NVIDIA GeForce GTX 760 vs NVIDIA GeForce GTX 980 Comparison
NVIDIA GeForce GTX 760
GeForce GTX 980
PERFORMANCE BENCHMARKS
Analysis: NVIDIA GeForce GTX 760 vs NVIDIA GeForce GTX 980
Head-to-Head Benchmarks
The recorded head-to-head data contains three benchmark tests, and the NVIDIA GeForce GTX 980 wins all of them. The largest margin appears in Geekbench Metal, where the GTX 980 scores 15,163 against the GTX 760’s 4,524, a delta of 70.2% in favor of the newer card. That is a decisive gap, more than triple the older card’s output. In Geekbench OpenCL, the GTX 980 posts 34,676 versus 11,299, a 67.4% advantage. The smallest relative gap is in Geekbench Vulkan, where the GTX 980 scores 22,543 against 12,551, still a 44.3% lead. The direction of the results is consistent across all three APIs: the GTX 980 is substantially faster in every measured workload.
The delta percentages reflect the GTX 760’s score as a fraction of the GTX 980’s score. For Metal, the 70.2% delta means the GTX 760 trails by that amount, not that it achieves 70% of the performance. The OpenCL delta of 67.4% similarly indicates a large shortfall. The Vulkan delta of 44.3% is the closest contest, but that still leaves the GTX 980 ahead by a wide margin. No benchmark in the database shows the GTX 760 winning, and the wins tally stands at 3 for the GTX 980 and 0 for the GTX 760.
The average benchmark scores further contextualize these results. The GTX 760 has an average score of 9,458, while the GTX 980’s average is 8,167. That might seem contradictory given the head-to-head results, but the averages include different test sets. The GTX 980’s average is pulled down by several lower-scoring tests, such as Passmark DirectX 12 at 46, Passmark DirectX 10 at 53, and Passmark DirectX 11 at 83. The GTX 760’s average is computed from only three Geekbench tests, all of which are moderate. So the head-to-head comparisons are the more direct measure of relative performance, and they clearly favor the GTX 980.
Percentile ranks also differ. The GTX 760 sits at the 46th percentile among all GPUs, while the GTX 980 is at the 43rd percentile. Despite the GTX 980’s superior head-to-head performance, its broader benchmark profile places it slightly lower in the overall distribution. This is because the GTX 980’s Passmark scores, particularly the DirectX 12 result of 46 and DirectX 10 result of 53, are very low relative to its Geekbench numbers. The GTX 760, lacking those low scores, appears more consistent in its limited test set.
Where Each One Wins
Looking at the benchmark breakdown, the GTX 980 wins in every category where direct comparisons exist. Its Geekbench Metal score of 15,163 is 70.2% higher than the GTX 760’s 4,524. For users running Metal-based applications, the GTX 980 offers a dramatic performance uplift. In OpenCL workloads, the GTX 980’s 34,676 score dwarfs the GTX 760’s 11,299, a 67.4% lead. OpenCL is common in compute tasks, video encoding, and some scientific applications, so the GTX 980 is clearly the stronger choice there.
For Vulkan, the GTX 980 scores 22,543 versus 12,551, a 44.3% advantage. Vulkan-based games and compute workloads will run noticeably faster on the GTX 980, though the gap is smaller than in Metal or OpenCL. The GTX 760 has no winning benchmark in the recorded data. There is no scenario in the database where the GTX 760 outperforms the GTX 980.
Beyond the direct comparisons, the GTX 980’s additional benchmark scores show where it excels in isolation. Its Passmark G3D score is 11,095, and its Passmark GPU Compute score is 4,753. The GTX 760 does not have those tests recorded, so no direct comparison is possible, but the GTX 980’s numbers indicate solid performance in older DirectX workloads as well. The GTX 980 also has a Passmark G2D score of 792, which reflects 2D rendering and desktop compositing, a category where the GTX 760 has no recorded result.
In summary, the GTX 980 wins all three head-to-head tests, and its broader benchmark suite shows strength across Metal, OpenCL, Vulkan, and legacy DirectX APIs. The GTX 760 is left without a single recorded victory.
Architecture Differences
The two cards come from different NVIDIA architectures and generations. The GTX 760 uses the GK104 chip, built on the Kepler architecture, and belongs to the GeForce 700 generation. The GTX 980 uses the GM204 chip, built on the Maxwell 2.0 architecture, and belongs to the GeForce 900 generation. Both are manufactured by TSMC on a 28 nm process, but the transistor counts differ substantially. The GTX 760 has 3,540 million transistors on a 294 mm² die, giving a transistor density of 12.0M per mm². The GTX 980 has 5,200 million transistors on a 398 mm² die, with a density of 13.1M per mm². The GTX 980 packs 47% more transistors into a 35% larger die, which partly explains its higher performance.
Core counts scale accordingly. The GTX 760 has 1,152 shading units, 96 texture mapping units, and 32 render output units. The GTX 980 has 2,048 shading units, 128 TMUs, and 64 ROPs. That is 896 more shading units, 32 more TMUs, and double the ROPs. The ROP count doubling is particularly relevant for fill-rate-bound scenarios, such as high-resolution rendering with heavy anti-aliasing. Pixel rate and texture rate reflect these differences. The GTX 760 achieves 24.77 GPixel/s and 99.07 GTexel/s, while the GTX 980 reaches 77.82 GPixel/s and 155.6 GTexel/s. The GTX 980’s pixel rate is more than triple that of the GTX 760, and its texture rate is 57% higher.
Clock speeds also favor the GTX 980. The GTX 760 has a base clock of 980 MHz and a boost clock of 1,032 MHz. The GTX 980 runs at a base of 1,127 MHz and boosts to 1,216 MHz. That is a 147 MHz higher base and 184 MHz higher boost. Combined with the larger core, the GTX 980’s FP32 throughput is 4.981 TFLOPS, more than double the GTX 760’s 2.378 TFLOPS. The theoretical compute advantage is clear.
Memory configurations differ as well. The GTX 760 has 2 GB of GDDR5 on a 256-bit bus, with memory clocked at 1,502 MHz and 6 Gbps effective, yielding 192.3 GB/s of bandwidth. The GTX 980 has 4 GB of GDDR5 on the same 256-bit bus, but memory runs at 1,753 MHz and 7 Gbps effective, raising bandwidth to 224.4 GB/s. The GTX 980 offers double the capacity and 16.7% more bandwidth. For modern games with large textures, the 4 GB frame buffer is a meaningful advantage over 2 GB.
API support also differs. The GTX 760 supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The GTX 980 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The higher DirectX feature level and newer Vulkan version give the GTX 980 better compatibility with recent titles and features. Display outputs also vary: the GTX 760 has 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2, while the GTX 980 has 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2. The GTX 980’s HDMI 2.0 supports higher resolutions and refresh rates over a single cable.
Power consumption is similar despite the performance gap. The GTX 760 has a TDP of 170 W, and the GTX 980 has a TDP of 165 W. Both use dual-slot coolers, 2x 6-pin power connectors, and a suggested PSU of 450 W. The GTX 980 delivers roughly double the FP32 throughput while drawing 5 W less, a strong efficiency improvement. Physical dimensions differ: the GTX 760 is 241 mm long, while the GTX 980 is 267 mm long, 111 mm tall, and 40 mm wide. The GTX 980 is longer, so case clearance should be considered.
The Verdict
The data points to one clear conclusion: the NVIDIA GeForce GTX 980 is the superior card in every recorded head-to-head test. It wins Geekbench Metal by 70.2%, OpenCL by 67.4%, and Vulkan by 44.3%. Its architecture, with more shading units, more ROPs, higher clocks, and double the memory, supports those results. The GTX 980 also offers higher API support, including DirectX 12_1 and Vulkan 1.4, which makes it more future-proof for modern software.
However, the overall percentile ranks complicate a simple recommendation. The GTX 760 sits at the 46th percentile among all GPUs, while the GTX 980 is at the 43rd percentile. That means the GTX 980’s average benchmark score of 8,167 is lower than the GTX 760’s 9,458. The reason is the GTX 980’s poor showing in certain Passmark tests, particularly DirectX 12 (46) and DirectX 10 (53). Those low scores drag its average down, but they do not reflect the card’s performance in the head-to-head tests where it dominates.
For users who prioritize the specific workloads measured in the head-to-head set, the GTX 980 is the obvious pick. Metal, OpenCL, and Vulkan all show massive gains. For users who rely on legacy DirectX 10 or 12 workloads, the GTX 980’s Passmark scores are concerning, but the GTX 760 has no recorded Passmark scores to compare against. The data available does not allow a direct comparison in those legacy tests.
The GTX 760 is an end-of-life product from 2013, while the GTX 980 is also end-of-life but from a later generation. The GTX 980’s launch MSRP was 549 USD, compared to the GTX 760’s 249 USD. That price difference is substantial, but the performance difference in the recorded benchmarks is even larger. The GTX 980 costs more than double, yet it delivers more than double the performance in Metal and OpenCL. The GTX 980 is the stronger choice for anyone running the tested workloads.
FAQ
Q: Which card wins in Geekbench Metal?
A: The NVIDIA GeForce GTX 980 wins with a score of 15,163, which is 70.2% higher than the GTX 760’s 4,524.
Q: What is the memory capacity difference?
A: The GTX 760 has 2 GB of GDDR5, while the GTX 980 has 4 GB of GDDR5, both on a 256-bit bus.
Q: How do the pixel rates compare?
A: The GTX 980 achieves 77.82 GPixel/s, while the GTX 760 achieves 24.77 GPixel/s, a difference of more than three times.
Q: Which card has more shading units?
A: The GTX 980 has 2,048 shading units, while the GTX 760 has 1,152, a difference of 896 units.
Q: What is the power draw of each card?
A: The GTX 760 has a TDP of 170 W, and the GTX 980 has a TDP of 165 W, so the GTX 980 draws slightly less power despite higher performance.
Q: How do the average benchmark scores compare?
A: The GTX 760 has an average score of 9,458, while the GTX 980 has an average of 8,167, but the GTX 980 wins all three head-to-head tests.