New video explains quantization basics: how 405B weights shrink from 810GB to ~200GB
arpit_bhayani · x · 2026-09-08
Engineer arpitbhayani published a fundamentals video on quantization, covering what it is, what model weights actually are in memory, and how converting 16-bit floats to 4-bit integers makes inference faster and cheaper. Example: Llama 3.1's 405B parameters need roughly 810GB at 16-bit, dropping to 200GB at 4-bit — still huge, but finally approachable.
Related event: Engineer Releases Video Explaining Quantization Basics(2 posts)→
More from Infra
- Ubicloud publishes "Fixing Tool-Call Accuracy in vLLM" — petrusenko_max · 2026-10-26
- Meta approves nonvoting Class C shares as 2026 capex guidance hits $135B–$145B — thione · 2026-10-26
- Unsloth Ships MiniMax-M3.1-Flash Quants: 2-bit Retains 91.5% Accuracy, Runs in 150GB RAM — danielhanchen · 2026-10-26
- Engineer with thousands of fabricated devices pours cold water on photonic crystal modulator yields — jwt0625 · 2026-10-26
- Anthropic S-1 reveals 48 cents of compute cost per dollar of Q2 revenue — nmasc_ · 2026-10-26
- Persistent AI joins UR network to route private AI inference with encrypted, IP-hidden traffic — 0xSammy · 2026-10-26