Intel UHD Graphics 750 vs NVIDIA GeForce GTX 980 Comparison

# FAQ

Q: How do the two GPUs compare on raw compute performance?

A: In the Geekbench OpenCL test, the NVIDIA GeForce GTX 980 scores 34,676 versus 6,743 for the Intel UHD Graphics 750. That is a delta of 414.3% in favor of the GTX 980, meaning the NVIDIA part delivers over five times the OpenCL throughput of the integrated Intel solution.

Q: Which GPU has better Vulkan performance, and by how much?

A: The GeForce GTX 980 wins the Geekbench Vulkan test with a score of 22,543 against 8,138 for the Intel UHD Graphics 750. This represents a 177% advantage for the NVIDIA card, roughly tripling the Intel iGPU’s Vulkan result.

Q: What do the percentile rankings indicate about each GPU’s standing?

A: The GTX 980 sits at the 43rd percentile of all GPUs, while the UHD 750 ranks at the 40th percentile. Despite the massive performance gap between them in head-to-head tests, both parts land in similar percentiles because the GTX 980 is compared against discrete desktop graphics cards, whereas the UHD 750 is measured against a broader field that includes other integrated solutions.

Q: Are there any benchmark categories where the Intel UHD Graphics 750 wins?

A: No. Across the two head-to-head benchmark tests available (Geekbench OpenCL and Geekbench Vulkan), the GTX 980 wins both. The win count stands at 2 for NVIDIA and 0 for Intel.

Q: What are the closest rivals for each GPU?

A: The GTX 980’s nearest rivals include the NVIDIA GeForce GTX 950M (avg score 8,135, 0.4% faster), AMD Radeon R9 M360 (8,129, 0.5% faster), and NVIDIA GeForce 945M (8,099, 0.8% faster). The UHD 750’s nearest rivals include the AMD Radeon HD 8850M (7,447, 0.1% faster), NVIDIA GeForce GTX 1650 (7,472, 0.4% faster), and Intel Arc A310 (7,550, 1.4% faster).

Q: How does the average benchmark score compare between the two?

A: The GTX 980 has an average benchmark score of 8,167 across all its tested workloads, while the UHD 750 averages 7,441. This puts the NVIDIA card about 9.8% higher on average, though the gap is much larger in the specific compute and graphics tests where both were measured.

# Architecture Differences

The architectural gap between these two GPUs is fundamental. The NVIDIA GeForce GTX 980 is built on the Maxwell 2.0 architecture with the GM204 chip, fabricated on a 28 nm process at TSMC. It packs 5,200 million transistors into a 398 mm² die, yielding a transistor density of 13.1 million transistors per square millimeter. The Intel UHD Graphics 750, by contrast, uses the Generation 12.1 architecture with the Rocket Lake chip, built on Intel’s 14 nm+++ process. It is an integrated graphics processor (IGP) with no standalone die size or transistor count listed, as it shares the CPU package.

The core configurations diverge sharply. The GTX 980 features 2,048 shading units, 128 texture mapping units (TMUs), and 64 raster output units (ROPs). The UHD 750 has 256 shading units, 16 TMUs, and 8 ROPs. This 8:1 ratio in shading units and TMUs, and an 8:1 ratio in ROPs, explains the massive throughput differences seen in pixel and texture rates. The GTX 980 delivers 77.82 GPixel/s pixel fill rate and 155.6 GTexel/s texture fill rate, while the UHD 750 manages 10.40 GPixel/s and 20.80 GTexel/s, respectively.

Memory architecture is another major divider. The GTX 980 uses 4 GB of GDDR5 memory on a 256-bit bus, with 224.4 GB/s of bandwidth and a memory clock of 1753 MHz (7 Gbps effective). The UHD 750 relies on System Shared memory, with its type, bus width, and bandwidth all listed as "System Shared" or "System Dependent." This means the Intel iGPU has no dedicated video memory and must borrow from system RAM, which inherently limits bandwidth compared to a dedicated GDDR5 solution.

Feature sets for modern APIs are similar on paper: both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. However, the physical implementation differs. The GTX 980 is a dual-slot card with 2x 6-pin power connectors, a 165 W TDP, and a recommended 450 W PSU. The UHD 750 is an IGP with a 15 W TDP, no power connectors, and no suggested PSU, as it draws power through the motherboard. The GTX 980 uses a PCIe 3.0 x16 interface, while the UHD 750 connects via Ring Bus.

# Where Each One Wins

The data paints a clear picture: the GTX 980 wins decisively across every shared benchmark. In the Geekbench OpenCL test, the NVIDIA card scores 34,676 versus 6,743 for Intel, a 414.3% advantage. In Geekbench Vulkan, the GTX 980 scores 22,543 versus 8,138, a 177% lead. These are not marginal victories; they represent different performance tiers entirely.

For gaming and compute workloads that leverage OpenCL or Vulkan, the GTX 980 is the only viable option between the two. Its dedicated 4 GB GDDR5 frame buffer and 224.4 GB/s bandwidth allow it to handle textures and geometry that the UHD 750’s system-shared memory cannot realistically accommodate. The GTX 980’s 4.981 TFLOPS of FP32 throughput dwarfs the UHD 750’s 665.6 GFLOPS — a 7.5x gap in raw floating-point compute.

The UHD 750, however, has its niche in low-power, integrated scenarios. Its 15 W TDP makes it suitable for compact systems where a discrete GPU is impractical. It supports the same modern API feature set (DirectX 12_1, OpenGL 4.6, Vulkan 1.4), so it can run modern titles at reduced settings, but the benchmark data shows no scenario where it outperforms the GTX 980. The Intel part’s FP16 capability of 1,331.2 GFLOPS (2:1) is notable for a chip of its size, but it does not translate into a win in any measured test.

# Specification Differences

| Specification | NVIDIA GeForce GTX 980 | Intel UHD Graphics 750 |

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

| Architecture | Maxwell 2.0 | Generation 12.1 |

| Chip | GM204 | Rocket Lake |

| Process Node | 28 nm | 14 nm+++ |

| Foundry | TSMC | Intel |

| Transistors | 5,200 million | Not listed |

| Die Size | 398 mm² | Not listed |

| Transistor Density | 13.1M / mm² | Not listed |

| Base Clock | 1127 MHz | 300 MHz |

| Boost Clock | 1216 MHz | 1300 MHz |

| Memory Size | 4 GB | System Shared |

| Memory Type | GDDR5 | System Shared |

| Memory Bus Width | 256 bit | System Shared |

| Memory Bandwidth | 224.4 GB/s | System Dependent |

| Shading Units | 2048 | 256 |

| TMUs | 128 | 16 |

| ROPs | 64 | 8 |

| Pixel Rate | 77.82 GPixel/s | 10.40 GPixel/s |

| Texture Rate | 155.6 GTexel/s | 20.80 GTexel/s |

| FP32 | 4.981 TFLOPS | 665.6 GFLOPS |

| FP16 | Not listed | 1,331.2 GFLOPS (2:1) |

| TDP | 165 W | 15 W |

| Slot Width | Dual-slot | IGP |

| Power Connectors | 2x 6-pin | None |

| Suggested PSU | 450 W | None |

| Bus Interface | PCIe 3.0 x16 | Ring Bus |

| Display Outputs | 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.2 | Motherboard Dependent |

| Release Date | 2014-09-18 | 2021-03-29 |

| Launch MSRP | 549 USD | None |

The specification table shows a generational and architectural chasm. The GTX 980 is a discrete, high-power card with dedicated memory and a physical footprint of 267 mm in length, 111 mm in height, and 40 mm in width. The UHD 750 is an integrated solution with no physical dimensions listed, no power connectors, and a TDP that is 11x lower. Notably, the UHD 750 has a higher boost clock (1300 MHz vs 1216 MHz), but this is meaningless given its 8x fewer shading units and shared memory.

# Head-to-Head Benchmarks

The head-to-head benchmark data is limited to two tests, but both are lopsided. In Geekbench OpenCL, the GTX 980 scores 34,676 against 6,743 for the UHD 750. The delta is 414.3%, meaning the NVIDIA card is more than five times faster. This test measures general-purpose compute using OpenCL, and the result reflects the GTX 980’s 2,048 shading units and dedicated memory bandwidth.

In Geekbench Vulkan, the GTX 980 scores 22,543 versus 8,138 for the Intel part, a 177% delta. Vulkan is a low-level graphics API, and the GTX 980’s advantage here is smaller than in OpenCL but still overwhelming. The UHD 750’s 8,138 score is respectable for an integrated GPU, but it sits far below the discrete card’s output.

Looking beyond the head-to-head tests, the GTX 980’s broader benchmark suite shows strength across legacy and modern APIs. It scores 164 in Passmark DirectX 9, 83 in DirectX 11, 53 in DirectX 10, and 46 in DirectX 12. Its Passmark G3D score is 11,095, with a GPU compute score of 4,753 and a G2D score of 792. It also achieves 15,163 in Geekbench Metal and 474 in 3DMark Steel Nomad DX12. The UHD 750 has no results in these tests, so no direct comparison is possible, but the GTX 980’s average benchmark score of 8,167 versus 7,441 for the Intel part confirms its overall superiority.

The closest rivals for each GPU reinforce the gap. The GTX 980’s nearest competitors — GTX 950M, R9 M360, GeForce 945M, and GRID K2 — all score within 1.1% of the GTX 980’s average. The UHD 750’s nearest rivals — HD 8850M, R7 350, GTX 1650, and Arc A310 — cluster within 1.4% of its average. Notably, the UHD 750 trails the GTX 1650 by 0.4% and the Arc A310 by 1.4%, showing it is competitive with low-end discrete parts from its era, but the GTX 980 operates in a different league entirely.

# The Verdict

The data is unambiguous: the NVIDIA GeForce GTX 980 is the superior GPU by every measurable metric. It wins both head-to-head benchmarks, with a 414.3% lead in OpenCL and a 177% lead in Vulkan. Its average benchmark score is 9.8% higher, and its percentile ranking is 3 points better (43rd vs 40th). The GTX 980’s 4.981 TFLOPS FP32 performance, 224.4 GB/s memory bandwidth, and 64 ROPs make it a capable discrete solution for gaming and compute workloads from its 2014 era, and it remains competitive with mid-range mobile parts like the GTX 950M and R9 M360.

The Intel UHD Graphics 750 is not a competitor to the GTX 980; it is an integrated fallback. Its 15 W TDP, shared memory, and 256 shading units place it in a different performance class. It is suitable for basic display output and light compute in systems without a discrete GPU, and its Vulkan score of 8,138 shows it can handle some modern API workloads. But the benchmark results show no scenario where it wins against the GTX 980.

Who should pick which? Users who need dedicated graphics performance — for gaming, GPU-accelerated rendering, or compute tasks — should choose the GTX 980, provided their system can accommodate its dual-slot footprint, 165 W TDP, and 450 W PSU requirement. Users who prioritize low power consumption, minimal physical footprint, and are building a system where a discrete GPU is unnecessary should consider the UHD 750. The GTX 980 was launched with a 549 USD MSRP in 2014 and is now end-of-life, while the UHD 750, also end-of-life, has no listed MSRP. For pure performance, the GTX 980 wins outright; for integration and efficiency, the UHD 750 has its place, but not in any head-to-head comparison.