NVIDIA L40S vs NVIDIA RTX 4000 (Ada Generation), AI & Machine Learning Comparison
NVIDIA L40S wins 11 of 11 benchmarks, averaging 153.6% faster.
Both cards were measured first-party on our bench, same suite, same test rig.
What the numbers say
The gap is widest in LTX-Video (distilled), where NVIDIA L40S leads by 289% (8.01 vs 2.06 frames/s); the closest fight is Qwen3-4B (92% apart); VRAM decides part of this one: NVIDIA RTX 4000 (Ada Generation) runs 152 of our 12 AI workloads while the other card runs 6, models that don't fit score zero.
Benchmark results head-to-head
| Benchmark | NVIDIA L40S | NVIDIA RTX 4000 (Ada Generation) | Difference |
|---|---|---|---|
| Qwen3-4B tok/s | 212.05 | 110.18 | +92% |
| Llama-3.1-8B tok/s | 135.51 | 66.59 | +103% |
| Qwen2.5-Coder-14B tok/s | 74.6 | 36.34 | +105% |
| Qwen3-32B tok/s | 34.44 | 0 | n/a |
| Llama-3.3-70B tok/s | 16.49 | 0 | n/a |
| Stable Diffusion XL images/min | 17.02 | 6.56 | +159% |
| FLUX.1 dev images/min | 3.836 | 0 | n/a |
| FLUX.1 Kontext dev images/min | 1.693 | 0 | n/a |
| Qwen-Image-Edit images/min | 1.06 | 0 | n/a |
| LTX-Video (distilled) frames/s | 8.01 | 2.06 | +289% |
| Wan 2.2 5B (720p) frames/s | 0.49 | 0.18 | +172% |
Whole-job comparison
How long each card takes to finish a complete pipeline, not just one model. NVIDIA L40S is faster on 1 of 1; NVIDIA RTX 4000 (Ada Generation) on 0.
| Workflow | NVIDIA L40S | NVIDIA RTX 4000 (Ada Generation) | Difference | Cost per run |
|---|---|---|---|---|
| Full codebase review | 13.5 min | 27.5 min | NVIDIA L40S 2.04x faster | $0.177 vs $0.092 |
Renting by the hour, NVIDIA RTX 4000 (Ada Generation) finishes 1 of 1 cheaper. The quicker card is not automatically the cheaper way to get the work done.
Cost to rent
| Card | Per hour |
|---|---|
| NVIDIA L40S | $0.790 |
| NVIDIA RTX 4000 (Ada Generation) | $0.200 |
NVIDIA RTX 4000 (Ada Generation) is 3.95x cheaper per hour. Cheapest on-demand rate we see across RunPod and Vast.
Specifications compared
| NVIDIA L40S | NVIDIA RTX 4000 (Ada Generation) | |
|---|---|---|
| VRAM | 48GB | 20GB |
| Architecture | Ada Lovelace | Ada Lovelace |
| Memory bandwidth | 864 GB/s | 360 GB/s |
| Boost clock | 2,520 MHz | 2,175 MHz |
| TDP | 350 W | 130 W |
| Launch MSRP | $7,500 | $1,250 |
| Release | 2023-08-08 | 2023-08-09 |
FAQ
- Which is better for ai & machine learning: NVIDIA L40S or NVIDIA RTX 4000 (Ada Generation)?
- NVIDIA L40S performs better for ai & machine learning, winning 11 of 11 benchmarks in our suite with an average 153.6% advantage.
- What are the main hardware differences between NVIDIA L40S and NVIDIA RTX 4000 (Ada Generation)?
- NVIDIA L40S has 48GB VRAM and a 350W TDP, while NVIDIA RTX 4000 (Ada Generation) has 20GB VRAM and a 130W TDP.
- Does VRAM matter more than speed between NVIDIA L40S and NVIDIA RTX 4000 (Ada Generation)?
- For AI, yes, NVIDIA RTX 4000 (Ada Generation) fits 3 more of our 12 workloads than NVIDIA L40S. A model that exceeds VRAM doesn't run slower, it doesn't run at all, so the card that fits the model wins that workload outright.
- Where is the biggest performance difference between NVIDIA L40S and NVIDIA RTX 4000 (Ada Generation)?
- LTX-Video (distilled): NVIDIA L40S leads by roughly 289% (8.01 vs 2.06 frames/s) in our testing.
NVIDIA L40S full review · NVIDIA RTX 4000 (Ada Generation) full review · All AI & Machine Learning rankings