Intel UHD Graphics 750 vs NVIDIA GeForce 945M Comparison

NVIDIA GeForce 945M and Intel UHD Graphics 750 represent two distinct approaches to mobile graphics: a discrete MXM module from the Maxwell era versus an integrated Gen 12.1 part built into Rocket Lake processors. The benchmark data shows a clear overall leader in the NVIDIA GeForce 945M, which wins the single head-to-head comparison, but the Intel part counters with superior API support and a different performance profile. The GeForce 945M posts an average benchmark score of 8099 against the UHD 750's 7441, a difference driven largely by the NVIDIA chip's 20.1% lead in the Geekbench OpenCL test. However, the Intel solution offers a DirectX 12 (12_1) feature level, a faster boost clock, and a notably lower power draw, making the choice between them dependent on workload and system constraints.

Where Each One Wins

The NVIDIA GeForce 945M is the clear winner in raw compute throughput as measured by the Geekbench OpenCL benchmark. Its score of 8099 versus 6743 for the Intel UHD Graphics 750 translates to a 20.1% advantage, which is substantial for a single workload metric. This makes the discrete NVIDIA part the better option for OpenCL-accelerated tasks such as general-purpose GPU computing, video encoding filters, or any application that leverages the GPU for non-graphics parallel processing. Its 42nd percentile ranking among all GPUs also edges out the Intel part's 40th percentile, indicating that the GeForce 945M sits slightly higher in the overall performance distribution.

The Intel UHD Graphics 750, while losing the compute comparison, wins on architectural modernity and feature support. It is the only one of the two to support Vulkan 1.4 and DirectX 12 (12_1), whereas the NVIDIA part is limited to DirectX 12 (11_0). This means the Intel iGPU is better positioned for modern games and applications that require the latest graphics API features, even if its raw throughput is lower. The UHD 750 also demonstrates strength in the Vulkan API specifically, scoring 8138 in the Geekbench Vulkan test, which actually exceeds the NVIDIA chip's OpenCL score of 8099. For users running Vulkan-based titles or workloads, the Intel part's API efficiency could narrow or reverse the performance gap seen in OpenCL.

The Intel UHD Graphics 750 also wins decisively on power efficiency. With a TDP of just 15 W compared to the GeForce 945M's 75 W, the integrated solution consumes a fraction of the power. This makes it the only viable choice for thin-and-light laptops where battery life and thermal management take priority over peak performance. The NVIDIA part's 75 W TDP and MXM module form factor require a larger chassis with active cooling, confining it to more substantial gaming or workstation laptops.

Architecture Differences

The two GPUs could not be more different in their fundamental design. The NVIDIA GeForce 945M is built on the Maxwell architecture using the GM107 chip, fabricated on a 28 nm process at TSMC. This chip packs 1,870 million transistors into a 148 mm² die, yielding a transistor density of 12.6 million per square millimeter. The Intel UHD Graphics 750, by contrast, uses Intel's Generation 12.1 architecture on a 14 nm+++ process from Intel's own fabs, with no transistor or die size data provided.

The compute resources differ dramatically. The GeForce 945M fields 640 shading units, 40 texture mapping units, and 16 ROPs, while the Intel part has 256 shading units, 16 TMUs, and 8 ROPs. This 2.5x advantage in shading units and 2.5x in TMUs explains the NVIDIA chip's higher fill rates: 16.32 GPixel/s pixel rate and 40.80 GTexel/s texture rate versus 10.40 GPixel/s and 20.80 GTexel/s for Intel. The FP32 throughput follows the same pattern, with the GeForce 945M delivering 1,305.6 GFLOPS against the UHD 750's 665.6 GFLOPS. Interestingly, the Intel part supports FP16 at 1,331.2 GFLOPS with a 2:1 ratio, a feature the NVIDIA chip does not list, potentially benefiting certain machine learning or media workloads.

Memory architecture is another major divide. The GeForce 945M uses 2 GB of dedicated DDR3 memory on a 128-bit bus, providing 28.80 GB/s of bandwidth. The Intel UHD Graphics 750 uses system shared memory, with bandwidth listed as system dependent. This means the NVIDIA part has guaranteed, dedicated memory bandwidth, while the Intel solution's performance is contingent on the host system's RAM speed and configuration. The clock speeds also differ: the GeForce 945M runs at a 928 MHz base and 1020 MHz boost, whereas the Intel part starts at 300 MHz but boosts to 1300 MHz, giving it a higher peak clock despite the lower base.

Head-to-Head Benchmarks

The only direct benchmark comparison available is the Geekbench OpenCL test, and it shows a decisive NVIDIA victory. The GeForce 945M scores 8099, while the Intel UHD Graphics 750 scores 6743, resulting in a 20.1% delta in favor of the discrete GPU. This is a significant margin that aligns with the hardware specifications: the NVIDIA part has 2.5x more shading units, 2.5x more TMUs, and nearly double the FP32 throughput. The GeForce 945M's dedicated memory bandwidth of 28.80 GB/s also avoids the contention issues that can plague shared-memory iGPUs.

However, the Intel UHD Graphics 750 shows a different side in the Vulkan benchmark, where it scores 8138. While this test was not run on the NVIDIA part, the Intel score alone is notable because it exceeds the NVIDIA chip's OpenCL score of 8099. This suggests that the Intel architecture is more efficient in Vulkan, potentially due to its newer Generation 12.1 design and DirectX 12 (12_1) feature level. If a Vulkan head-to-head existed, the margin could be much closer, or even favor Intel, depending on driver optimization and workload characteristics.

Looking at the nearest rivals provides additional context. The GeForce 945M's closest competitors in the database are the NVIDIA GRID K2 (8080, 0.2% lower), the AMD Radeon R9 M360 (8129, 0.4% higher), and the NVIDIA GeForce GTX 950M (8135, 0.4% higher). This clustering shows the GeForce 945M sits in a tight performance band where 0.4% differences are the norm. The Intel UHD Graphics 750's rivals include the AMD Radeon HD 8850M (7447, 0.1% higher) and the NVIDIA GeForce GTX 1650 (7472, 0.4% higher), placing the Intel iGPU in a lower but still competitive tier. The 20.1% gap between the two reviewed parts is far larger than the gaps within their respective rival clusters.

Specification Differences

The specification sheets reveal several key differences beyond the obvious performance gap. The most striking is the power envelope: the GeForce 945M draws 75 W while the UHD 750 draws only 15 W, a 5x difference that fundamentally alters their deployment scenarios. The NVIDIA part is an MXM-B (3.0) module with no power connectors, while the Intel part is an IGP connected via Ring Bus. The GeForce 945M has a 28 nm process, 1,870 million transistors, and a 148 mm² die; the Intel part's process is 14 nm+++, with no transistor or die size data available.

Memory configuration is another major divergence. The GeForce 945M uses 2 GB of DDR3 on a 128-bit bus with 28.80 GB/s bandwidth. The UHD 750 uses system shared memory with a system dependent bandwidth and bus width. This means the NVIDIA part has predictable, fixed memory performance, while the Intel solution's memory throughput varies with the host system's RAM. The clock speeds differ as well: the NVIDIA part has a 928 MHz base and 1020 MHz boost, while the Intel part has a 300 MHz base and 1300 MHz boost. The Intel boost clock is 27.5% higher than the NVIDIA boost, yet the NVIDIA part still wins on throughput due to its massive hardware advantage.

API support also differs. Both support OpenGL 4.6 and Vulkan 1.4, but the NVIDIA part is limited to DirectX 12 (11_0) while the Intel part supports DirectX 12 (12_1). This makes the Intel UHD 750 more future-proof for DirectX 12 games that use feature level 12_1, such as those requiring conservative rasterization or rasterizer-ordered views. The GeForce 945M also has a higher pixel rate (16.32 GPixel/s vs 10.40 GPixel/s) and texture rate (40.80 GTexel/s vs 20.80 GTexel/s), but the Intel part has the FP16 advantage at 1,331.2 GFLOPS.

FAQ

Q: Which GPU is faster in OpenCL compute?

A: The NVIDIA GeForce 945M scores 8099 in Geekbench OpenCL, which is 20.1% higher than the Intel UHD Graphics 750's 6743. This aligns with the NVIDIA chip's 640 shading units versus Intel's 256, and its 1,305.6 GFLOPS FP32 throughput versus 665.6 GFLOPS.

Q: Does the Intel UHD Graphics 750 support newer graphics APIs?

A: Yes, the Intel part supports DirectX 12 (12_1), while the NVIDIA GeForce 945M only supports DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4, but the Intel part's higher DirectX feature level gives it an edge in modern DirectX 12 titles.

Q: How do their power requirements compare?

A: The NVIDIA GeForce 945M has a TDP of 75 W, while the Intel UHD Graphics 750 has a TDP of 15 W. This 60 W difference means the Intel iGPU is far more suitable for battery-powered devices, whereas the NVIDIA part requires a larger cooling solution.

Q: What is the memory configuration difference?

A: The GeForce 945M uses 2 GB of dedicated DDR3 memory on a 128-bit bus with 28.80 GB/s bandwidth. The Intel UHD Graphics 750 uses system shared memory with system dependent bandwidth, meaning its performance relies on the host system's RAM speed.

Q: Can the Intel part win in any benchmark scenario?

A: The Intel UHD Graphics 750 scores 8138 in Geekbench Vulkan, which is higher than the NVIDIA part's OpenCL score of 8099. However, no direct Vulkan comparison exists between the two, so a head-to-head result in that API is unverified.

Q: How do their overall performance percentiles compare?

A: The NVIDIA GeForce 945M sits at the 42nd percentile among all GPUs, while the Intel UHD Graphics 750 sits at the 40th percentile. The NVIDIA part's average benchmark score is 8099, versus 7441 for the Intel part, reflecting its higher compute performance.