NVIDIA GeForce RTX 4080 SUPER vs NVIDIA GeForce RTX 5070 Ti Comparison

The NVIDIA GeForce RTX 4080 SUPER and the NVIDIA GeForce RTX 5070 Ti represent two distinct generations of NVIDIA's flagship-tier offerings. The benchmark data shows a clear split: the RTX 4080 SUPER, built on Ada Lovelace, wins the majority of head-to-head tests, while the RTX 5070 Ti, built on Blackwell 2.0, counters with victories in specific compute and 2D workloads. Both cards sit at the 86th percentile among all GPUs, but their average benchmark scores diverge significantly, with the older card holding a higher overall average.

Head-to-Head Benchmarks

The most dramatic margin in the entire comparison belongs to the RTX 4080 SUPER in the Geekbench Vulkan test. It scores 260,075 against the RTX 5070 Ti's 225,122, a gap of 15.5%. That is not a marginal win; it is a substantial lead in a modern graphics API workload. The RTX 4080 SUPER also takes the Geekbench OpenCL test with 219,065 versus 212,363, a 3.2% advantage. These two results alone establish that the older card holds a clear edge in general-purpose compute tasks that leverage these APIs.

In the DirectX legacy tests, the RTX 4080 SUPER wins all four, but the margins vary. The largest of these is in Passmark DirectX 9, where the RTX 4080 SUPER scores 381 against 351, an 8.5% lead. The DirectX 12 test shows a 5.5% advantage (134 vs. 127), while the DirectX 10 and DirectX 11 tests are closer, with the RTX 4080 SUPER leading by just 0.5% (193 vs. 192) and 0.3% (301 vs. 300), respectively. These results suggest that the architectural advantages of the RTX 4080 SUPER become more pronounced in older, less parallelized workloads.

The Passmark G3D score, a holistic gaming-oriented metric, also favors the RTX 4080 SUPER. It posts 34,245 points against the RTX 5070 Ti's 32,974, a 3.9% victory. This aligns with the pattern seen in the DirectX tests, reinforcing the idea that the RTX 4080 SUPER is the stronger raw rasterizer in most scenarios.

However, the RTX 5070 Ti does secure three wins. Its most notable victory is in Passmark G2D, a 2D graphics test, where it scores 1,332 against the RTX 4080 SUPER's 1,270, a 4.7% margin. It also wins the Passmark GPU Compute test with 20,203 points versus 19,822, a 1.9% advantage. Finally, in the 3DMark Steel Nomad DX12 test, the two cards are virtually tied, with the RTX 5070 Ti scoring 6,604 and the RTX 4080 SUPER scoring 6,600, a razor-thin 0.1% difference that mathematically goes to the newer card.

The overall tally is 7 wins for the RTX 4080 SUPER and 3 for the RTX 5070 Ti. The average benchmark scores reflect this: the RTX 4080 SUPER averages 54,209 points, while the RTX 5070 Ti averages 49,957. This is an 8.5% difference in the average, which is a significant gap when considering the two cards' near-identical performance in the most modern synthetic test.

Architecture Differences

The two GPUs come from different architectures and chips. The RTX 4080 SUPER is based on the AD103 chip using the Ada Lovelace architecture, while the RTX 5070 Ti uses the GB203 chip with the Blackwell 2.0 architecture. Both are fabricated on a 5 nm process at TSMC, and their transistor counts are nearly identical: 45,900 million for the RTX 4080 SUPER and 45,600 million for the RTX 5070 Ti. The die sizes are also close, at 379 mm² and 378 mm², respectively, leading to a marginal transistor density difference of 121.1M / mm² versus 120.6M / mm².

The core configurations differ substantially. The RTX 4080 SUPER has 10,240 shading units, 320 TMUs, and 112 ROPs, while the RTX 5070 Ti has 8,960 shading units, 280 TMUs, and 96 ROPs. This represents a 14.3% reduction in shading units for the newer card. The RTX 4080 SUPER also has more RT cores (80 vs. 70) and more tensor cores (320 vs. 280). Consequently, the RTX 4080 SUPER's peak FP32 throughput is 52.22 TFLOPS compared to the RTX 5070 Ti's 43.94 TFLOPS. The same 1:1 ratio applies to FP16 performance for both cards.

The memory subsystems show a different story. Both cards feature 16 GB of VRAM on a 256-bit bus, but the type and speed differ. The RTX 4080 SUPER uses GDDR6X at 23 Gbps effective, yielding 736.3 GB/s of bandwidth. The RTX 5070 Ti uses GDDR7 at 28 Gbps effective, yielding a much higher 896.0 GB/s. This 21.7% bandwidth advantage for the RTX 5070 Ti does not translate into a win in most gaming tests, but it likely contributes to its win in the GPU compute test.

Clock speeds are similar at the base, with both at 2295 MHz. However, the boost clocks differ, with the RTX 4080 SUPER boosting to 2550 MHz and the RTX 5070 Ti to 2452 MHz. This higher boost clock helps explain the RTX 4080 SUPER's raw performance advantage. The RTX 5070 Ti has a lower TDP of 300 W versus 320 W for the RTX 4080 SUPER, and it is a dual-slot card instead of a triple-slot card. The RTX 5070 Ti also uses a newer PCIe 5.0 x16 interface, while the RTX 4080 SUPER is limited to PCIe 4.0 x16.

The Verdict

The data presents a straightforward picture for most users. The RTX 4080 SUPER wins 7 of 10 head-to-head benchmarks, including all DirectX tests and both Geekbench compute tests. Its average benchmark score of 54,209 is significantly higher than the RTX 5070 Ti's 49,957. If the priority is maximum performance in legacy DirectX APIs, general compute via OpenCL and Vulkan, and raw 3D rendering as measured by Passmark G3D, the RTX 4080 SUPER is the clear choice. Its 15.5% lead in Vulkan and 8.5% lead in DirectX 9 are decisive.

However, the RTX 5070 Ti is not without merit. It matches the RTX 4080 SUPER in the 3DMark Steel Nomad DX12 test, a modern workload, with a score of 6,604 versus 6,600. It also wins in 2D performance and GPU compute, suggesting that its Blackwell architecture and faster GDDR7 memory provide advantages in specific tasks. Its lower TDP and dual-slot design are also practical benefits. For users focused on the most modern API performance, 2D workloads, or pure compute, the RTX 5070 Ti is a viable alternative, even if its overall average score is lower.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce RTX 4080 SUPER has a higher average benchmark score of 54,209, compared to the RTX 5070 Ti's 49,957.

Q: In which test does the RTX 4080 SUPER have its largest margin of victory?

A: The RTX 4080 SUPER's largest win is in the Geekbench Vulkan test, where it scores 260,075 versus the RTX 5070 Ti's 225,122, a 15.5% difference.

Q: Does the RTX 5070 Ti win any benchmark tests?

A: Yes, the RTX 5070 Ti wins three tests: Passmark G2D (1,332 vs. 1,270), Passmark GPU Compute (20,203 vs. 19,822), and 3DMark Steel Nomad DX12 (6,604 vs. 6,600).

Q: How do their memory bandwidths compare?

A: The RTX 5070 Ti has a higher memory bandwidth of 896.0 GB/s using GDDR7, while the RTX 4080 SUPER has 736.3 GB/s using GDDR6X.

Q: What is the difference in their shading unit counts?

A: The RTX 4080 SUPER has 10,240 shading units, while the RTX 5070 Ti has 8,960 shading units.

Q: Which card has a higher boost clock speed?

A: The RTX 4080 SUPER has a higher boost clock of 2550 MHz, compared to the RTX 5070 Ti's 2452 MHz.

Where Each One Wins

The RTX 4080 SUPER wins in every DirectX benchmark category, including a notable 8.5% lead in DirectX 9. It also dominates in Geekbench OpenCL (3.2% lead) and Vulkan (15.5% lead). Its win in Passmark G3D (3.9% lead) confirms its strength in general 3D rendering. This card is the superior choice for users prioritizing legacy DirectX titles, OpenCL compute, or Vulkan-based applications. Its higher shading unit count and boost clock are the primary drivers of this performance profile.

The RTX 5070 Ti wins in Passmark G2D, indicating better performance in 2D graphical operations. It also wins in Passmark GPU Compute, a raw computational workload, and it edges out the RTX 4080 SUPER in the modern 3DMark Steel Nomad DX12 test, albeit by a negligible 0.1%. This suggests that the RTX 5070 Ti's architecture and faster GDDR7 memory make it more competitive in current-generation APIs and compute-heavy tasks. Its lower TDP of 300 W and dual-slot form factor also make it a more physically accommodating card.

Specification Differences

| Specification | NVIDIA GeForce RTX 4080 SUPER | NVIDIA GeForce RTX 5070 Ti |

|---|---|---|

| Architecture | Ada Lovelace | Blackwell 2.0 |

| Chip | AD103 | GB203 |

| Process Node | 5 nm | 5 nm |

| Transistors | 45,900 million | 45,600 million |

| Die Size | 379 mm² | 378 mm² |

| Shading Units | 10240 | 8960 |

| TMUs | 320 | 280 |

| ROPs | 112 | 96 |

| RT Cores | 80 | 70 |

| Tensor Cores | 320 | 280 |

| Base Clock | 2295 MHz | 2295 MHz |

| Boost Clock | 2550 MHz | 2452 MHz |

| Memory Type | GDDR6X | GDDR7 |

| Memory Speed | 23 Gbps effective | 28 Gbps effective |

| Memory Bandwidth | 736.3 GB/s | 896.0 GB/s |

| FP32 Performance | 52.22 TFLOPS | 43.94 TFLOPS |

| Pixel Rate | 285.6 GPixel/s | 235.4 GPixel/s |

| Texture Rate | 816.0 GTexel/s | 686.6 GTexel/s |

| TDP | 320 W | 300 W |

| Slot Width | Triple-slot | Dual-slot |

| Bus Interface | PCIe 4.0 x16 | PCIe 5.0 x16 |

| Display Outputs | 1x HDMI 2.1, 3x DisplayPort 1.4a | 1x HDMI 2.1b, 3x DisplayPort 2.1b |

| Dimensions (LxHxW) | 310 mm x 140 mm x 61 mm | 304 mm x 137 mm x 48 mm |

| Production Status | End-of-life | Active |

| Launch MSRP | 999 USD | 749 USD |