2024-12-19
5 min read
Understanding Landau Symbols: The Language of Algorithm Efficiency
Ever wondered how computer scientists compare algorithms without getting bogged down in messy details? Enter Landau symbols — an elegant notation system that lets us describe how algorithms scale.
read more →
2024-12-03
20 min read
NVIDIA GPU Management
From cold starts to memory error protection - five essential GPU management concepts every ML engineer should know. Covers persistence mode, MPS, MIG, clock speed management, and ECC memory.
read more →
2024-11-30
15 min read
Introducing InferenceX: Open Source vLLM Benchmarking for Multi-GPU Inference
I built InferenceX, an open-source tool for benchmarking LLM inference across different GPU configurations. It measures throughput, latency, TTFT, and real power efficiency using nvidia-smi. Works with any HuggingFace model that vLLM supports.
read more →