Comparison · GPU vs GPU · read 21:35 UTCTesla V100 16GB vs Tesla V100 32GB
Specifications, today's prices from 32 providers, price per unit of memory and compute, and who rents each cheapest. Better values are highlighted; lower is better for prices.
Side by sideSpecs, prices, value
Highlighted cells are the better of the two: more memory and compute, lower prices, lower cost per unit.
| Tesla V100 16GB | Tesla V100 32GB | Difference | |
|---|---|---|---|
| Memory & compute | |||
| VRAM | 16 GB | 32 GB | -50% |
| Memory bandwidth | 0.90 TB/s | 0.90 TB/s | 0% |
| FP16 tensor | 112 TFLOPS | 125 TFLOPS | -10% |
| FP32 | 14 TFLOPS | 15.7 TFLOPS | -11% |
| Power | 250 W | 300 W | -17% |
| Prices today | |||
| Cheapest on-demand | $0.33 | $0.55 | -41% |
| Median on-demand | $1.71 | $2.48 | -31% |
| Highest on-demand | $3.06 | $4.39 | -30% |
| Cheapest spot | $0.04 | $0.51 | -92% |
| Cheapest reserved | $0.35 | $0.38 | -6% |
| Current offers (n) | 36 | 63 | -43% |
| Providers with a fresh offer | 15 | 17 | -12% |
| Value | |||
| Per GB of VRAM · hour (cheapest) | $0.020 | $0.017 | +18% |
| Per GB of VRAM · hour (median) | $0.107 | $0.078 | +38% |
| Per PFLOP FP16 · hour (median) | $15.268 | $19.862 | -23% |
| Per TB/s · hour (median) | $1.900 | $2.759 | -31% |
| Trend | |||
| 30-day median change | -8.2% | 0.0% | — |
| 90-day median change | -12.6% | -45.2% | — |
90 days · median on-demand price · rebased to 100How the two have moved
Both medians indexed to 100 ninety days ago, so the slopes compare even though the price levels differ.
Cheapest providers today · on-demand · current offersWho rents each one cheapest
- 1AtmosCompute2× $0.33
- 2
OVHcloudTrust 792× $0.78 - 3LambdaTrust 808× $0.79
- 1
LiumTrust 408× $0.55 - 2
Cyfuture AI8× $0.55 - 3
VoltageGPU8× $1.10
Verdict · computed from the numbers aboveChoose one, choose the other
Choose the Tesla V100 16GB if…
- the hourly price matters most: median $1.71 against $2.48, 31% less.
- you serve LLMs: memory bandwidth per dollar is better ($1.900 per TB/s-hour against $2.759).
Choose the Tesla V100 32GB if…
- your model needs more than 16 GB per GPU: 32 GB against 16 GB.
- you want choice: 17 providers have a fresh offer today against 15.
Questions about Tesla V100 16GB vs Tesla V100 32GB
Which is cheaper to rent, the Tesla V100 16GB or the Tesla V100 32GB?
Today the cheapest current on-demand Tesla V100 16GB is $0.33 per GPU-hour and the cheapest Tesla V100 32GB is $0.55; medians are $1.71 and $2.48. Cheapest excludes offers marked unavailable and counts one floor per marketplace.
How much memory do the Tesla V100 16GB and Tesla V100 32GB have?
16 GB of memory on the Tesla V100 16GB, 32 GB of memory on the Tesla V100 32GB, at 0.90 and 0.90 TB/s.
Where do these numbers come from?
From 99 current offers read on the providers' own pages and APIs over the last 45 days (reference 21:35 UTC), or in our October 2026 market study for the few providers our collector does not read. Specifications come from manufacturer datasheets. Every price on the GPU pages links to its receipt and shows its read date.