…
L40S vs A100 SXM 80GB
Cloud rental price, compute per dollar and specs, side by side.
Open in the interactive comparer →
L40S$0.60lowest on-demand per GPU·h
$1.66 per PFLOP·h · 17 providersA100 SXM 80GB$0.81lowest on-demand per GPU·h
$2.60 per PFLOP·h · 17 providersWhat actually differs
- The L40S has 1.2× more dense FP16/BF16 throughput than the A100 SXM 80GB (0.362 PF vs 0.312 PF).
- The A100 SXM 80GB has 2.4× more memory bandwidth than the L40S (864 vs 2,039 GB/s), which usually limits LLM token generation. Memory: 48 GB vs 80 GB.
- Only the L40S has native FP8 tensor cores.
- The L40S is 26% cheaper per hour than the A100 SXM 80GB at the lowest on-demand prices ($0.60 vs $0.81).
- Per unit of compute, the L40S is the better deal: $1.66 vs $2.60 per PFLOP-hour of dense FP16/BF16.
| Spec / price | L40S | A100 SXM 80GB |
|---|---|---|
| Memory | 48 GB GDDR6 | 80 GB HBM2ebetter |
| Memory bandwidth | 864 GB/s | 2,039 GB/sbetter |
| FP16 / BF16 (dense) | 0.362 PFbetter | 0.312 PF |
| FP8 (dense) | 0.733 PF | — |
| FP4 (dense) | — | — |
| INT8 (dense) | 0.733 PFbetter | 0.624 PF |
| TF32 (dense) | 0.183 PFbetter | 0.156 PF |
| FP32 (dense) | 0.0916 PFbetter | 0.0195 PF |
| FP64 (dense) | 0.0014 PF | 0.0195 PFbetter |
| Power | 350 W | 400 W |
| Architecture | Ada Lovelace (2023) | Ampere (2020) |
| Lowest on-demand | $0.60better | $0.81 |
| Median on-demand | $1.50better | $2.35 |
| Lowest spot | $0.38better | $0.90 |
| FP16/BF16 $ per PFLOP·h | $1.66better | $2.60 |
| Providers | 17 | 17 |
L40S vs A100 SXM 80GB: questions
Is the L40S faster than the A100 SXM 80GB?
The L40S has 1.2× more dense FP16/BF16 compute than the A100 SXM 80GB. The A100 SXM 80GB has 2.4× more memory bandwidth than the L40S. Real-world speedups depend on the workload: training and prefill scale with compute, while LLM decoding scales mostly with memory bandwidth.
Which is cheaper per FLOP, the L40S or the A100 SXM 80GB?
Right now the L40S is cheaper per unit of compute: $1.66 vs $2.60 per PFLOP-hour at the lowest on-demand prices.
How much does it cost to rent an L40S vs an A100 SXM 80GB for a month?
Running one GPU around the clock at the lowest on-demand price costs about $440 per month for the L40S and $591 for the A100 SXM 80GB.
Which has more memory, the L40S or the A100 SXM 80GB?
The A100 SXM 80GB has more memory: 80 GB vs 48 GB.