KV cache, not parameter count, may be the real bottleneck for long-context local models

jonejy · reddit · 2026-09-03

The author argues local-LLM discussions overfocus on parameter count while KV cache is the real memory bottleneck at long context: key/value states grow per token, so a model that fits in VRAM can break at 100k–200k context. GQA/MQA and KV cache quantization help but don't change linear growth. Future optimization will target memory movement and persistent state — designing models around 'how much they need to remember' rather than parameter count.

Original post →

More from Infra

Infra channel →