AMD Radeon RX 6800 vs NVIDIA GeForce GTX 980 Ti Comparison

# AMD Radeon RX 6800 vs NVIDIA GeForce GTX 980 Ti

The AMD Radeon RX 6800 is decisively the faster card, winning three of four head-to-head benchmark comparisons and delivering lead margins that range from 141% to 687%. The GTX 980 Ti retains one narrow victory in OpenCL compute, but that result does not offset its overall standing: the RX 6800 sits at the 75th percentile of all GPUs versus the GTX 980 Ti's 73rd, a gap that understates how lopsided the modern workload results actually are.

FAQ

Q: Which card wins more head-to-head benchmarks?

A: The AMD Radeon RX 6800 wins three of the four head-to-head tests, with victories in 3DMark Steel Nomad DX12, Geekbench Metal, and Geekbench Vulkan. The NVIDIA GeForce GTX 980 Ti wins only the Geekbench OpenCL test.

Q: What is the biggest performance gap between the two cards?

A: The largest gap is in Geekbench Metal, where the RX 6800 scores 153,635 versus 19,520 for the GTX 980 Ti, a lead of 687.1%. The smallest gap is in Geekbench OpenCL, where the GTX 980 Ti leads by 43.7%.

Q: How do their average benchmark scores compare?

A: The RX 6800 has an average benchmark score of 30,095, while the GTX 980 Ti averages 28,020. This places the RX 6800 1.1% ahead of the AMD Radeon RX 6700 and 0.5% ahead of the NVIDIA GeForce RTX 3070 Ti.

Q: What generation and architecture does each card use?

A: The RX 6800 uses the RDNA 2.0 architecture on a 7 nm process, part of the Navi II (RX 6000) generation. The GTX 980 Ti uses Maxwell 2.0 on a 28 nm process, part of the GeForce 900 generation.

Q: Do both cards support the same API levels?

A: No. The RX 6800 supports DirectX 12 Ultimate (12_2), while the GTX 980 Ti supports DirectX 12 (12_1). Both cards support OpenGL 4.6 and Vulkan 1.4.

Q: Which card has more memory and bandwidth?

A: The RX 6800 has 16 GB of GDDR6 memory on a 256-bit bus with 512.0 GB/s bandwidth. The GTX 980 Ti has 6 GB of GDDR5 memory on a 384-bit bus with 336.6 GB/s bandwidth.

Architecture Differences

The architectural gap between these two cards spans five years of GPU development, and the data reflects a fundamental shift in design priorities. The RX 6800 is built on TSMC's 7 nm process, packing 26,800 million transistors into a 520 mm² die, a transistor density of 51.5M per mm². The GTX 980 Ti uses the same foundry but a much older 28 nm process, with 8,000 million transistors spread across a larger 601 mm² die, yielding a density of only 13.3M per mm². That density difference explains why the RX 6800 delivers far higher raw throughput despite having a smaller physical die.

The RX 6800's RDNA 2.0 architecture introduces hardware ray tracing with 60 dedicated RT cores, a feature entirely absent from the GTX 980 Ti's Maxwell 2.0 design. The RX 6800 also supports DirectX 12 Ultimate (12_2), whereas the GTX 980 Ti tops out at DirectX 12 (12_1), meaning the newer card can handle the full feature set of modern DX12 titles including variable rate shading and mesh shaders. The older card has no ray tracing capability and no equivalent to the RX 6800's 60 RT cores.

Clock speeds tell a similar story of generational progress. The RX 6800 runs at a 1700 MHz base and boosts to 2105 MHz, with a game clock of 1815 MHz. The GTX 980 Ti operates at 1000 MHz base and 1076 MHz boost, with no game clock defined. Memory clocks also diverge sharply: the RX 6800 uses 2000 MHz memory with 16 Gbps effective data rate, while the GTX 980 Ti's memory runs at 1753 MHz with 7 Gbps effective. This, combined with the RX 6800's modern GDDR6 versus the GTX 980 Ti's GDDR5, drives a 512.0 GB/s versus 336.6 GB/s bandwidth advantage.

The RX 6800 also offers more of every compute resource: 3840 shading units versus 2816, 240 texture mapping units versus 176, and an identical 96 ROPs. Pixel rate on the RX 6800 hits 202.1 GPixel/s versus 103.3 GPixel/s, and texture rate reaches 505.2 GTexel/s versus 189.4 GTexel/s. FP32 compute is 16.17 TFLOPS versus 6.060 TFLOPS, and the RX 6800 lists FP16 at 32.33 TFLOPS (2:1) while the GTX 980 Ti has no FP16 specification. Both cards are rated at 250 W TDP with a 600 W suggested PSU, though the RX 6800 requires two 8-pin connectors versus the GTX 980 Ti's single 6-pin plus 8-pin configuration.

Head-to-Head Benchmarks

The 3DMark Steel Nomad DX12 test shows a 141.3% lead for the RX 6800, scoring 3,188 versus 1,321. This is a modern DirectX 12 workload, and the result aligns with the architectural advantages: the RX 6800's newer process node, higher clocks, and DX12 Ultimate support all contribute to a score that is more than double the GTX 980 Ti's. The RX 6800's 16.17 TFLOPS FP32 throughput versus 6.060 TFLOPS is the raw compute engine behind this margin.

Geekbench Vulkan shows a nearly identical margin at 141.2%, with the RX 6800 scoring 115,107 against 47,724. Vulkan is another modern API, and the RX 6800's RDNA 2.0 architecture was designed with Vulkan 1.4 support in mind. The GTX 980 Ti also supports Vulkan 1.4, but its older Maxwell architecture cannot extract the same level of performance from the API, resulting in a score that is less than half of the RX 6800's.

The most extreme result is Geekbench Metal, where the RX 6800 scores 153,635 versus 19,520, a 687.1% lead. Metal is Apple's graphics API, and this test measures both raw compute and memory bandwidth efficiency. The RX 6800's 512.0 GB/s bandwidth and 16 GB GDDR6 memory give it a massive advantage over the GTX 980 Ti's 336.6 GB/s and 6 GB GDDR5. This is the single most lopsided comparison in the dataset, indicating that the RX 6800 is dramatically better suited to Metal-based workloads.

The one test the GTX 980 Ti wins is Geekbench OpenCL, scoring 43,513 versus 24,508 for the RX 6800, a 43.7% lead. OpenCL is an older, more CPU-driven compute standard, and the GTX 980 Ti's Maxwell architecture was optimized for this workload class. The RX 6800's RDNA 2.0 design appears to prioritize newer APIs, and its OpenCL performance lags its DirectX and Vulkan results. This is the only bright spot for the GTX 980 Ti, and it is a narrow one compared to the RX 6800's dominant showings elsewhere.

Specification Differences

The two cards differ across nearly every specification field. The RX 6800 uses a 7 nm process node; the GTX 980 Ti uses 28 nm. Transistor count is 26,800 million versus 8,000 million, with die size at 520 mm² versus 601 mm². The RX 6800 has 3840 shading units, 240 TMUs, 96 ROPs, and 60 RT cores; the GTX 980 Ti has 2816 shading units, 176 TMUs, 96 ROPs, and no RT cores. Base clock is 1700 MHz versus 1000 MHz, boost clock is 2105 MHz versus 1076 MHz, and the RX 6800 adds a 1815 MHz game clock not present on the GTX 980 Ti. Memory is 16 GB GDDR6 versus 6 GB GDDR5, with bus width 256-bit versus 384-bit, and bandwidth 512.0 GB/s versus 336.6 GB/s. Pixel rate is 202.1 versus 103.3 GPixel/s, texture rate is 505.2 versus 189.4 GTexel/s, and FP32 is 16.17 versus 6.060 TFLOPS. The RX 6800 lists FP16 at 32.33 TFLOPS (2:1); the GTX 980 Ti has no FP16 specification. Power connectors differ: two 8-pin on the RX 6800 versus one 6-pin plus one 8-pin on the GTX 980 Ti. Bus interface is PCIe 4.0 x16 versus PCIe 3.0 x16. Display outputs are 1x HDMI 2.1, 2x DisplayPort 1.4a, and 1x USB Type-C on the RX 6800 versus 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2 on the GTX 980 Ti. The RX 6800 is 120 mm tall versus 111 mm; both are 267 mm long and 40 mm wide. DirectX support is 12 Ultimate (12_2) versus 12 (12_1), with both at OpenGL 4.6 and Vulkan 1.4.

Where Each One Wins

The RX 6800 wins in every modern workload category represented in the benchmarks. In DirectX 12, the 3DMark Steel Nomad result shows a 141.3% advantage, making it the clear choice for current and upcoming DX12 titles. In Vulkan, the 141.2% lead in Geekbench Vulkan reinforces this pattern, suggesting that the RX 6800 is the better option for games and applications built on either modern API. In Metal, the 687.1% margin is so large that the RX 6800 is effectively in a different performance class, which matters for macOS or iOS-adjacent workflows that rely on Metal acceleration.

The GTX 980 Ti wins only in OpenCL compute, where its 43.7% lead over the RX 6800 indicates that this older API still plays to Maxwell's strengths. For users running legacy OpenCL workloads that cannot migrate to Vulkan or DirectX 12, the GTX 980 Ti offers better performance in that specific test. However, this is a single data point against three substantial defeats, and the GTX 980 Ti's overall average score of 28,020 versus 30,095 for the RX 6800 reflects its narrower applicability.

The RX 6800 also wins on memory capacity and bandwidth, with 16 GB versus 6 GB and 512.0 GB/s versus 336.6 GB/s. This makes it more suitable for high-resolution textures and large datasets, while the GTX 980 Ti's smaller frame buffer could become a bottleneck in memory-intensive scenarios. The RX 6800's ray tracing hardware and DX12 Ultimate support further extend its useful lifespan, whereas the GTX 980 Ti lacks these features entirely.

The Verdict

The data is unambiguous: the AMD Radeon RX 6800 is the superior card for nearly all workloads. It wins three of four head-to-head benchmarks, with margins ranging from 141.2% to 687.1%, and it holds a higher average benchmark score at 30,095 versus 28,020. The RX 6800 also ranks at the 75th percentile of all GPUs, two points above the GTX 980 Ti's 73rd, and its nearest rival comparison shows it trading blows with the RTX 3070 Ti and RTX 2080 Ti rather than the older GTX 980 Ti's peers like the Radeon Pro W5500X.

Anyone building a system for modern DirectX 12 or Vulkan gaming should choose the RX 6800 without hesitation. Its 141.3% lead in 3DMark Steel Nomad and 141.2% lead in Geekbench Vulkan demonstrate that it is more than twice as fast in the APIs that define contemporary game development. The ray tracing hardware, 16 GB memory, and DX12 Ultimate support also future-proof it against upcoming titles that will leverage these features.

The GTX 980 Ti's only remaining justification is the Geekbench OpenCL result, where its 43.7% lead could matter for specialized compute tasks locked to that API. But even there, the margin is modest compared to the RX 6800's dominance elsewhere, and the GTX 980 Ti's older 28 nm process, 6 GB memory, and lack of RT cores make it a poor choice for any forward-looking workload. The verdict is straightforward: the RX 6800 is the modern performer, and the GTX 980 Ti is a legacy card whose best days are behind it.