Chart shows LLM context windows growing far faster than RAM or CPU clocks
nbaschez · x · 2026-07-27
A Reddit post with a chart compares the growth curves of LLM context windows, PC RAM, CPU clock speed, and transistor counts, arguing that context windows have scaled far faster than traditional hardware trends.
- The chart plots “growth relative to start” on a log scale versus years since the technology began.
- LLM context windows jump rapidly, far outpacing CPU clock speed growth.
- PC RAM follows a stronger growth curve than CPU clocks, but still does not match the steep rise in context windows.
- The implicit takeaway is that model-side context expansion is moving much faster than the hardware trends people usually compare against.
More from Infra
- AI chip startups are attacking Nvidia’s data-moving bottleneck, with a $58B private-paper stack — FinanceYF5 · 2026-07-27
- AMD and South Korea Partner on Heterogeneous Computing and Local NPU R&D — JungWooHa2 · 2026-07-27
- NVIDIA and dozens of firms back a letter arguing open frontier models matter — ying11231 · 2026-07-27
- CXMT reportedly jumps nearly 500% on market debut amid AI memory-chip demand — Polymarket · 2026-07-27
- Local vs. Cloud: Evaluating image generation costs for indie game devs — Simple-Evidence-9125 · 2026-07-27
- A tweet argues closed models may cost $10,000 in compute for every megabyte of weights — JosephJacks_ · 2026-07-27