2048 TPUv4s and deep domain expertise: professor dissects how a gene-regulation model was trained
anshulkundaje · x · 2026-11-12
Stanford professor Anshul Kundaje details how this gene-regulation model was trained on 2048 TPUv4s with meticulous data/model design and top-notch engineering squeezing every drop of value from the dataset. It wasn't brute-force scaling: architecture, training scheme, losses and normalization were crafted by deep domain experts, and the single-cell encoder was just a tweak on scVI — no billion-parameter foundation world model needed.
More from Infra
- Curated 9-paper reading list for LLM quantization beginners, from QLoRA to BitNet — East-Muffin-6472 · 2026-11-12
- llama.cpp PR #31915 fixes Gemma4 tool calling, user reports it "completely fixed" agentic coding — jacek2023 · 2026-11-12
- Nvidia board reportedly working with outside firm to find Jensen Huang successor — kimmonismus · 2026-11-12
- RDNA3-tuned llama.cpp fork doubles Qwen3.8-27B speed on RX 7900 XTX to ~100 tok/s — Jebbyk1 · 2026-11-12
- Ming Chi Kuo: Huang is irreplaceable to Nvidia; Rubin mass production slips to 2027 on HBM4 — zephyr_z9 · 2026-11-12
- Breakdown: Only 3% of Anthropic's 2.68M Chips Are Nvidia, Lowest in AI — teortaxesTex · 2026-11-12