L40S vs A40
L40S starts at $0.97/GPU-hour on-demand; A40 starts at $0.59/GPU-hour on-demand. Compare the exact variants, memory and published specifications below. GPU spec numbers are not benchmarks.
39.2% cheaper at A40.
checked 16 October 2026. Price alone does not establish equivalent capacity or performance.
L40S
A40
Spec sheet
| L40S | A40 | |
|---|---|---|
| Spec variant | L40S | A40 |
| Architecture | Ada Lovelace | Ampere |
| VRAM (spec variant) | 48 GB | 48 GB |
| FP16 TFLOPS (dense) | 362 | 150 |
| Memory bandwidth | 864 GB/s | 696 GB/s |
| TDP | 350 W | 300 W |
| Interconnect | PCIe Gen4 | NVLink 112.5GB/s |
| Launched | 2023 | 2020 |
Rental price per manufacturer-rated FP16 compute: L40S $2.68 per PFLOP-hour vs A40 $3.93. Formula: cheapest $/hr ÷ dense FP16 PFLOPS (1 PFLOP = 1000 TFLOPS). This uses L40S and A40, whose datasheets may differ from the variants above. It is not measured job throughput. Specs: L40S manufacturer, A40 manufacturer.
What they cost to rent (16 October 2026)
| L40S | A40 | |
|---|---|---|
| Cheapest on-demand $/hr | $0.97 | $0.59 |
| Cheapest spot/community $/hr | $0.79 | – |
| Cheapest source | Massed Compute | RunPod |
| Providers offering it | 16 | 2 |
Cheapest on-demand rate; the cheaper one is highlighted. Spot/marketplace tiers shown on the individual pages.
FAQ
A40 is cheaper: the L40S rents from $0.97 per GPU-hour and the A40 from $0.59 on-demand, checked 16 October 2026.
The largest recorded L40S variant has 48 GB and the largest A40 variant has 48 GB. Other variants may have less memory. They match on maximum VRAM.
Compare more rental options: the screener · price movers · methodology.