awesome-hpc
Another "awesome-*" list, for random things I find interesting in HPC / ML
https://github.com/arthurfeeney/awesome-hpc
Last synced: 15 days ago
JSON representation
-
Aggregators
-
Conference Proceedings
-
ML Performance
-
Frameworks
-
Efficient Implementations
- Data movement is all you need
- Data movement is all you need
- The Hardware Lottery
- The Hardware Lottery
- IO Complexity of sorting and related problems
- Online Softmax Normalizer
- Self-attention does not need $O(n^2)$ memory
- FlashAttention 1
- FlashAttention 2
- FlashAttention 3, for H100. Uses Asynchrony and low precision
- IO Complexity of sorting and related problems
- Online Softmax Normalizer
- Self-attention does not need $O(n^2)$ memory
- FlashAttention 1
- FlashAttention 2
- FlashAttention 3, for H100. Uses Asynchrony and low precision
-
-
Blogs
-
Lectures
-
Matrix Multiplication and Linear Algebra
-
NVIDIA
Categories
Sub Categories