Learn Notes: 200+ Tensor/CUDA Cores Kernels🎉, HGEMM, FA2 via MMA and CuTe, 98~100% TFLOPS of cuBLAS/FA2.

FlashInfer: Kernel Library for LLM Serving

LLM training in simple, raw C/CUDA

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

CUTLASS and CuTe Examples

SplatAD: Real-Time Lidar and Camera Rendering with 3D Gaussian Splatting for Autonomous Driving

flash attention tutorial written in python, triton, cuda, cutlass

The modified differential Gaussian rasterization in the CVPR 2024 highlight paper: GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting.

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

📚Modern CUDA Learn Notes: 200+ Tensor/CUDA Cores Kernels🎉, HGEMM, FA2 via MMA and CuTe, 98~100% TFLOPS of cuBLAS/FA2.

LLM training in simple, raw C/CUDA

FlashInfer: Kernel Library for LLM Serving

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

SplatAD: Real-Time Lidar and Camera Rendering with 3D Gaussian Splatting for Autonomous Driving

No description

A tool for examining GPU scheduling behavior.

CUDA 算子手撕与面试指南

This package contains the original 2012 AlexNet code.

📚Modern CUDA Learn Notes: 200+ Tensor/CUDA Cores Kernels🎉, HGEMM, FA2 via MMA and CuTe, 98~100% TFLOPS of cuBLAS/FA2.

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

LLM training in simple, raw C/CUDA

FlashInfer: Kernel Library for LLM Serving

This package contains the original 2012 AlexNet code.

DDN-SLAM: Real-time Dense Dynamic Neural Implicit SLAM (RA-L 2025)

No description

SpInfer: Leveraging Low-Level Sparsity for Efficient Large Language Model Inference on GPUs

No description

DeepEP: an efficient expert-parallel communication library

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

This package contains the original 2012 AlexNet code.

[ICLR2025 Spotlight] SVDQuant: Absorbing Outliers by Low-Rank Components for 4-Bit Diffusion Models

Quantized Attention that achieves speedups of 2.1-3.1x and 2.7-5.1x compared to FlashAttention2 and xformers, respectively, without lossing end-to-end metrics across various models.

LLM training in simple, raw C/CUDA

DeepEP: an efficient expert-parallel communication library

DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling

A massively parallel, optimal functional runtime in Rust

📚Modern CUDA Learn Notes: 200+ Tensor/CUDA Cores Kernels🎉, HGEMM, FA2 via MMA and CuTe, 98~100% TFLOPS of cuBLAS/FA2.

No description

Learnings and programs related to CUDA

DDN-SLAM: Real-time Dense Dynamic Neural Implicit SLAM (RA-L 2025)

A quantization algorithm for LLM

llama3.cuda is a pure C/CUDA implementation for Llama 3 model.