A data center GPU against a workstation card, one Hopper and one Blackwell: 80 GB HBM3 against 24 GB GDDR7, 5x the memory bandwidth on the H100, and NVLink 900 GB/s on the H100 where the RTX PRO 4000 is PCIe only.
At a glance
Updated 7 minutes ago- Memory
- 80 GB
- HBM3 · 94 GB on NVL
- Memory bandwidth
- 3.35 TB/s
- 2,000 GB/s on PCIe · 3,938 GB/s on NVL
- From
- $1.50 /GPU/hr
- 53 providers
- Memory
- 24 GB
- GDDR7
- Memory bandwidth
- 672 GB/s
- From
- $20.00 /GPU/hr
- 3 providers
Price comparison
Cheapest listed price per GPU per hour by billing type, and the median across providers.
| Billing | H100 | RTX PRO 4000 | Difference |
|---|---|---|---|
| On-demand | $1.50 | $20.00 | 13x more expensive on RTX PRO 4000 |
| Spot | $0.13 | $13.33 | 100x more expensive on RTX PRO 4000 |
| Reserved | $1.53 | $0.26 | 5.8x more expensive on H100 |
| Median | $3.39 | $20.00 | 5.9x more expensive on RTX PRO 4000 |
Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.
Price history
H100 vs RTX PRO 4000 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | H100 | RTX PRO 4000 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 80 GB HBM3 94 GB on NVL | 24 GB GDDR7 | 3.3x more on H100 |
| Bandwidth | 3,350 GB/s 2,000 GB/s on PCIe 3,938 GB/s on NVL | 672 GB/s | 5x higher on H100 |
| Bus width | 5,120-bit | 192-bit | 27x wider on H100 |
| Compute · dense | |||
| FP4 | Not supported | 589 TFLOPS | RTX PRO 4000 only |
| FP8 | 1,979 TFLOPS | 294.5 TFLOPS | 6.7x higher on H100 |
| FP16 / BF16 | 989.5 TFLOPS | 147.3 TFLOPS | 6.7x higher on H100 |
| INT8 | 1,979 TOPS | 294.5 TOPS | 6.7x higher on H100 |
| FP32 | 67 TFLOPS | 37 TFLOPS | 1.8x higher on H100 |
| FP64 | 34 TFLOPS | Not supported | H100 only |
| Platform | |||
| Interconnect | NVLink 900 GB/s NVLink 600 GB/s on PCIe, NVL | PCIe only | NVLink on H100 only |
| Board power | 700 W | 145 W | 4.8x higher on H100 |
| Released | Q3 2022 | Q1 2025 | 2 years newer on RTX PRO 4000 |
| Precision | |||
| Supported formats | FP8FP16BF16TF32FP32FP64INT8 | FP4FP8FP16BF16TF32FP32INT8 | |
Figures are for each model's headline configuration. Where a form factor carries a different number, it is named under the figure.
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: H100 datasheet and RTX PRO 4000 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | H100 /GPU/hr | RTX PRO 4000 /GPU/hr |
|---|---|---|
Vast.ai
|
On-Demand $2.13 | On-Demand $20.00 |
Lium
|
On-Demand $1.50 | -- |
HyperAI
|
On-Demand $1.80 | -- |
GPU.ai
|
On-Demand $1.98 | -- |
Koyeb
|
On-Demand $2.50 | -- |
Spheron
|
On-Demand $2.65 | -- |
Massed Compute
|
On-Demand $2.73 | -- |
Runcrate
|
On-Demand $2.75 | -- |
Database Mart
|
-- | Reserved (24mo) $0.26 |
Runpod
|
-- | Reservation on request |
53 providers list the H100 and 3 the RTX PRO 4000. Every listing is on the H100 page and the RTX PRO 4000 page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia H100 was last updated on Sept. 7, 2026, and for Nvidia RTX PRO 4000 on Sept. 7, 2026.