Skip to content
VRAMWATCH

NVIDIA Tesla V100 32GB rental price

Datacenter

Volta, still rentable a decade on. Priced against its electricity bill more than its performance.

live · updated 17d ago

VRAM

32 GB HBM2

Architecture

Volta

Segment

Datacenter

Providers quoting

4

best rate

$0.028/GPU-hr at Vast.ai

24h

7d

30d

cheapest 8× node

$12.24/hr

The cheapest NVIDIA Tesla V100 32GB you can rent right now is $0.028 per GPU-hour at Vast.ai, against $3.901 at the dearest of the 4 providers quoting it — a 140.5× spread for identical silicon. A full 8-GPU node starts at $12.24/hr at Microsoft Azure. Rates below refresh automatically.

Rent vs. buy

buy outright

$669

rent

$0.028/hr

breakeven

2.7 years

Buying a NVIDIA Tesla V100 32GB at today’s street price of $669 costs the same as 24,084 hours (2.7 years) of renting one at $0.028 an hour. Run it less than that and renting is simply cheaper; run it more and the card starts paying itself back — but this is a floor, not a total cost of ownership. It counts only the purchase price against the rental rate, and ignores power, cooling, rack space, networking, depreciation and the cost of tying up the capital, all of which push the real breakeven further out.

NVIDIA Tesla V100 32GB purchase prices and history →

V100-32 hourly rate history

$0.00$1.00$2.00$3.00$4.00Aug 29

Daily low $/GPU-hour per provider, cheapest provider first — hover the chart for detail.

Where to rent the NVIDIA Tesla V100 32GB

Vast.ai

Marketplace

$0.028/GPU-hr

2× V100-32 · $0.056/hr per node · Washington, US

  • 2× Tesla V100 x2$0.056/hr
  • 2× Tesla V100 x2$0.056/hr
  • 2× Tesla V100 x2$0.056/hr

Microsoft Azure

Hyperscaler

$1.530/GPU-hr

8× V100-32 · $12.24/hr per node · eastus

  • 8× ND40rs_v2$22.03/hr

Paperspace

GPU cloud

$2.340/GPU-hr

1× V100-32 · $2.340/hr per node · pcie · US, New York, NY

  • 4× V100-32Gx4$9.360/hr
  • 2× V100-32Gx2$4.680/hr

AWS

Hyperscaler

$3.901/GPU-hr

8× V100-32 · $31.21/hr per node · us-east-1

V100-32 cluster nodes

Multi-GPU configurations, cheapest per GPU-hour first. For distributed training the interconnect matters as much as the rate — an NVLink or InfiniBand node and the same count of loose PCIe cards are not the same machine.

ProviderConfigurationInterconnect$/GPU-hrNode/hr
Vast.ai2× Tesla V100 x2 · Washington, US$0.028$0.056/hr
Microsoft Azure8× ND40s_v2 · eastus$1.530$12.24/hr
Paperspace4× V100-32Gx4 · US, New York, NYpcie$2.340$9.360/hr
Paperspace2× V100-32Gx2 · US, New York, NYpcie$2.340$4.680/hr
AWS8× p3dn.24xlarge · us-east-1$3.901$31.21/hr
NVIDIA Tesla V100 32GB Rental Price — $0.028/hr (Live, Vast.ai) | VRAMWATCH