CUDA Templates and Python DSLs for High-Performance Linear Algebra

Cute layout visualization

Triton-based Symmetric Memory operators and examples

A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.

Tile primitives for speedy kernels

Fastest kernels written from scratch

Tensors and Dynamic neural networks in Python with strong GPU acceleration

Helpful tools and examples for working with flex-attention

A versatile pairwise aligner for genomic and spliced nucleotide sequences

A framework for implementing Block Devices in User Space

Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.

Hyrise is a research in-memory database.

Puzzles for learning Triton

DRAMsim3: a Cycle-accurate, Thermal-Capable DRAM Simulator

NoSQL Redis and Memcache traffic generation and benchmarking tool.

ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator

Sampled simulation of multi-threaded applications using LoopPoint methodology