Side-by-side GPU comparison: specs, memory, compute performance, and live cloud pricing.
Verdict
L40 has 16 GB more VRAM, making it better suited for large models and long context windows. For compute-bound workloads like training, RTX 5090 delivers 1.2× higher FP16 throughput. At $0.44/hr vs $0.99/hr, RTX 5090 is the more cost-efficient choice for inference. L40 supports a broader range of models (624 vs 577 from this catalog), giving more flexibility.
Specifications
| L40 | RTX 5090 | |
|---|---|---|
| VRAM | 48 GB | 32 GB |
| VRAM Type | GDDR6 | GDDR7 |
| Memory Bandwidth | 0.9 TB/s | 1.8 TB/s |
| FP16 Performance | 181 TFLOPS | 210 TFLOPS |
| Manufacturer | NVIDIA | NVIDIA |
| FP8 Support | Yes | Yes |
| FP4 Support | No | Yes |
Price / Performance
Based on cheapest single-GPU on-demand pricing. Lower $/TFLOP = better compute value; lower $/GB = better memory value.
| L40 | RTX 5090 | |
|---|---|---|
| $/hr (cheapest) | $0.99 | $0.44✓ best |
| $/TFLOP (compute value) | $0.0055 | $0.0021✓ best |
| $/GB VRAM (memory value) | $0.0206 | $0.0138✓ best |
Cloud Pricing
Cheapest on-demand price per provider (single GPU).
Model Compatibility
Models from the catalog that fit on each GPU, grouped by required precision.
L40 (624 models)
RTX 5090 (577 models)
You might also compare…
Pricing data refreshed hourly · Last updated April 11, 2026 · Browse all comparisons