inference/research
draftlow · 0.55updated 2026-10-17

NVIDIA Tesla P4

nvidia pascal (GP104) 2016
This record is a draft. Not yet verified. Provisional. Treat every number on this page as provisional.

Fields

id
accelerators/nvidia-tesla-p4
Vendor
nvidia
Architecture
pascal (GP104)
Release year
2016
Form factors
pcie
VRAM (GB)
8
Memory type
gddr5
Memory bus (bit)
256
Memory bandwidth (GB/s)
192
Bandwidth basis
GDDR5 6 Gbps x 256-bit = 192 GB/s (NVIDIA Tesla P4 spec)
TDP (W)
75
Unified memory
no
Consumer part
no
notes
Low-profile inference card, 8 GB, 75 W. Distinct from the Tesla P40.

Notes

Low-profile inference card, 8 GB, 75 W. Distinct from the Tesla P40.

Cross-references

Interconnect (1)