AMD Radeon RX 6600 XT vs NVIDIA GeForce GTX 980 Ti Comparison

Head-to-Head Benchmarks

The recorded data shows a decisive sweep in the head-to-head comparison, with the AMD Radeon RX 6600 XT winning all four common benchmark tests. The largest margin appears in Geekbench Metal, where the RX 6600 XT scores 88,720 against the GTX 980 Ti's 19,520, a difference of 78%. This is an enormous gap, reflecting not just generational improvements but also the much newer architecture's ability to handle Metal workloads efficiently.

In Geekbench OpenCL, the RX 6600 XT posts 80,503 versus 43,513 for the GTX 980 Ti, a 45.9% advantage. This test is particularly relevant for compute-heavy tasks, and the data suggests the RDNA 2.0 part has a clear edge in raw parallel throughput. The Geekbench Vulkan result follows a similar pattern: the RX 6600 XT scores 72,417, while the GTX 980 Ti manages 47,724, a 34.1% lead for the AMD card. Vulkan is a modern API that both cards support at version 1.4, so this delta reflects actual rendering performance rather than API compatibility differences.

The smallest but still significant margin is in 3DMark Steel Nomad DX12, where the RX 6600 XT scores 1,804 and the GTX 980 Ti scores 1,321. The 26.8% difference here is notable because this is a modern DirectX 12 workload, and the older Maxwell card remains competitive relative to its other results, but it still trails by more than a quarter. Across all four tests, the RX 6600 XT wins by an average that is heavily weighted by the Metal outlier, but even the most conservative test shows a substantial lead. The GTX 980 Ti records zero wins in the head-to-head set, while the RX 6600 XT takes all four.

Architecture Differences

The two cards represent entirely different eras of GPU design. The NVIDIA GeForce GTX 980 Ti uses the GM200 chip built on Maxwell 2.0 architecture, fabricated on a 28 nm process at TSMC. The AMD Radeon RX 6600 XT uses the Navi 23 chip with RDNA 2.0 architecture, fabricated on a 7 nm process, also at TSMC. The process node difference is stark: 28 nm versus 7 nm, which explains much of the efficiency and clock speed gap between the two.

The transistor counts tell a story of density improvements. The GTX 980 Ti packs 8,000 million transistors across a 601 mm² die, yielding a transistor density of 13.3 million per square millimeter. The RX 980 Ti's die is far larger physically, but the RX 6600 XT crams 11,060 million transistors into just 237 mm², achieving 46.7 million per square millimeter. This is more than triple the density, a direct result of the smaller manufacturing process.

Clock speeds differ dramatically. The GTX 980 Ti runs at a 1000 MHz base with a 1076 MHz boost. The RX 6600 XT operates at a 1968 MHz base clock with a 2589 MHz boost and a 2359 MHz game clock. Even the RX 6600 XT's base clock is nearly double the GTX 980 Ti's boost clock. Memory also reflects the generational leap: the GTX 980 Ti uses 6 GB of GDDR5 at 1753 MHz (7 Gbps effective) across a 384 bit bus for 336.6 GB/s bandwidth, while the RX 6600 XT uses 8 GB of GDDR6 at 2000 MHz (16 Gbps effective) on a 128 bit bus for 256.0 GB/s. Interestingly, the GTX 980 Ti actually has higher raw memory bandwidth due to its much wider bus, but the RX 6600 XT compensates with faster memory technology.

The compute configuration differs as well. The GTX 980 Ti has 2,816 shading units, 176 TMUs, and 96 ROPs. The RX 6600 XT has 2,048 shading units, 128 TMUs, and 64 ROPs, but adds 32 ray accelerators. Despite fewer shading units, the RX 6600 XT produces higher pixel and texture rates: 165.7 GPixel/s and 331.4 GTexel/s versus 103.3 GPixel/s and 189.4 GTexel/s for the GTX 980 Ti. FP32 compute is also higher on the AMD card: 10.60 TFLOPS versus 6.060 TFLOPS. The RX 6600 XT also supports FP16 at 21.21 TFLOPS (2:1 ratio), a feature the GTX 980 Ti does not expose.

API support shows the age gap. The GTX 980 Ti supports DirectX 12 (12_1), while the RX 6600 XT supports DirectX 12 Ultimate (12_2). Both support OpenGL 4.6 and Vulkan 1.4. The RX 6600 XT also uses PCIe 4.0 x8, compared to the GTX 980 Ti's PCIe 3.0 x16, though the practical impact depends on the platform.

Where Each One Wins

The data offers a clear picture: the AMD Radeon RX 6600 XT wins every benchmark where both cards were tested. There are no tests in the database where the GTX 980 Ti comes out ahead in the head-to-head comparison. The RX 6600 XT's largest wins come in compute-oriented workloads, particularly Metal and OpenCL, where its modern architecture and higher FP32 throughput shine. The smallest win is in 3DMark Steel Nomad DX12, but even there the margin is 26.8%.

The GTX 980 Ti does have some theoretical advantages in the specification sheet. Its memory bandwidth of 336.6 GB/s is higher than the RX 6600 XT's 256.0 GB/s, and it has more ROPs (96 versus 64). In memory-intensive workloads that saturate the bus, the older card could potentially narrow the gap, but no benchmark in the recorded data shows this happening. The GTX 980 Ti's 384 bit bus is a legacy design choice that provided high bandwidth for its era, but the RX 6600 XT's faster GDDR6 and larger 8 GB frame buffer give it a capacity advantage that matters more for modern games.

For use cases, the RX 6600 XT is the pick for any application that leverages DirectX 12 Ultimate features, ray tracing (via its 32 ray accelerators), or FP16 compute. The GTX 980 Ti remains a capable card for older DirectX 11 titles or scenarios where its higher memory bandwidth might help, but the benchmark evidence does not support choosing it over the RX 6600 XT in any tested workload. The percentile rankings reinforce this: the GTX 980 Ti sits at the 73rd percentile of all GPUs, while the RX 6600 XT is at the 70th, but the average benchmark scores tell a different story, with the GTX 980 Ti averaging 28,020 versus 24,442 for the RX 6600 XT. This discrepancy is explained by the different benchmark sets recorded for each card in the database, which include additional legacy tests for the AMD card.

FAQ

Q: Which card wins in 3DMark Steel Nomad DX12?

A: The AMD Radeon RX 6600 XT wins with a score of 1,804 versus 1,321 for the NVIDIA GeForce GTX 980 Ti, a 26.8% advantage.

Q: How large is the performance gap in Geekbench Metal?

A: The RX 6600 XT scores 88,720, while the GTX 980 Ti scores 19,520. The RX 6600 XT leads by 78%, the largest margin of any head-to-head test.

Q: Does the GTX 980 Ti have higher memory bandwidth?

A: Yes. The GTX 980 Ti has 336.6 GB/s bandwidth over a 384 bit bus, while the RX 6600 XT has 256.0 GB/s over a 128 bit bus. However, the RX 6600 XT uses faster GDDR6 memory and has 8 GB versus 6 GB capacity.

Q: Which card supports DirectX 12 Ultimate?

A: Only the AMD Radeon RX 6600 XT supports DirectX 12 Ultimate (12_2). The NVIDIA GeForce GTX 980 Ti supports DirectX 12 (12_1).

Q: What is the average benchmark score for each card?

A: The GTX 980 Ti has an average benchmark score of 28,020, while the RX 6600 XT has an average of 24,442. Note that the database records different benchmark suites for each card, so this comparison is not direct.

Q: Which card has more shading units?

A: The GTX 980 Ti has 2,816 shading units, compared to 2,048 for the RX 6600 XT. Despite fewer shading units, the RX 6600 XT produces higher FP32 throughput: 10.60 TFLOPS versus 6.060 TFLOPS.

Specification Differences

| Specification | NVIDIA GeForce GTX 980 Ti | AMD Radeon RX 6600 XT |

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

| Architecture | Maxwell 2.0 | RDNA 2.0 |

| Process node | 28 nm | 7 nm |

| Transistors | 8,000 million | 11,060 million |

| Die size | 601 mm² | 237 mm² |

| Transistor density | 13.3M / mm² | 46.7M / mm² |

| Base clock | 1000 MHz | 1968 MHz |

| Boost clock | 1076 MHz | 2589 MHz |

| Game clock | None | 2359 MHz |

| Memory size | 6 GB | 8 GB |

| Memory type | GDDR5 | GDDR6 |

| Memory bus | 384 bit | 128 bit |

| Memory bandwidth | 336.6 GB/s | 256.0 GB/s |

| Shading units | 2816 | 2048 |

| TMUs | 176 | 128 |

| ROPs | 96 | 64 |

| Ray accelerators | None | 32 |

| Pixel rate | 103.3 GPixel/s | 165.7 GPixel/s |

| Texture rate | 189.4 GTexel/s | 331.4 GTexel/s |

| FP32 | 6.060 TFLOPS | 10.60 TFLOPS |

| FP16 | None | 21.21 TFLOPS (2:1) |

| TDP | 250 W | 160 W |

| Power connectors | 1x 6-pin + 1x 8-pin | 1x 8-pin |

| Suggested PSU | 600 W | 450 W |

| Bus interface | PCIe 3.0 x16 | PCIe 4.0 x8 |

| DirectX | 12 (12_1) | 12 Ultimate (12_2) |

| Release date | 2015-06-01 | 2021-07-29 |

| Launch MSRP | 649 USD | 379 USD |

The RX 6600 XT also has a smaller physical footprint at 190 mm length versus 267 mm for the GTX 980 Ti, though both are dual-slot cards with similar height and width. The display outputs differ: the GTX 980 Ti has 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2, while the RX 6600 XT has 1x HDMI 2.1 and 3x DisplayPort 1.4a.

The Verdict

The benchmark data is unambiguous: the AMD Radeon RX 6600 XT outperforms the NVIDIA GeForce GTX 980 Ti in every shared test, with margins ranging from 26.8% to 78%. The RX 6600 XT also draws less power (160 W versus 250 W), requires a smaller PSU (450 W versus 600 W), and offers modern features like DirectX 12 Ultimate and ray acceleration. Its 8 GB of GDDR6 memory doubles the capacity advantage for modern game textures, even though the GTX 980 Ti retains a bandwidth lead.

The GTX 980 Ti's only remaining advantages are its higher memory bandwidth and larger number of ROPs, but no recorded benchmark shows these translating into a win. Its 28 nm process and Maxwell 2.0 architecture place it firmly in an older era. For anyone choosing between these two based on the database measurements, the RX 6600 XT is the clear pick for every workload tested. The GTX 980 Ti might appeal to collectors or those running very old software that specifically benefits from its configuration, but the data does not support that choice for general use. The RX 6600 XT also launched at a lower MSRP, making its performance advantage even more decisive in absolute terms.