Two Nvidia data center GPUs, one Ampere and one Grace Blackwell: 80 GB HBM2e against 186 GB HBM3e and 4.1x the memory bandwidth on the GB200.
At a glance
Memory
- A100
- 80 GB
- GB200
- 186 GB
2.3x more on GB200
Memory bandwidth
- A100
- 1.94 TB/s
- GB200
- 8.00 TB/s
4.1x higher on GB200
Cheapest on-demand
- A100
- $0.89 /GPU/hr
- GB200
- $10.50 /GPU/hr
12x more expensive on GB200
Price comparison
Median price per GPU per hour by billing type, with the cheapest listing under it.
| Billing | A100 | GB200 | Difference |
|---|---|---|---|
| On-demand | $1.81 /GPU/hr from $0.89 | $16.00 /GPU/hr from $10.50 | 8.8x more expensive on GB200 |
| Reserved | $1.60 /GPU/hr from $0.66 (24 mo) | -- /GPU/hr from $10.58 | |
| Spot | $1.20 /GPU/hr from $0.67 | -- /GPU/hr from $27.04 |
Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Medians are across providers, one vote each, so a large catalog does not outweigh a small one. No median for reserved and spot on the GB200: we only show one when at least 3 providers list the GPU on that billing type. Verify before provisioning. More on how we price.
Price history
A100 vs GB200 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | A100 | GB200 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 80 GB HBM2e 40 GB on SXM, PCIe | 186 GB HBM3e | 2.3x more on GB200 |
| Bandwidth | 1,935 GB/s 2,039 GB/s on SXM 1,555 GB/s on SXM, PCIe | 8,000 GB/s | 4.1x higher on GB200 |
| Bus width | 5,120-bit | ||
| Compute · dense | |||
| FP4 | Not supported | 10,000 TFLOPS | GB200 only |
| FP8 | Not supported | 5,000 TFLOPS | GB200 only |
| FP16 / BF16 | 312 TFLOPS | 2,500 TFLOPS | 8x higher on GB200 |
| INT8 | 624 TOPS | 5,000 TOPS | 8x higher on GB200 |
| FP32 | 19.5 TFLOPS | 80 TFLOPS | 4.1x higher on GB200 |
| FP64 | 9.7 TFLOPS | 40 TFLOPS | 4.1x higher on GB200 |
| Platform | |||
| Interconnect | NVLink 600 GB/s | NVLink 1,800 GB/s | 3x higher on GB200 |
| Board power | 300 W | 1,200 W | 4x higher on GB200 |
| Released | Q2 2020 | Q1 2024 | 4 years newer on GB200 |
| Precision | |||
| Supported formats | FP16BF16TF32FP32FP64INT4INT8 | FP4FP6FP8FP16BF16TF32FP32FP64INT8 | |
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: A100 datasheet and GB200 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | A100 /GPU/hr | GB200 /GPU/hr |
|---|---|---|
GPU.ai
|
On-Demand $0.89 | -- |
Thunder Compute
|
On-Demand $1.09 | -- |
Massed Compute
|
On-Demand $1.35 | -- |
Packet·ai
|
On-Demand $1.43 | -- |
Spheron
|
On-Demand $1.43 | -- |
Oblivus
|
On-Demand $1.47 | -- |
Sesterce
|
On-Demand $1.49 | -- |
Geodd
|
On-Demand $1.49 | -- |
CoreWeave
|
-- | On-Demand $10.50 |
Oracle Cloud
|
-- | On-Demand $16.00 |
Azure
|
-- | On-Demand $27.04 |
AWS
|
-- | Reservation $10.58 |
CUDO
|
-- | Custom on request |
Canopy Wave
|
-- | Custom on request |
Nebius
|
-- | Custom on request |
Together AI
|
-- | Custom on request |
43 providers list the A100 and 8 the GB200. Every listing is on the A100 page and the GB200 page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia A100 was last updated on Sept. 7, 2026, and for Nvidia GB200 on Sept. 7, 2026.