L40S vs A40
Datasheet specs side by side, plus what each card actually costs to rent today across every provider we track.
L40S
A40
Spec sheet
| L40S | A40 | |
|---|---|---|
| Architecture | Ada Lovelace | Ampere |
| VRAM (max 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 |
Computed price-performance at today's cheapest on-demand rates: L40S $2.43 per PFLOP-hour vs A40 $2.93. Formula: cheapest $/hr ÷ dense FP16 PFLOPS (1 PFLOP = 1000 TFLOPS). Specs from vendor datasheets (source, source).
What they cost to rent (2026-07-18)
| L40S | A40 | |
|---|---|---|
| Cheapest on-demand $/hr | $0.88 | $0.44 |
| Cheapest spot/community $/hr | $0.79 | $0.35 |
| Cheapest source | Massed Compute | RunPod |
| Providers offering it | 12 | 2 |
Cheapest verified on-demand rate per side from today's snapshot; highlighted = cheaper. Spot/marketplace tiers shown on the individual pages.
FAQ
Is the L40S cheaper to rent than the A40?
A40 is cheaper: the L40S rents from $0.88 per GPU-hour and the A40 from $0.44 on-demand, verified 2026-07-18.
Which has more VRAM, L40S or A40?
The L40S has 48 GB and the A40 has 48 GB. They match on maximum VRAM.
Prices render from the live dataset each build. More: the screener · price movers · methodology.