Humansand's weight teleportation cuts RL sampler pause to <0.23s, bottleneck shifts to GPU decompression ASIC
gharik · x · 2026-11-03
Humansand shared weight teleportation for large-scale RL: faster trainer-to-sampler weight updates than intra-cluster techniques with internet speeds. Using GPUs for massively parallel compression plus heterogeneous decompression pipelines, sampler pause is <0.23s (3s non-blocking prepare), enabling samplers anywhere. The new bottleneck is the GPU's decompression ASIC.
More from Infra
- Apple Q4: $114.5B revenue up 12%, AI/Siri spending to squeeze margins — mitchdeg · 2026-11-03
- Darkbloom passes 1T tokens as Mac compute network pivots to CI/CD — gajesh · 2026-11-03
- OpenRouter Hits 100T Weekly Tokens; Surplus Jumps to 2.87T/Week in 4 Weeks — 0xJeff · 2026-11-03
- Claude Code caches are model-specific: switching models, compacting and idle time all bust the cache — shoumikgoswami · 2026-11-03
- Strata update doubles local inference speed for quad-R9700 setups: 70 to 140-160 tok/s — poofph · 2026-11-03
- Space-based distributed training gains traction as researchers push orbital datacenter vision — MillionInt · 2026-11-03