A100 vs L40S
Datasheet specs side by side, plus what each card actually costs to rent today across every provider we track.
A100
L40S
Spec sheet
| A100 | L40S | |
|---|---|---|
| Architecture | Ampere | Ada Lovelace |
| VRAM (max variant) | 80 GB | 48 GB |
| FP16 TFLOPS (dense) | 312 | 362 |
| Memory bandwidth | 1935 GB/s | 864 GB/s |
| TDP | 300 W | 350 W |
| Interconnect | PCIe Gen4 | PCIe Gen4 |
| Launched | 2020 | 2023 |
Computed price-performance at today's cheapest on-demand rates: A100 $2.85 per PFLOP-hour vs L40S $2.43. Formula: cheapest $/hr ÷ dense FP16 PFLOPS (1 PFLOP = 1000 TFLOPS). Specs from vendor datasheets (source, source).
What they cost to rent (2026-08-22)
| A100 | L40S | |
|---|---|---|
| Cheapest on-demand $/hr | $0.89 | $0.88 |
| Cheapest spot/community $/hr | $0.68 | $0.79 |
| Cheapest source | Jarvislabs | Massed Compute |
| Providers offering it | 21 | 16 |
Cheapest verified on-demand rate per side from today's snapshot; highlighted = cheaper. Spot/marketplace tiers shown on the individual pages.
FAQ
Is the A100 cheaper to rent than the L40S?
L40S is cheaper: the A100 rents from $0.89 per GPU-hour and the L40S from $0.88 on-demand, verified 2026-08-22.
Which has more VRAM, A100 or L40S?
The A100 has 80 GB and the L40S has 48 GB. The A100 has more.
Prices render from the live dataset each build. More: the screener · price movers · methodology.