NVIDIA GeForce GTX 960 vs NVIDIA GeForce GTX 980 Comparison

NVIDIA GeForce GTX 960 and NVIDIA GeForce GTX 980 are both Maxwell 2.0 parts from the GeForce 900 generation. The data shows a clear performance hierarchy, but the GTX 960 holds a notable position in the database due to its average benchmark score. The GTX 980, despite having a lower percentile ranking, delivers substantially higher raw performance in every recorded test. This analysis breaks down where each card wins, the architectural differences that explain the gap, and the benchmark results that quantify the separation.

Where Each One Wins

The benchmark records show a complete sweep in favor of the NVIDIA GeForce GTX 980. Out of the four head-to-head tests available in the database, the GTX 980 wins all four, leaving the GTX 960 with zero wins. This is not a close contest; the GTX 980 dominates across different API workloads, including DirectX 12, Metal, OpenCL, and Vulkan.

The GTX 960, however, is not without its own standing. In the database, the GTX 960 posts an average benchmark score of 9273, which places it in the 45th percentile of all GPUs. Its nearest rival, the NVIDIA GeForce GTX 465, scores 9294, a delta of only -0.2%, meaning the GTX 960 is essentially tied with that older card. The GTX 960 also sits close to the AMD Radeon R7 M380 (9313, -0.4%) and the NVIDIA GeForce GTX 850M (9302, -0.3%). Interestingly, the AMD Radeon Vega 8 scores 9221, which is 0.6% lower than the GTX 960, so the GTX 960 actually leads that integrated part.

The GTX 980, despite its stronger raw scores, sits at the 43rd percentile with an average benchmark score of 8167. This is lower than the GTX 960's average, which is a quirk of the database's weighting across all recorded tests. The GTX 980's nearest rivals include the NVIDIA GeForce GTX 950M (8135, 0.4% higher), the AMD Radeon R9 M360 (8129, 0.5% higher), and the NVIDIA GeForce 945M (8099, 0.8% higher). The GTX 980 leads the NVIDIA GRID K2 (8080) by 1.1%. So, while the GTX 980 wins every direct comparison against the GTX 960, its overall database percentile is dragged down by other factors, likely the inclusion of more diverse test suites.

In terms of use-case split, the GTX 980 is the clear choice for any workload that stresses raw graphics throughput. The GTX 960 is competitive only in the context of its own tier, where it trades blows with older or lower-end parts. The data does not support any scenario where the GTX 960 outperforms the GTX 980.

Architecture Differences

Both cards are built on the Maxwell 2.0 architecture and use the 28 nm process at TSMC, but they are fundamentally different chips. The GTX 960 uses the GM206 chip, while the GTX 980 uses the GM204 chip. This is the primary source of the performance gap.

The transistor counts differ substantially. The GTX 960 packs 2,940 million transistors on a die size of 228 mm², resulting in a transistor density of 12.9 million per mm². The GTX 980, in contrast, houses 5,200 million transistors on a 398 mm² die, with a density of 13.1 million per mm². The GTX 980 has nearly double the transistor budget, which allows for a much larger execution engine.

Memory configurations also diverge. The GTX 960 comes with 2 GB of GDDR5 on a 128-bit bus, yielding a bandwidth of 112.2 GB/s. The GTX 980 offers 4 GB of GDDR5 on a 256-bit bus, doubling the bandwidth to 224.4 GB/s. Both cards run their memory at the same effective speed of 7 Gbps, but the wider bus on the GTX 980 is the decisive factor.

The compute resources scale accordingly. The GTX 960 has 1024 shading units, 64 texture mapping units (TMUs), and 32 render output units (ROPs). The GTX 980 doubles these: 2048 shading units, 128 TMUs, and 64 ROPs. This doubling is reflected in the fill rates and compute throughput. The GTX 960 achieves a pixel rate of 37.70 GPixel/s and a texture rate of 75.39 GTexel/s, with FP32 performance of 2.413 TFLOPS. The GTX 980 reaches 77.82 GPixel/s, 155.6 GTexel/s, and 4.981 TFLOPS, all of which are more than double the GTX 960's figures.

Physical and power characteristics also differ. The GTX 960 has a TDP of 120 W and requires a single 6-pin power connector, with a suggested PSU of 300 W. The GTX 980 draws 165 W, needs two 6-pin connectors, and suggests a 450 W PSU. Both are dual-slot cards, but the GTX 980 is physically larger: 267 mm long, 111 mm high, and 40 mm wide, compared to the GTX 960's 241 mm length. The display outputs are identical (1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.2), and both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4.

The release dates show the GTX 980 came first, on September 18, 2014, while the GTX 960 followed on January 21, 2015. Both are end-of-life products, with the same predecessor (GeForce 700) and successor (GeForce 10) lines.

Head-to-Head Benchmarks

The head-to-head results in the database are unambiguous. In the 3DMark Steel Nomad DX12 test, the GTX 960 scores 162, while the GTX 980 scores 474. This gives the GTX 980 a win with a delta of -65.8% for the GTX 960, meaning the GTX 980 is roughly 2.9 times faster in this specific workload.

The Geekbench Metal test shows a similar pattern. The GTX 960 records 8773 points, and the GTX 980 records 15163 points. The delta is -42.1%, indicating the GTX 980 is about 73% faster. In Geekbench OpenCL, the GTX 960 scores 18925, while the GTX 980 scores 34676, a delta of -45.4%, which translates to the GTX 980 being roughly 83% faster.

The Geekbench Vulkan test produces the largest relative gap. The GTX 960 scores 9231, and the GTX 980 scores 22543. The delta is -59.1%, meaning the GTX 980 is approximately 144% faster, or nearly 2.4 times the GTX 960's score.

These results highlight that the GTX 980's advantage is not uniform but is consistently large. The smallest gap is in Metal, where the GTX 980 leads by 42.1%, and the largest is in Steel Nomad DX12, where it leads by 65.8%. The Vulkan and OpenCL gaps fall in between at 59.1% and 45.4%, respectively. In every case, the GTX 980's doubling of shading units, TMUs, and ROPs, along with double the memory bandwidth, translates into a commanding lead.

The GTX 960's best relative showing is in Geekbench Metal, where it still loses by a wide margin. There are no tests where the GTX 960 comes close to matching the GTX 980. The database also lists additional PassMark tests for the GTX 980 alone, including DirectX 9 (164), DirectX 10 (53), DirectX 11 (83), DirectX 12 (46), G2D (792), G3D (11095), and GPU Compute (4753), but no corresponding GTX 960 scores are available for those tests, so they cannot be compared directly.

FAQ

Q: Which card has a higher average benchmark score in the database?

A: The GTX 960 has a higher average benchmark score of 9273, compared to the GTX 980's 8167. However, the GTX 980 wins every head-to-head test recorded.

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

A: The largest gap is in the 3DMark Steel Nomad DX12 test, where the GTX 980 scores 474 versus the GTX 960's 162, a delta of -65.8% for the GTX 960.

Q: Do both cards support the same APIs?

A: Yes, both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. Their display outputs are also identical: 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2.

Q: How do the memory bandwidths compare?

A: The GTX 960 has a 128-bit bus with 112.2 GB/s bandwidth, while the GTX 980 has a 256-bit bus with 224.4 GB/s bandwidth. Both use GDDR5 at 7 Gbps effective speed.

Q: Which card has a lower TDP?

A: The GTX 960 has a TDP of 120 W and requires a single 6-pin power connector. The GTX 980 has a TDP of 165 W and requires two 6-pin connectors.

Q: What are the nearest rivals for each card in the database?

A: The GTX 960's nearest rival is the NVIDIA GeForce GTX 465 (9294, -0.2% delta). The GTX 980's nearest rival is the NVIDIA GeForce GTX 950M (8135, 0.4% delta).

Specification Differences

| Specification | NVIDIA GeForce GTX 960 | NVIDIA GeForce GTX 980 |

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

| Chip | GM206 | GM204 |

| Transistors | 2,940 million | 5,200 million |

| Die Size | 228 mm² | 398 mm² |

| Transistor Density | 12.9M / mm² | 13.1M / mm² |

| Base Clock | 1127 MHz | 1127 MHz |

| Boost Clock | 1178 MHz | 1216 MHz |

| Memory Size | 2 GB | 4 GB |

| Memory Bus Width | 128 bit | 256 bit |

| Memory Bandwidth | 112.2 GB/s | 224.4 GB/s |

| Shading Units | 1024 | 2048 |

| TMUs | 64 | 128 |

| ROPs | 32 | 64 |

| Pixel Rate | 37.70 GPixel/s | 77.82 GPixel/s |

| Texture Rate | 75.39 GTexel/s | 155.6 GTexel/s |

| FP32 | 2.413 TFLOPS | 4.981 TFLOPS |

| TDP | 120 W | 165 W |

| Power Connectors | 1x 6-pin | 2x 6-pin |

| Suggested PSU | 300 W | 450 W |

| Dimensions | 241 mm length | 267 mm length, 111 mm height, 40 mm width |

| Release Date | 2015-01-21 | 2014-09-18 |

| Launch MSRP | 199 USD | 549 USD |

The base clocks are identical at 1127 MHz, but the GTX 980 boosts higher at 1216 MHz versus 1178 MHz. The memory clock is the same at 1753 MHz (7 Gbps effective). Both use the same process node (28 nm), foundry (TSMC), architecture (Maxwell 2.0), generation (GeForce 900), bus interface (PCIe 3.0 x16), and slot width (Dual-slot). Both support the same APIs and display outputs.

The Verdict

The data is decisive: the NVIDIA GeForce GTX 980 is the superior card in every recorded benchmark. It wins all four head-to-head tests, with deltas ranging from -42.1% in Geekbench Metal to -65.8% in 3DMark Steel Nomad DX12. The GTX 980 doubles the GTX 960's shading units, TMUs, ROPs, and memory bandwidth, which explains the consistent performance lead. Its higher boost clock (1216 MHz vs 1178 MHz) and larger die (398 mm² vs 228 mm²) further cement its position.

For buyers or system builders choosing between these two, the GTX 980 is the obvious pick for any workload that demands maximum frame rates or compute throughput. The GTX 960, however, should not be dismissed outright. Its average benchmark score of 9273 is higher than the GTX 980's 8167, and it sits at a higher percentile (45th vs 43rd). This suggests that in the broader database, the GTX 960's performance profile is more consistent relative to its peers, possibly due to its lower TDP (120 W vs 165 W) and simpler power requirements (single 6-pin vs dual 6-pin).

The GTX 960's nearest rivals are all within a 0.6% delta, indicating it is a well-balanced card in its tier. The GTX 980's nearest rivals are also close, but they are lower-performing parts like the GTX 950M and R9 M360, which reflects the fact that the GTX 980's average score is pulled down by tests that may not favor its architecture. In direct competition, the GTX 980 is 73% to 144% faster than the GTX 960 depending on the test. There is no benchmark in the database where the GTX 960 wins. The verdict is straightforward: choose the GTX 980 for performance, or the GTX 960 for a lower-power, smaller-footprint alternative that still holds its own in its own class.