You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As per the datasheet ( https://www.nvidia.com/content/cloud-computing/pdf/nvidia-grid-datasheet-k1-k2.pdf ) , the K1 has 4 Kepler-GPUs with 4x4 GB DDR3 VRAM (and 768 total CUDA cores); the K2 has only 2, but hig-end Kepler GPUs with 2x4 GB DDR5 VRAM (with 3072 CUDA-cores). Which one is better if I mostly want to use them with stable-diffusion.cpp (and sometimes - very rare - to some llama.cpp based chat, and whisper.cpp) on Linux. Are they recommended at all? (These are fanless "GPU-accelerator-cards" - from the past, using the 470-based driver, and cuda-v11.)
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
As per the datasheet ( https://www.nvidia.com/content/cloud-computing/pdf/nvidia-grid-datasheet-k1-k2.pdf ) , the K1 has 4 Kepler-GPUs with 4x4 GB DDR3 VRAM (and 768 total CUDA cores); the K2 has only 2, but hig-end Kepler GPUs with 2x4 GB DDR5 VRAM (with 3072 CUDA-cores). Which one is better if I mostly want to use them with stable-diffusion.cpp (and sometimes - very rare - to some llama.cpp based chat, and whisper.cpp) on Linux. Are they recommended at all? (These are fanless "GPU-accelerator-cards" - from the past, using the 470-based driver, and cuda-v11.)
Beta Was this translation helpful? Give feedback.
All reactions