Account
Items 1-12 of 128
Views:
NVIDIA Tesla GPUs
Key takeaways
- 128 listings match this collection right now.
- Tesla cards need above 4G decoding and large BAR support enabled in host BIOS, a setting GPU-ready servers provide but older boards may lack.
- Power draw ranges from 70 watts with no external connector up to 250 watts needing a dedicated cable, so chassis power budget has to match the specific card.
- Maxwell is the oldest generation here, and its CUDA and driver support ends well before Volta or Turing does, which matters for a workload tied to a specific framework version.
Tesla is the NVIDIA datacenter line before the A-series naming. The cards are passively cooled and rely on server chassis airflow, so they belong in GPU-ready servers rather than workstations. Generation sets the capability: Turing (T4) for inference and video, Volta (V100) for training and HPC, Pascal for HPC (P100) and inference or virtual desktop (P40), Maxwell for virtual desktop and legacy compute.
128 listings in this collection, filtered from Graphics Cards. The part number and manufacturer are stated on every listing, and the condition wherever the catalogue records one. For quantity, request a volume quote.
How to choose gpus and accelerators
Datacenter cards (Tesla, A-series) are passively cooled and expect server airflow; workstation cards (Quadro, RTX A) carry their own fans and display outputs. Match the slot (PCIe generation and width), the auxiliary power the card needs, and the memory the workload needs: inference and video work fits in 16 GB to 24 GB, training and large models want 32 GB to 80 GB.