Popular repositories Loading

1. layer-tensor-parallel-bench layer-tensor-parallel-bench Public

   Layer split vs tensor parallelism in llama.cpp - measured on two platforms 16x apart in interconnect bandwidth

   Python 1

2. supraHex supraHex Public

   Forked from TinyChipHub/supraHex

   supraHex is the next iteration of the ultraHex featureing 6 BM1368 chips

   OpenSCAD

3. bitaxeHex-TCH bitaxeHex-TCH Public

   Forked from TinyChipHub/bitaxeHex-TCH

   multi-ASIC Bitcoin miner

   OpenSCAD

4. local-ai-cost-cutting local-ai-cost-cutting Public

   Cutting local AI server build cost with mixed used GPUs - measured data, reproduction scripts, and engine patches (KO/EN)

   Python

5. image-model-split-bench image-model-split-bench Public

   What it costs to split an image generation model across GPUs - layer vs row split vs worker parallel, measured on P104-100 x4 (PCIe Gen1 x4) and V100 x4 (PCIe Gen3 x16)

   Shell