Stories by shihab

AVX-512: First Impressions on Performance and Programmability

Mojo: MLIR-Based Performance-Portable HPC Science Kernels on GPUs

Efficient and Portable Mixture-of-Experts Communication

Chris Lattner- Democratizing AI Compute: What about AI Compilers (TVM and XLA)?

Democratizing AI Compute, Part 5: What about CUDA C++ alternatives?

Google "We have no moat, and neither does OpenAI" (2023)

Los Alamos Pushes Memory Wall with "Venado" Supercomputer

CUDA vs. ROCm: A case study

CUDA vs. ROCm: A Case Study

Google: What Do Terrorists Wear?

Spotify’s $1B Podcast Bet Turns into a Serial Drama

Yale Law School Abandons U.S. News Rankings