…
RTX 4090 vs A100 SXM 80GB
Cloud rental price, compute per dollar and specs, side by side.
Open in the interactive comparer →
RTX 4090$0.29lowest on-demand per GPU·h
$1.78 per PFLOP·h · 6 providersA100 SXM 80GB$0.81lowest on-demand per GPU·h
$2.60 per PFLOP·h · 17 providersWhat actually differs
- The A100 SXM 80GB has 1.9× more dense FP16/BF16 throughput than the RTX 4090 (0.165 PF vs 0.312 PF).
- The A100 SXM 80GB has 2.0× more memory bandwidth than the RTX 4090 (1,008 vs 2,039 GB/s), which usually limits LLM token generation. Memory: 24 GB vs 80 GB.
- Only the RTX 4090 has native FP8 tensor cores.
- The RTX 4090 is 64% cheaper per hour than the A100 SXM 80GB at the lowest on-demand prices ($0.29 vs $0.81).
- Per unit of compute, the RTX 4090 is the better deal: $1.78 vs $2.60 per PFLOP-hour of dense FP16/BF16.
| Spec / price | RTX 4090 | A100 SXM 80GB |
|---|---|---|
| Memory | 24 GB GDDR6X | 80 GB HBM2ebetter |
| Memory bandwidth | 1,008 GB/s | 2,039 GB/sbetter |
| FP16 / BF16 (dense) | 0.165 PF | 0.312 PFbetter |
| FP8 (dense) | 0.33 PF | — |
| FP4 (dense) | — | — |
| INT8 (dense) | 0.661 PFbetter | 0.624 PF |
| TF32 (dense) | 0.0826 PF | 0.156 PFbetter |
| FP32 (dense) | 0.0826 PFbetter | 0.0195 PF |
| FP64 (dense) | 0.0013 PF | 0.0195 PFbetter |
| Power | 450 W | 400 W |
| Architecture | Ada Lovelace (2022) | Ampere (2020) |
| Lowest on-demand | $0.29better | $0.81 |
| Median on-demand | $0.39better | $2.35 |
| Lowest spot | $0.16better | $0.90 |
| FP16/BF16 $ per PFLOP·h | $1.78better | $2.60 |
| Providers | 6 | 17 |
RTX 4090 vs A100 SXM 80GB: questions
Is the RTX 4090 faster than the A100 SXM 80GB?
The A100 SXM 80GB has 1.9× more dense FP16/BF16 compute than the RTX 4090. The A100 SXM 80GB has 2.0× more memory bandwidth than the RTX 4090. 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 RTX 4090 or the A100 SXM 80GB?
Right now the RTX 4090 is cheaper per unit of compute: $1.78 vs $2.60 per PFLOP-hour at the lowest on-demand prices.
How much does it cost to rent an RTX 4090 vs an A100 SXM 80GB for a month?
Running one GPU around the clock at the lowest on-demand price costs about $215 per month for the RTX 4090 and $591 for the A100 SXM 80GB.
Which has more memory, the RTX 4090 or the A100 SXM 80GB?
The A100 SXM 80GB has more memory: 80 GB vs 24 GB.