4 ms·
That's why I run a local Qwen3-Next model on an NVIDIA Thor dev kit (Apple Silicon and DGX Spark are other options but they are even more expensive for 128GB VR
by cat_plus_plus 9mo ago
That's why I run a local Qwen3-Next model on an NVIDIA Thor dev kit (Apple Silicon and DGX Spark are other options but they are even more expensive for 128GB VRAM)