Two Nvidia data center GPUs on the Ampere architecture: 80 GB HBM2e against 16 GB GDDR6, 9.7x the memory bandwidth on the A100, and NVLink 600 GB/s on the A100 where the A16 is PCIe only.
At a glance
Memory
- A100
- 80 GB
- A16
- 16 GB
5x more on A100
Memory bandwidth
- A100
- 1,935 GB/s
- A16
- 200 GB/s
9.7x higher on A100
Cheapest on-demand
- A100
- $0.74 /GPU/hr
- A16
- $0.47 /GPU/hr
36% cheaper on A16
Price comparison
Median price per GPU per hour by billing type, with the cheapest listing under it.
| Billing | A100 | A16 | Difference |
|---|---|---|---|
| On-demand | $1.77 /GPU/hr from $0.74 | $0.56 /GPU/hr from $0.47 | 3.2x more expensive on A100 |
| Reserved | $1.62 /GPU/hr from $0.66 (24 mo) | No reserved listings | |
| Spot | $1.14 /GPU/hr from $0.30 | No spot listings |
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. Verify before provisioning. More on how we price.
Price history
A100 vs A16 price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | A100 | A16 | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 80 GB HBM2e 40 GB on SXM, PCIe | 16 GB GDDR6 | 5x more on A100 |
| Bandwidth | 1,935 GB/s 2,039 GB/s on SXM 1,555 GB/s on SXM, PCIe | 200 GB/s | 9.7x higher on A100 |
| Bus width | 5,120-bit | 128-bit | 40x wider on A100 |
| Compute · dense | |||
| FP16 / BF16 | 312 TFLOPS | 17.9 TFLOPS | 17x higher on A100 |
| INT8 | 624 TOPS | 35.9 TOPS | 17x higher on A100 |
| FP32 | 19.5 TFLOPS | 4.5 TFLOPS | 4.3x higher on A100 |
| FP64 | 9.7 TFLOPS | Not supported | A100 only |
| Platform | |||
| Interconnect | NVLink 600 GB/s | PCIe only | NVLink on A100 only |
| Released | Q2 2020 | Q2 2021 | 11 months newer on A16 |
| Precision | |||
| Supported formats | FP16BF16TF32FP32FP64INT4INT8 | FP16BF16TF32FP32INT4INT8 | |
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 A16 datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | A100 /GPU/hr | A16 /GPU/hr |
|---|---|---|
Vast.ai
|
On-Demand $0.74 | -- |
GPU.ai
|
On-Demand $0.81 | -- |
Thunder Compute
|
On-Demand $1.09 | -- |
Lium
|
On-Demand $1.23 | -- |
Verda
|
On-Demand $1.29 | -- |
Massed Compute
|
On-Demand $1.35 | -- |
Packet·ai
|
On-Demand $1.43 | -- |
Oblivus
|
On-Demand $1.47 | -- |
Vultr
|
-- | On-Demand $0.47 |
Runcrate
|
-- | On-Demand $0.56 |
Sesterce
|
-- | On-Demand $0.56 |
44 providers list the A100 and 3 the A16. Every listing is on the A100 page and the A16 page.
Similar GPUs
Other comparisons in the same class.
-
Nvidia H100
From $1.55 /GPU/hr Compare -
Nvidia RTX 5090
From $0.35 /GPU/hr Compare -
Nvidia RTX 4090
From $0.23 /GPU/hr Compare
Our data for Nvidia A100 was last updated on Sept. 7, 2026, and for Nvidia A16 on Sept. 7, 2026.