NVIDIA
V100
SXM2
VRAM
32GB
FP32 TFLOPS
15.7 TFLOPS
CUDA Cores
5,120
TDP
300 W
Provider Marketplace
Best Value
—
Awaiting listings
No additional provider yet
All Cloud Providers
TensorDockCheapest
On-Demand
$0.17/ hour
Estimated Cost
Estimates only — rates are collected automatically from public provider pages and may be out of date. Prices vary by region, commitment term, and availability, and typically exclude storage, egress, and tax. Confirm current pricing with the provider before purchasing. Last collected .
Compute Performance
FP647.8 TFLOPS
FP3215.7 TFLOPS
FP16125 TFLOPS
Architecture
MicroarchitectureVolta
CUDA Cores5120
Tensor Cores640 Tensor Cores
Dynamic PrecisionSupported (mixed precision)
Memory & VRAM
Memory TypeHBM2
Total Capacity32GB
Bandwidth1134 GB/sec
ECC SupportYes
Connectivity & Scaling
InterconnectNVLink
IB Bandwidth300 GB/s
Max GPUs/Node8
Virtualization
vGPU ReadinessSupported (NVIDIA Virtual Compute Server (vCS) software)
GPU SharingIndependent thread scheduling; sharing resources among small jobs
Power & Efficiency
TDP300 W
Peak Power300
Thermal LimitsPassive
Physical Design
Form FactorSXM2
CoolingPassive
Software Ecosystem
PyTorchevery major deep learning framework
TensorFlowevery major deep learning framework
Dockercontainer=19.11-py3
Compiler StackCompute APIs CUDA, DirectCompute, OpenCL ™, OpenACC ®
Kernel Optimindependent thread scheduling enables finer-grain synchronization and improves GPU utilization by sharing resources among small jobs.
Server & Deployment
PreconfiguredNVIDIA HGX-2
DGX/HGXNVIDIA HGX-2
System Compatibility
MotherboardForm Factor: SXM2
Rack Power300 W
Benchmarks & Throughput
Scaling Efficiency
NVLink: up to 300 GB/s; up to eight V100 accelerators can be interconnected
Multi-GPU Scalability
Workload Readiness
Market Authority
Supercomputer Usage
NVIDIA HGX-2
Key Strengths
Limitations
Also in the Lineup
Expert Insight
The V100 represents a strategic leap in AI compute. When comparing cloud providers, consider not just the hourly rate, but also the interconnect bandwidth (InfiniBand/NVLink) and regional availability which can significantly impact total cost of ownership for large-scale training.
Glossary Terms
FP32 TFLOPS
VRAM
TDP
Cores
Information updated daily. Cloud pricing subject to vendor availability.