NVIDIA GeForce GTX 980 Ti vs NVIDIA GeForce RTX 2080 SUPER Comparison

Head-to-Head Benchmarks

The recorded data shows a clear sweep for the RTX 2080 SUPER across the three shared benchmark tests. In 3DMark Steel Nomad DX12, the RTX 2080 SUPER scores 1882 against 1321 for the GTX 980 Ti, a 29.8% advantage. That is a substantial margin in a modern DirectX 12 workload, suggesting the newer architecture extracts far more performance from the same class of graphics workload.

The gap widens dramatically in compute-oriented tests. In Geekbench OpenCL, the RTX 2080 SUPER delivers 99226 points versus 43513 for the GTX 980 Ti, a 56.1% difference. The Geekbench Vulkan result follows the same pattern: 111284 for the RTX 2080 SUPER against 47724 for the GTX 980 Ti, a 57.1% gap. These are not marginal improvements; the RTX 2080 SUPER more than doubles the GTX 980 Ti's output in both OpenCL and Vulkan compute scenarios.

What makes these numbers striking is that the GTX 980 Ti still holds a respectable 73rd percentile position among all GPUs in the database, while the RTX 2080 SUPER sits at the 69th percentile. The percentile ranking is based on average benchmark score across all recorded tests, not just the head-to-head set. The GTX 980 Ti's average score is 28020, while the RTX 2080 SUPER averages 24170. So despite losing every shared test by a wide margin, the GTX 980 Ti actually has a higher overall average score in the database. This suggests the GTX 980 Ti benefits from a broader set of test results, or that the specific tests where the RTX 2080 SUPER excels are not fully represented in its average.

The RTX 2080 SUPER's nearest rivals in the database include the GTX 780 Ti (24236, 0.3% higher), the GTX 1630 (24277, 0.4% higher), the RX 6800S (24063, 0.4% lower), and the RX 6600 XT (24442, 1.1% higher). The GTX 980 Ti, by contrast, sits closest to the Radeon Pro W5500X (27973, 0.2% lower), the FirePro S7150 (28117, 0.3% higher), the RX 7800M (27883, 0.5% lower), and the Radeon Pro Vega 20 (27839, 0.7% lower). The GTX 980 Ti's average sits in a cluster of professional and workstation cards, while the RTX 2080 SUPER's average is surrounded by older and lower-tier gaming cards. That is an odd inversion given the RTX 2080 SUPER's dominance in the shared tests.

Architecture Differences

The two cards come from fundamentally different architectural eras. The GTX 980 Ti uses the GM200 chip built on Maxwell 2.0 architecture, fabricated on a 28 nm process at TSMC. The RTX 2080 SUPER uses the TU104 chip built on Turing architecture, fabricated on a 12 nm process, also at TSMC. The process shrink from 28 nm to 12 nm is the single largest physical difference, allowing the RTX 2080 SUPER to pack far more transistors into a smaller die.

Transistor counts tell the story clearly. The GTX 980 Ti has 8,000 million transistors on a 601 mm² die, giving a transistor density of 13.3 million per square millimeter. The RTX 2080 SUPER has 13,600 million transistors on a 545 mm² die, a density of 25.0 million per square millimeter. The RTX 2080 SUPER is physically smaller but contains 70% more transistors, and its density is nearly double. That density increase is the direct result of the 12 nm process.

The memory subsystems are different in kind, not just degree. The GTX 980 Ti uses 6 GB of GDDR5 memory on a 384-bit bus, delivering 336.6 GB/s of bandwidth. The RTX 2080 SUPER uses 8 GB of GDDR6 memory on a 256-bit bus, delivering 495.9 GB/s. The RTX 2080 SUPER achieves 47% more bandwidth despite a narrower bus, proof of the higher data rates of GDDR6. Effective memory speed is 7 Gbps on the GTX 980 Ti versus 15.5 Gbps on the RTX 2080 SUPER.

The RTX 2080 SUPER also introduces hardware features that simply did not exist in the Maxwell generation. It has 48 RT cores for ray tracing and 384 tensor cores for AI workloads. The GTX 980 Ti has neither. This is not a small addition; it is a fundamental change in what the GPU can do. The GTX 980 Ti is limited to traditional rasterization, while the RTX 2080 SUPER can accelerate ray-traced rendering and tensor-based operations.

Compute throughput reflects the architectural leap. The GTX 980 Ti's FP32 throughput is 6.060 TFLOPS, while the RTX 2080 SUPER reaches 11.15 TFLOPS in FP32 and 22.30 TFLOPS in FP16 with a 2:1 ratio. The GTX 980 Ti has no listed FP16 figure. The RTX 2080 SUPER's texture rate is 348.5 GTexel/s against 189.4 GTexel/s for the GTX 980 Ti, and its pixel rate is 116.2 GPixel/s against 103.3 GPixel/s. The RTX 980 Ti actually has more ROPs, 96 against 64, yet still trails in pixel rate due to its lower clocks.

Where Each One Wins

The RTX 980 Ti's pixel rate advantage is worth noting, but the data does not favor it anywhere in the shared tests. Every head-to-head metric goes to the RTX 2080 SUPER. The GTX 980 Ti's only wins are in non-shared hardware attributes like ROP count and pixel fill rate, and even those are marginal in practical impact given its lower shading unit count and texture rate.

The RTX 2080 SUPER's wins fall into two clear categories. First, it wins in modern API performance, as shown by its 29.8% lead in 3DMark Steel Nomad DX12. Second, it wins decisively in compute throughput, as shown by its 56.1% lead in OpenCL and 57.1% lead in Vulkan. These are the workloads that stress the GPU's raw compute capabilities, and the Turing architecture's higher FP32 throughput and tensor cores clearly pay off.

The GTX 980 Ti does not win any category in the head-to-head set. Its higher average benchmark score in the database (28020 versus 24170) comes from tests not shared between the two cards, and its percentile rank is actually higher (73rd versus 69th). This creates a confusing picture for anyone trying to compare the two directly. The data suggests the GTX 980 Ti has been tested more extensively in the database, and its average reflects a wider range of workloads, many of which may not stress the features that make the RTX 2080 SUPER faster.

For practical use, the RTX 2080 SUPER is the clear choice for any workload that leverages modern APIs, compute shaders, ray tracing, or tensor operations. The GTX 980 Ti remains competitive only in scenarios that were already well-served by Maxwell-era hardware, and even then, its lower FP32 throughput and older memory technology place it well behind.

Specification Differences

The two cards differ in nearly every specification field. The GTX 980 Ti uses the GM200 chip on Maxwell 2.0 architecture, while the RTX 2080 SUPER uses the TU104 chip on Turing architecture. The process node is 28 nm versus 12 nm. Transistor count is 8,000 million versus 13,600 million. Die size is 601 mm² versus 545 mm². Transistor density is 13.3 million per square millimeter versus 25.0 million.

Clock speeds differ substantially. The GTX 980 Ti has a base clock of 1000 MHz and a boost clock of 1076 MHz. The RTX 2080 SUPER has a base clock of 1650 MHz and a boost clock of 1815 MHz. Memory clock is 1753 MHz (7 Gbps effective) on the GTX 980 Ti versus 1937 MHz (15.5 Gbps effective) on the RTX 2080 SUPER.

Memory configuration differs in size, type, bus width, and bandwidth. The GTX 980 Ti has 6 GB of GDDR5 on a 384-bit bus with 336.6 GB/s bandwidth. The RTX 2080 SUPER has 8 GB of GDDR6 on a 256-bit bus with 495.9 GB/s bandwidth. The GTX 980 Ti has 2816 shading units, 176 TMUs, and 96 ROPs. The RTX 2080 SUPER has 3072 shading units, 192 TMUs, and 64 ROPs. The RTX 2080 SUPER adds 48 RT cores and 384 tensor cores; the GTX 980 Ti has none.

Pixel rate is 103.3 GPixel/s for the GTX 980 Ti versus 116.2 GPixel/s for the RTX 2080 SUPER. Texture rate is 189.4 GTexel/s versus 348.5 GTexel/s. FP32 throughput is 6.060 TFLOPS versus 11.15 TFLOPS. The RTX 2080 SUPER also lists FP16 at 22.30 TFLOPS with a 2:1 ratio; the GTX 980 Ti has no FP16 figure.

Both cards have a TDP of 250 W and use the same power connectors: 1x 6-pin plus 1x 8-pin. Both suggest a 600 W power supply. Both are dual-slot cards with PCIe 3.0 x16 interfaces. Both are 267 mm long, but the GTX 980 Ti is 111 mm tall and 40 mm wide, while the RTX 2080 SUPER is 116 mm tall and 35 mm wide.

Display outputs differ. The GTX 980 Ti has 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2. The RTX 2080 SUPER has 1x HDMI 2.0, 3x DisplayPort 1.4a, and 1x USB Type-C. API support differs in DirectX version too: the GTX 980 Ti supports DirectX 12 (12_1 and 12_2 featured in the RTX 2080 SUPER's case. The RTX 2080 SUPER supports DirectX 12 Ultimate (12_2), while the GTX 980 Ti supports DirectX 12 (12_1. OpenGL 4.6 and Vulkan 1.4 are identical on both cards.

The GTX 980 Ti was released on June 1, 2015, and its launch MSRP was 649 USD. The RTX 2080 SUPER was released on July 22, 2019, with a launch MSRP of 699 USD.

FAQ

Q: Which card wins in raw compute performance?

The RTX 2080 SUPER leads by 56.1% in Geekbench OpenCL and 57.1% in Geekbench Vulkan. It also has nearly double the FP32 throughput: 11.15 TFLOPS versus 6.060 TFLOPS.

Q: How do the memory systems compare?

The GTX 980 Ti has 6 GB of GDDR5 on a 384-bit bus with 336.6 GB/s bandwidth. The RTX 2080 SUPER has 8 GB of GDDR6 on a 256-bit bus with 495.9 GB/s. The RTX 2080 SUPER delivers more bandwidth despite a narrower bus.

Q: Does the GTX 980 Ti have ray tracing support?

No. The GTX 980 Ti has no RT cores and no tensor cores. The RTX 2080 SUPER has 48 RT cores and 384 tensor cores.

Q: Why does the GTX 980 Ti have a higher average benchmark score if it loses every head-to-head test?

The GTX 980 Ti has an average score of 28020 across all recorded tests, while the RTX 2080 SUPER averages 24170. The RTX 2080 SUPER wins the three shared tests, but the GTX 980 Ti's average is pulled up by other tests in the database. Its percentile rank is 73rd versus 69th for the RTX 2080 SUPER.

Q: What is the power draw of each card?

Both cards have a TDP of 250 W and use a 1x 6-pin plus 1x 8-pin power connector setup. Both suggest a 600 W power supply.

Q: Which card has better DirectX support?

The RTX 2080 SUPER supports DirectX 12 Ultimate (12_2), while the GTX 980 Ti supports DirectX 12 (12_1. OpenGL 4.6 and Vulkan 1.4 are supported by both.

The Verdict

The RTX 2080 SUPER is the faster card in every shared benchmark, with margins ranging from 29.8% in 3DMark Steel Nomad DX12 to 57.1% in Geekbench Vulkan. Its architectural advantages are substantial: a 12 nm process, 13,600 million transistors, GDDR6 memory, 48 RT cores, 384 tensor cores, and 11.15 TFLOPS of FP32 throughput. Anyone needing modern API performance, ray tracing, or high compute throughput should choose the RTX 2080 SUPER without hesitation.

The GTX 980 Ti is not without merit. Its average benchmark score in the database is higher at 28020 versus 24170, and it holds a better percentile rank at 73rd versus 69th. It has more ROPs (96 versus 64) and a wider memory bus (384-bit versus 256-bit), though its memory bandwidth is lower. For rasterization workloads that do not leverage modern features, the GTX 980 Ti remains a capable card, but the recorded data shows it lagging significantly in every test that matters for current-generation software.

The choice is clear: the RTX 2080 SUPER is the superior GPU by a wide margin in all shared tests. The GTX 980 Ti's higher average score is an artifact of the database's test mix, not a sign of real-world superiority. The RTX 2080 SUPER's launch MSRP was 699 USD, compared to 649 USD for the GTX 980 Ti, a small premium for a massive performance increase.