Architecting Sparse Tensor Computations on NVIDIA Hopper: A Guide to Efficient Memory Management
A practical guide to managing memory and optimizing sparse tensor operations on NVIDIA Hopper GPUs for AI and HPC workloads.
Cuda
Mastering CUDA Graphs: Low-Overhead Kernel Launching for High-Performance Computing
Learn how CUDA Graphs reduce kernel launch overhead, enable dynamic task scheduling, and improve throughput in high-performance computing applications.
Optimizing GPU Memory Pools: Inside PyTorch's CUDA Caching Allocator
PyTorch’s CUDA caching allocator hides the cost of repeated allocations by recycling VRAM in fixed-size blocks. Here’s how it works, why it matters at scale, and how to squeeze more performance out of it.
Build a GPTQ-Style Int4 Quantization Engine From Scratch
Build a real int4 quantization engine from scratch: grouped scales, zero points, and a fast dequant matmul kernel that runs without CUDA. A substantive portfolio project that signals real systems skill.
Building a Rotary Positional Embedding Kernel from Scratch: A Portfolio Project That Signals Real Systems Skill
Build a real rotary positional embedding kernel in PyTorch with a custom autograd node, fused CUDA-style kernels, and a tiny test harness. The project you ship on a CV to signal GPU, systems, and ML-engineering depth.