2025-03-31Mind Maps March 2025 – CUTLASS, CuTe Layout Algebra deep dive, The Ultra-Scale Playbook sneak peek, picotron.
2025-03-01Mind Maps February 2025 – simplifying OCANNL with gradient tensors, more in-depth MLIR, deep dive on polyhedral optimization for loop program transformations / polyhedral schedulers-compilers, Integer Set Library, machine learning loop optimizations in Tiramisu and Halide, matrix multiplication: on CPU / in CUDA / on AMD GPUs, the FineWeb dataset.
2025-02-01Mind Maps January 2025 – Apple’s MLX, maintaining habits, Andrej Karpathy’s educational resources, llm.c (optimized no-framework GPT2), CUDA warp-level primitives and cooperative groups, cuDNN (optimized NNs library / framework), MPI (Message Passing Interface), NCCL (NVIDIA Collective Communications Library similar to MPI), Caten.