Show HN: Hey, Reachy – near-real-time voice for the Reachy Mini desktop robotrunreachyrun.com·2 pts·RyeCatcher·0
Warp Decode vs. vLLM's Triton kernel: where each wins (crossover analysis)ai.rundatarun.io·2 pts·RyeCatcher·0
Show HN: Dendrite – O(1) KV cache forking for tree-structured LLM inferencegithub.com·3 pts·RyeCatcher·1
Fixing macOS window chaos: display-aware layout restore with Hammerspoonai.rundatarun.io·2 pts·RyeCatcher·0
Nvidia DGX Spark: When benchmark numbers meet production realitypublish.obsidian.md·152 pts·RyeCatcher·117