Book-Length Deep Dive Explains Virtual Memory From First Principles: Page Tables, TLBs, NUMA
abhi9u · x · 2026-09-30
Abhinav Upadhyay published a months-in-the-making, near-book-length article on virtual memory (with a 60-page PDF/EPUB edition), aimed at building a complete mental model for high-performance, data-intensive systems.
Key coverage:
- Why virtual memory exists: process isolation, memory protection, and the illusion of abundant memory
- Virtual address space layout and how a process's memory is organized into segments
- Demand paging for lazy allocation, copy-on-write for fast fork
- mmap file I/O avoiding the page-cache-to-user-buffer copy
- Page reclaim, swap, and the page cache
- Performance effects: access patterns, huge pages, TLB shootdowns, and NUMA placement
The piece also systematically covers page tables, MMU, TLB, L1/L2/L3 caches, MESI cache coherence, false sharing, and mmap/munmap — a practical reference for engineers debugging memory performance.
Related event: Book-Length Deep Dive Explains Virtual Memory from First Principles(2 posts)→
More from Infra
- Miles v0.1.1 Ships Multi-LoRA Sharing and Native Claude Code Agentic Training — ying11231 · 2026-09-30
- vLLM gains RAM offloading: DeepSeek-V4-Flash-Vision-Exp runs on 4x R9700 locally — sloptimizer · 2026-09-30
- Buy 4 Mac Studios and run a 1T-param FP4 model locally to beat API pricing — DimitrisPapail · 2026-09-30
- NVIDIA releases Nemotron Labs for running Nemotron models locally on DGX Spark & Station — NVIDIAAI · 2026-09-30
- AI infra will look like a distributed energy grid, not a few giant plants — sarahookr · 2026-09-30
- Efficient Computer Raises $97M to Scale Chips Claiming 10x Energy Efficiency for AI — rebeccakaden · 2026-09-30