

NVIDIA GeForce RTX 4070 Super wins 3 of 11 benchmarks, averaging 18.3% faster.
Both cards were measured first-party on our bench, same suite, same test rig.
The gap is widest in Qwen3 4B, where NVIDIA GeForce RTX 4070 Super leads by 21% (126.36 vs 152.84 tok/s); the closest fight is Llama 3.1 8B (16% apart); VRAM decides part of this one: NVIDIA GeForce RTX 4070 Super runs 3 of our 12 AI workloads while the other card runs 3, models that don't fit score zero.
| Benchmark | NVIDIA GeForce RTX 3070 Founders Edition | NVIDIA GeForce RTX 4070 Super | Difference |
|---|---|---|---|
| Qwen3 4B tok/s | 126.36 | 152.84 | -17% |
| Llama 3.1 8B tok/s | 80.95 | 93.62 | -14% |
| Qwen2.5-Coder 14B tok/s | 0 | 51.23 | n/a |
| Qwen3 32B tok/s | 0 | 0 | n/a |
| Llama 3.3 70B tok/s | 0 | 0 | n/a |
| Z-Image Turbo images/min | 0 | 0 | n/a |
| FLUX.1 dev images/min | 0 | 0 | n/a |
| FLUX.1 Kontext dev images/min | 0 | 0 | n/a |
| Qwen-Image-Edit images/min | 0 | 0 | n/a |
| LTX-Video (distilled) frames/s | 0 | 0 | n/a |
| Wan 2.2 5B (720p) frames/s | 0 | 0 | n/a |
| NVIDIA GeForce RTX 3070 Founders Edition | NVIDIA GeForce RTX 4070 Super | |
|---|---|---|
| VRAM | 8GB | 12GB |
| Transistors | 17,400M | 35,800M |
| Die size | 392.5 mm² | 294.5 mm² |
| Transistor density | 44.3 M/mm² | 121.6 M/mm² |
| Architecture | Ampere | Ada Lovelace |
| Memory bandwidth | 448 GB/s | 504 GB/s |
| Boost clock | 1,725 MHz | 2,475 MHz |
| TDP | 220 W | 220 W |
| Launch MSRP | $499 | $599 |
| Release | 2020-10-29 | 2024-01-17 |
NVIDIA GeForce RTX 3070 Founders Edition full review · NVIDIA GeForce RTX 4070 Super full review · All AI & Machine Learning rankings