Your Data. Your Hardware. Your AI.

Practical guides for running AI locally—from someone who figured it out on a budget,
not a Silicon Valley lab.

Latest

LLM on a Microcontroller: ESP32-S3 at 22.6 tok/s

esp32-ai's 28.9M-parameter model on an $8 ESP32-S3: 10.3 to 22.6 tok/s from five exact patches, same output to the byte. PSRAM bandwidth is the limit.

Oct 11, 2026

Strata vs llama.cpp on an RTX 3090: 78 tok/s vs 18 on the Full Flash-Next

The full Qwen3.8-Flash-Next on a 24 GB RTX 3090 with 64 GB of DDR4. Strata decodes at 78 tok/s against llama.cpp's 18, and reads prompts about 3x faster.

Oct 10, 2026

Does RAM Speed Matter for Local LLMs? 2133 vs 3000, Measured

DDR4-2133 vs 3000 on an RTX 3090 box: no change for models that fit on the GPU, 13 to 29% faster with MoE experts offloaded. Plus why 3200 wouldn't train.

Oct 10, 2026

Qwen 3.8 35B-A3B Is a Distill: Tested Against Qwen 3.6

There is no official Qwen 3.8 35B-A3B. I ran empero-ai's distill against Qwen 3.6 35B on an RTX 3090: same score, same speed, a fifth of the reasoning tokens.

Oct 8, 2026

Strata vs llama.cpp on a 3060: The 7x Is Real. Two Flags Halve It

Qwen3.8-Flash-Next Coder on a 12 GB RTX 3060 and 32 GB of DDR4. Strata prefills 6-7x faster than stock llama.cpp. One batch setting closes most of that.

Oct 5, 2026

Bonsai 2 vs Qwen3.8-27B on an RTX 3090: Speed Doubled, Scope Broke

Bonsai 2 27B, PrismML's ternary Qwen3.8, on a 3090: 1.8x the decode of Q4 from a 7.2 GB file, then 12 of 47 against 19 on a routing split. One field broke.

Sep 24, 2026