A workstation GPU against a data center card, one Ampere and one Hopper: 48 GB GDDR6 against 96 GB HBM3, 5.2x the memory bandwidth on the H20, and NVLink 112 GB/s on the RTX A6000 where the H20 is PCIe only.
At a glance
Updated 6 minutes ago- Memory
- 48 GB
- GDDR6
- Memory bandwidth
- 768 GB/s
- From
- $0.35 /GPU/hr
- 19 providers
- Memory
- 96 GB
- HBM3
- Memory bandwidth
- 4.00 TB/s
- From
- --
Price comparison
Cheapest listed price per GPU per hour by billing type, and the median across providers.
| Billing | RTX A6000 | H20 | Difference |
|---|---|---|---|
| On-demand | $0.35 | -- | |
| Spot | $0.40 | No spot listings | |
| Reserved | $0.35 | No reserved listings | |
| Median | $0.56 | -- |
Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.
Price history
RTX A6000 vs H20 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | RTX A6000 | H20 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 48 GB GDDR6 | 96 GB HBM3 | 2x more on H20 |
| Bandwidth | 768 GB/s | 4,000 GB/s | 5.2x higher on H20 |
| Bus width | 384-bit | ||
| Platform | |||
| Interconnect | NVLink 112 GB/s | PCIe only | NVLink on RTX A6000 only |
| Board power | 300 W | 400 W | 1.3x higher on H20 |
| Released | Q4 2020 | Q1 2024 | 3 years newer on H20 |
| Precision | |||
| Supported formats | FP16BF16TF32FP32INT4INT8 | FP8FP16BF16TF32FP32FP64INT8 | |
Source: RTX A6000 datasheet .
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | RTX A6000 /GPU/hr | H20 /GPU/hr |
|---|---|---|
|
|
On-Demand $0.35 | -- |
|
|
On-Demand $0.40 | -- |
|
|
On-Demand $0.54 | -- |
|
|
On-Demand $0.55 | -- |
|
|
On-Demand $0.55 | -- |
|
|
On-Demand $0.55 | -- |
|
|
On-Demand $0.55 | -- |
|
|
On-Demand $0.75 | -- |
19 providers list the RTX A6000 and 0 the H20. Every listing is on the RTX A6000 page and the H20 page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia RTX A6000 was last updated on Sept. 7, 2026, and for Nvidia H20 on Sept. 7, 2026.