L40S vs A10G
Compare NVIDIA L40S and NVIDIA A10G specs, performance, and cloud pricing
L40S
48GB
From $0.820/hr
A10G
24GB
From $0.540/hr
Architecture
Ada Lovelace
vs Ampere
FP16 Gap
1.5x
L40S leads
| Specification | L40S | A10G |
|---|---|---|
| VRAM | 48 GB | 24 GB |
| VRAM Type | GDDR6X | GDDR6X |
| FP16 TFLOPS | 366.5 TFLOPS | 250 TFLOPS |
| FP8 TFLOPS | 733 TFLOPS | N/A |
| Memory Bandwidth | 864 GB/s | 600 GB/s |
| TDP | 350W | 150W |
| Interconnect | PCIe Gen4 | PCIe Gen4 |
| Architecture | Ada Lovelace | Ampere |
Price Comparison
| Metric | L40S | A10G |
|---|---|---|
| Cheapest On-Demand | $0.820/hr | $0.540/hr |
| Cheapest Spot | $0.440/hr | $0.270/hr |
| Providers Available | 5 | 2 |
Verdict
Best for Training
NVIDIA L40S
366.5 TFLOPS FP16 with 48GB VRAM
Best Value
NVIDIA A10G
463 TFLOPS per $/hr
Best for Inference
NVIDIA L40S
733 TFLOPS FP8/FP16
Use-Case Recommendations
Large-Scale Training
Training LLMs and large multi-modal models
Winner
L40S
366.5 TFLOPS FP16 with 48GB GDDR6X provides the best training throughput.
Inference at Scale
Deploying models in production for real-time inference
Winner
L40S
733 TFLOPS FP8/FP16 gives superior inference throughput.
Budget-Conscious Workloads
Getting the best performance per dollar
Winner
A10G
Starting at $0.540/hr delivers the best TFLOPS per dollar.