All GPUs
DatacenterNone available
V100 32GB
NVIDIA Tesla V100 32GB
Proven datacenter GPU with 32GB HBM2 for AI training and inference.
Specifications
VRAM32 GB
Memory TypeHBM2
Bandwidth900 GB/s
CUDA Cores5,120
Tensor Cores640
FP32 Performance15.7 TFLOPS
FP16 Performance31.4 TFLOPS
TDP300 W
ArchitectureVolta
GPU ChipGV100
Best For
AI TrainingInferenceScientific ComputingHPC
Price per hour
$0.12
Cost Estimate
per hour$0.12
per day$2.88
per week$20.16
per month$86.40