NVIDIA Tesla M40 24GB
Very cheap 24GB, but Maxwell (2015): no fast FP16, compute 5.2, dropped by many newer frameworks/CUDA releases. ~250W passive. Only for tinkering with llama.cpp-style FP32/quantized inference.
NVIDIA · gpu-server · 24 GB VRAM · 250 W · tags: server, maxwell, passive, budget, legacy, cuda
EU / DE trend
No history points for this filter.
US trend
No history points for this filter.
Accepted offers (0)
No live price yet. We only show prices we could read from a shop page — no estimates.
Search used listings
Server GPUs are scam-prone. Check seller history and photos; anything under about €60 is likely broken, parts-only or fake.