Two Nvidia data center GPUs, one Ampere and one Ada Lovelace: 16 GB GDDR6 against 48 GB GDDR6 and 4.3x the memory bandwidth on the L40S.
At a glance
Updated 4 minutes ago- Memory
- 16 GB
- GDDR6
- Memory bandwidth
- 200 GB/s
- From
- $0.47 /GPU/hr
- 3 providers
- Memory
- 48 GB
- GDDR6
- Memory bandwidth
- 864 GB/s
- From
- $0.38 /GPU/hr
- 36 providers
Price comparison
Cheapest listed price per GPU per hour by billing type, and the median across providers.
| Billing | A16 | L40S | Difference |
|---|---|---|---|
| On-demand | $0.47 | $0.38 | 19% cheaper on L40S |
| Spot | No spot listings | $0.47 | |
| Reserved | No reserved listings | $0.60 | |
| Median | $0.56 | $1.28 | 2.3x more expensive on L40S |
Median: Median across providers, one vote each, so a large catalog does not outweigh a small one.
Price history
A16 vs L40S price history
Weekly median on-demand price per GPU per hour
Specs comparison
| Spec | A16 | L40S | Difference |
|---|---|---|---|
| Memory | |||
| Capacity | 16 GB GDDR6 | 48 GB GDDR6 | 3x more on L40S |
| Bandwidth | 200 GB/s | 864 GB/s | 4.3x higher on L40S |
| Bus width | 128-bit | 384-bit | 3x wider on L40S |
| Compute · dense | |||
| FP8 | Not supported | 733 TFLOPS | L40S only |
| FP16 / BF16 | 17.9 TFLOPS | 362.1 TFLOPS | 20x higher on L40S |
| INT8 | 35.9 TOPS | 733 TOPS | 20x higher on L40S |
| FP32 | 4.5 TFLOPS | 91.6 TFLOPS | 20x higher on L40S |
| Platform | |||
| Interconnect | PCIe only | PCIe only | |
| Released | Q2 2021 | Q3 2023 | 2 years newer on L40S |
| Precision | |||
| Supported formats | FP16BF16TF32FP32INT4INT8 | FP8FP16BF16TF32FP32INT4INT8 | |
Compute: Vendor peak figures, dense and per GPU. Not measured throughput.
Source: A16 datasheet and L40S datasheet.
Where to rent them
Providers listing these GPUs, those with both first, so you can switch without moving clouds.
| Provider | A16 /GPU/hr | L40S /GPU/hr |
|---|---|---|
Sesterce
|
On-Demand $0.56 | On-Demand $0.97 |
Vultr
|
On-Demand $0.47 | -- |
Runcrate
|
On-Demand $0.56 | -- |
Lium
|
-- | On-Demand $0.38 |
Novita
|
-- | On-Demand $0.55 |
Vast.ai
|
-- | On-Demand $0.60 |
GPU.ai
|
-- | On-Demand $0.79 |
Runpod
|
-- | On-Demand $0.79 |
EmpirioLabs AI
|
-- | On-Demand $0.80 |
Packet·ai
|
-- | On-Demand $0.92 |
3 providers list the A16 and 36 the L40S. Every listing is on the A16 page and the L40S page.
Similar GPUs
Other comparisons in the same class.
Our data for Nvidia A16 was last updated on Sept. 7, 2026, and for Nvidia L40S on Sept. 7, 2026.