Data center

NVIDIA Tesla V100 PCIe 32GB

Specs for the NVIDIA Tesla V100 PCIe 32GB, with the source and date of each value.

NVIDIA

Data center

Contact for availability

Overview

This catalog entry covers the NVIDIA Tesla V100 PCIe 32GB. Review the sourced specifications below and confirm software, power and cooling requirements before choosing hardware.

Memory capacity

32 GB

Model name as supplied by the business · Checked 2026-10-20

Key Specifications

Spec	Value
Dimensions	Dual-slot, length 10½ in
Brand	NVIDIA
Graphics card usage	Compute
Graphics card generation	Tesla
Graphics card ports	None
Profile size	Dual-slot
Interface	PCI Express 3.0 x16
Memory size	32 GB
Model	NVIDIA Tesla V100
GPU architecture	Volta
Manufacturing process	TSMC 12 nm FFN (FinFET NVIDIA)
Transistor count (billions)	21.1
CUDA cores	5,120
Tensor cores	640
Base clock	1,230 MHz
Boost clock	1,380 MHz
Memory type	HBM2
Memory bandwidth	900 GB/s
Memory bus width	4,096 bit
Power draw	250 W
FP32 compute	14 TFLOPS
FP64 compute	7 TFLOPS
Tensor compute (FP16)	112 TFLOPS
Multi-GPU link	Not supported (PCIe)
Cooling	Passive
Primary applications	Data center, HPC, Deep learning, Artificial intelligence, Scientific simulation
Operating ambient temperature	0–45 °C
Maximum GPU operating temperature	83 °C
Maximum memory operating temperature	85 °C
Software support	NVIDIA CUDA Toolkit, cuDNN, TensorRT
Form factor	Dual-slot, 10.5 in, passive heat sink