Intel UHD Graphics 750 vs NVIDIA GeForce GTX 970 Comparison

Intel UHD Graphics 750 is an integrated graphics solution built into Intel's Rocket Lake processors, while the NVIDIA GeForce GTX 970 is a dedicated, dual-slot discrete graphics card from the Maxwell 2.0 generation. The benchmark data is unequivocal: the GTX 970 wins both head-to-head tests by massive margins, delivering roughly 4.4x the OpenCL score and 2.5x the Vulkan score of the Intel UHD 750. However, the GTX 970's average benchmark score of 7157 is actually slightly lower than the Intel UHD 750's 7441 average, a quirk driven by the inclusion of different test suites for each card. The Intel part holds a 40th percentile ranking versus the GTX 970's 39th percentile, placing both in the lower-middle tier of all GPUs, but the nature of their performance is entirely different.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The Intel UHD Graphics 750 has a higher average benchmark score of 7441, compared to the NVIDIA GeForce GTX 970's average of 7157. This positions the Intel part at the 40th percentile, one point above the GTX 970's 39th percentile, though both are near the middle of the database.

Q: How does the GTX 970 perform in the Geekbench OpenCL test?

A: The GTX 970 achieves a score of 29982, which is 77.5% higher than the Intel UHD 750's score of 6743. This represents a dominant win for the NVIDIA card in this compute-oriented test.

Q: What is the difference in their memory configurations?

A: The GTX 970 uses 4 GB of dedicated GDDR5 memory on a 256-bit bus with 224.4 GB/s bandwidth, while the Intel UHD 750 uses System Shared memory with a System Dependent bandwidth. The GTX 970's memory clock is 1753 MHz (7 Gbps effective).

Q: Which GPU is more power-efficient based on TDP?

A: The Intel UHD Graphics 750 has a 15 W TDP, which is dramatically lower than the GTX 970's 148 W TDP. This makes the Intel part suitable for systems without dedicated power connectors, as it uses a Ring Bus interface and is an IGP.

Q: What are the closest rivals to each GPU according to the data?

A: The Intel UHD 750's nearest rival is the AMD Radeon HD 8850M with a deltaPct of -0.1%, while the GTX 970's nearest rival is the Intel Iris Pro Graphics P580 with a deltaPct of -0.2%. The GTX 970's other rivals include the NVIDIA GeForce GTX 560 SE and the AMD Radeon Vega 8 Mobile.

Q: Which GPU has more shading units and texture mapping units?

A: The NVIDIA GeForce GTX 970 has 1664 shading units and 104 TMUs, while the Intel UHD 750 has only 256 shading units and 16 TMUs. This specification gap directly explains the GTX 970's superior raw throughput.

Where Each One Wins

The NVIDIA GeForce GTX 970 wins every single head-to-head benchmark listed in the data. In the Geekbench OpenCL test, the GTX 970 scores 29982 versus the Intel UHD 750's 6743, a delta of -77.5% from the Intel part's perspective. In the Geekbench Vulkan test, the GTX 970 scores 20005 versus 8138, a delta of -59.3%. These are not close contests; the GTX 970's dedicated memory bandwidth of 224.4 GB/s and its 3.920 TFLOPS of FP32 performance dwarf the Intel part's 665.6 GFLOPS and system-dependent memory.

The Intel UHD Graphics 750 wins only in the context of its average benchmark score and percentile ranking. Its average of 7441 beats the GTX 970's 7157, and its 40th percentile is one point higher. This is a statistical artifact of the different benchmark suites used; the Intel part has only two benchmarks (OpenCL and Vulkan) while the GTX 970 has eleven, including several legacy DirectX tests where it scores poorly (e.g., Passmark DirectX 9 score of 143). For any modern workload, the GTX 970 is the clear winner, but the Intel part's lower TDP and integrated nature make it the only option for systems without a PCIe slot.

Architecture Differences

The two GPUs are built on fundamentally different architectures and processes. The Intel UHD 750 uses Intel's Generation 12.1 architecture on a 14 nm+++ process node, fabricated by Intel itself. It is integrated into the Rocket Lake chip and has no discrete transistor count or die size listed. In contrast, the GTX 970 uses NVIDIA's Maxwell 2.0 architecture on a 28 nm process node from TSMC, with 5,200 million transistors on a 398 mm² die, giving a transistor density of 13.1M per mm².

The Intel UHD 750 has 256 shading units, 16 TMUs, and 8 ROPs, while the GTX 970 has 1664 shading units, 104 TMUs, and 56 ROPs. This is a 6.5x difference in shading units and a 6.5x difference in TMUs. The GTX 970's pixel rate is 65.97 GPixel/s versus the Intel's 10.40 GPixel/s, and its texture rate is 122.5 GTexel/s versus 20.80 GTexel/s. The Intel part does support FP16 at 1,331.2 GFLOPS (2:1), while the GTX 970 has no listed FP16 capability. Both support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4.

Specification Differences

The most obvious specification difference is the memory subsystem. The GTX 970 has 4 GB of GDDR5 on a 256-bit bus with 224.4 GB/s bandwidth, while the Intel UHD 750 uses System Shared memory with System Dependent bandwidth. The GTX 970's memory clock is 1753 MHz (7 Gbps effective), whereas the Intel part's memory clock is listed as "System Shared".

The GTX 970 is a dual-slot card measuring 267 mm in length, 111 mm in height, and 40 mm in width, requiring two 6-pin power connectors and a 300 W suggested PSU. The Intel UHD 750 is an IGP with no slot width, no power connectors, and motherboard-dependent display outputs. The GTX 970's bus interface is PCIe 3.0 x16, while the Intel part uses a Ring Bus. The GTX 970 was released on 2014-09-18 with a launch MSRP of 329 USD, while the Intel UHD 750 was released on 2021-03-29 with no launch MSRP. The GTX 970's predecessor is GeForce 700 and successor is GeForce 10, while the Intel part has no listed predecessor or successor.

Head-to-Head Benchmarks

The two head-to-head benchmarks show an overwhelming victory for the GTX 970. In Geekbench OpenCL, the GTX 970 scores 29982, which is 23239 points higher than the Intel UHD 750's 6743. This represents a deltaPct of -77.5% for the Intel part, meaning the GTX 970 is roughly 4.45x faster. This massive gap is consistent with the GTX 970's 3.920 TFLOPS FP32 throughput versus the Intel's 665.6 GFLOPS, a 5.9x difference.

In Geekbench Vulkan, the GTX 970 scores 20005 versus the Intel UHD 750's 8138, a deltaPct of -59.3%. The GTX 970 leads by 11867 points, making it approximately 2.46x faster. While the Vulkan gap is smaller than the OpenCL gap, it is still a decisive win for the discrete card. The Intel UHD 750's Vulkan score of 8138 is actually higher than its OpenCL score, suggesting the integrated part handles modern APIs relatively better, but it remains far behind the GTX 970's dedicated hardware.

Outside these two tests, the GTX 970 has additional benchmarks that the Intel part lacks, including a Passmark G3D score of 9638 and a Geekbench Metal score of 13595. These scores are not directly comparable to the Intel part but reinforce the GTX 970's broader capability set.

The Verdict

The data is clear: the NVIDIA GeForce GTX 970 is the superior GPU for any task requiring raw performance. It wins both head-to-head benchmarks by enormous margins, with a 77.5% lead in OpenCL and a 59.3% lead in Vulkan. The GTX 970's 1664 shading units, 104 TMUs, and 4 GB of dedicated GDDR5 memory with 224.4 GB/s bandwidth make it a far more capable compute and rendering device than the Intel UHD 750's 256 shading units and system-shared memory.

However, the Intel UHD Graphics 750 has its own domain. Its 15 W TDP, Ring Bus interface, and IGP form factor mean it requires no additional power connectors, no PCIe slot, and no cooling solution beyond what the motherboard provides. It is the only choice for systems built around Rocket Lake processors that lack a discrete GPU. Its average benchmark score of 7441 and 40th percentile ranking are slightly better than the GTX 970's 7157 and 39th percentile, but this is driven by the GTX 970's inclusion of legacy DirectX 9, 10, 11, and 12 tests where it scores between 41 and 143.

For gamers or professionals needing a dedicated graphics card, the GTX 970 is the only viable option from this comparison, despite its 148 W TDP and dual-slot footprint. For a low-power integrated solution, the Intel UHD 750 is adequate but will struggle with any demanding graphical workload. The verdict is not about which is "better" in absolute terms; it is about which fits the use case. The GTX 970 wins on performance, the Intel UHD 750 wins on integration and power efficiency. If you have a PCIe slot and power connectors, choose the GTX 970. If you are building a compact or low-power system, the Intel UHD 750 is the only one of the two that can exist there at all.