

NVIDIA L40S wins 11 of 11 benchmarks, averaging 43% faster.
Both cards were measured first-party on our bench, same suite, same test rig.
The gap is widest in Qwen2.5-Coder 14B, where NVIDIA L40S leads by 46% (51.23 vs 74.6 tok/s); the closest fight is Qwen3 4B (39% apart).
| Benchmark | NVIDIA GeForce RTX 4070 Super | NVIDIA L40S | Difference |
|---|---|---|---|
| Qwen3 4B tok/s | 152.84 | 212.05 | -28% |
| Llama 3.1 8B tok/s | 93.62 | 135.51 | -31% |
| Qwen2.5-Coder 14B tok/s | 51.23 | 74.6 | -31% |
| Qwen3 32B tok/s | 0 | 34.44 | n/a |
| Llama 3.3 70B tok/s | 0 | 16.49 | n/a |
| Z-Image Turbo images/min | 0 | 8.325 | n/a |
| FLUX.1 dev images/min | 0 | 3.836 | n/a |
| FLUX.1 Kontext dev images/min | 0 | 1.693 | n/a |
| Qwen-Image-Edit images/min | 0 | 1.06 | n/a |
| LTX-Video (distilled) frames/s | 0 | 8.01 | n/a |
| Wan 2.2 5B (720p) frames/s | 0 | 0.49 | n/a |
How long each card takes to finish a complete pipeline, not just one model. NVIDIA GeForce RTX 4070 Super is faster on 0 of 1; NVIDIA L40S on 1.
| Workflow | NVIDIA GeForce RTX 4070 Super | NVIDIA L40S | Difference |
|---|---|---|---|
| Full codebase review | 19.5 min | 13.5 min | NVIDIA L40S 1.45x faster |
| NVIDIA GeForce RTX 4070 Super | NVIDIA L40S | |
|---|---|---|
| VRAM | 12GB | 48GB |
| Architecture | Ada Lovelace | Ada Lovelace |
| Memory bandwidth | 504 GB/s | 864 GB/s |
| Boost clock | 2,475 MHz | 2,520 MHz |
| TDP | 220 W | 350 W |
| Launch MSRP | $599 | $7,500 |
| Release | 2024-01-17 | 2023-08-08 |
NVIDIA GeForce RTX 4070 Super full review · NVIDIA L40S full review · All AI & Machine Learning rankings