Видео с ютуба Moe-Cache-Mib

Внутри llama.cpp: как запускать огромные MoE-модели (новый кэш экспертов)

The LRU Cache That Doubled Local MoE Speeds: llama.cpp Expert Offloading

llama.cpp's MoE Cache: 3x On 16GB. Off By Default.

Your MoE Model Is 2× Slower Without This llama.cpp Flag

This New Cache Makes Huge Local AI Models 2x Faster

Moe's Best Map! (Cache Match Making)

llama.cpp увеличил скорость вашего GPU в 2 раза на больших моделях (включите это)

Run 125B Models on 8GB Ram: llama.cpp Expert Cache

Broken in 12 Days: How a 1993 Idea Fixed llama.cpp MoE Caching

llama.cpp Just Got an Expert Cache. Do You Still Need Strata?

Карта CACHE официально вернулась! (Новая карта)

llama.cpp Just Made 125B MoE Models 4× Faster!

В llama.cpp добавили Expert Cache... Победит ли он Strata на 12 ГБ?!

llama.cpp Just Made Huge MoE Models 4× Faster

llama.cpp Cut 31 Wasted GPU Passes

Cache Pro AWP Tutorial [B SITE]

llama.cpp runs MoE models up to 2.2x faster when VRAM runs out

The MoE Trap: Why 3B Active Doesn't Mean 3GB of VRAM (Qwen 35B)

This llama.cpp Fork Just Made Qwen 3.3× Faster