Intel UHD Graphics 750 vs NVIDIA GeForce GTX 1650 Comparison

The NVIDIA GeForce GTX 1650 and Intel UHD Graphics 750 sit at nearly the same point in the aggregate benchmark hierarchy, with average scores of 7472 and 7441 respectively, yet they achieve that parity through radically different designs. The GTX 1650 is a discrete, 75 W Turing-based card from NVIDIA, while the UHD 750 is a 15 W integrated GPU built into Intel’s Rocket Lake processors. Despite the 0.4% average score gap favoring the NVIDIA part, the two GPUs are separated by over 300% in specific compute workloads, making their near-identical averages a statistical coincidence rather than a sign of equivalent capability.

FAQ

Q: How do the average benchmark scores of the GTX 1650 and UHD 750 compare?

A: The GTX 1650 has an average benchmark score of 7472, while the UHD 750 scores 7441. That puts the NVIDIA card 0.4% ahead, with the Intel part trailing by the same margin. Both GPUs share the 40th percentile ranking among all GPUs.

Q: Which GPU wins in Geekbench OpenCL performance?

A: The GTX 1650 dominates, scoring 29629 versus the UHD 750’s 6743. That is a 339.4% advantage for the NVIDIA card, representing a more than fourfold difference in raw compute throughput.

Q: What about Vulkan performance?

A: The GTX 1650 again wins decisively with a score of 33042, compared to the UHD 750’s 8138. The delta is 306%, meaning the discrete card delivers roughly four times the Vulkan performance of the integrated solution.

Q: Are these GPUs considered rivals based on the data?

A: Yes, they are listed as nearest rivals to each other. The GTX 1650’s nearest rivals include the UHD 750 (deltaPct 0.4%), and the UHD 750’s nearest rivals include the GTX 1650 (deltaPct -0.4%). Their average scores are nearly identical.

Q: What is the TDP difference between the two?

A: The GTX 1650 has a TDP of 75 W, while the UHD 750 is rated at 15 W. That makes the Intel integrated GPU five times more power-efficient on paper, though it lacks the performance headroom of the discrete card.

Q: Which GPU supports more display outputs?

A: The GTX 1650 provides 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a. The UHD 750’s display outputs are listed as “Motherboard Dependent,” meaning they vary based on the specific motherboard implementation.

Architecture Differences

The GTX 1650 is built on NVIDIA’s Turing architecture using the TU117 chip, manufactured on a 12 nm process at TSMC. The UHD 750 uses Intel’s Generation 12.1 architecture, implemented on a 14 nm+++ process and integrated into the Rocket Lake chip. The NVIDIA part contains 4,700 million transistors on a 200 mm² die, giving it a transistor density of 23.5M per mm². Intel does not disclose transistor count, die size, or density for the UHD 750 in the data.

The GTX 1650 is a discrete card with 4 GB of GDDR5 memory on a 128-bit bus, delivering 128.1 GB/s of bandwidth. The UHD 750 uses System Shared memory, with its bus width and bandwidth both listed as “System Shared” and “System Dependent” respectively. That means the Intel GPU’s memory performance is tied entirely to the host system’s RAM configuration.

The two differ fundamentally in their compute resources. The GTX 1650 has 896 shading units, 56 texture mapping units, and 32 ROPs. The UHD 750 has 256 shading units, 16 TMUs, and 8 ROPs. The NVIDIA card’s clock speeds are 1485 MHz base and 1665 MHz boost, while the Intel part runs at 300 MHz base and 1300 MHz boost. The GTX 1650’s pixel rate is 53.28 GPixel/s and texture rate is 93.24 GTexel/s, versus 10.40 GPixel/s and 20.80 GTexel/s for the UHD 750.

Both GPUs support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. Neither includes ray tracing or tensor cores. The GTX 1650 is a dual-slot card measuring 229 mm in length, 111 mm in height, and 35 mm in width, with no power connectors required. The UHD 750 is an IGP with no physical dimensions listed, naturally requiring no power connectors. The NVIDIA card interfaces via PCIe 3.0 x16, while the Intel part uses a Ring Bus.

Head-to-Head Benchmarks

The only two head-to-head benchmarks available are Geekbench OpenCL and Geekbench Vulkan, and the GTX 1650 wins both outright. In OpenCL, the NVIDIA card scores 29629 against the UHD 750’s 6743, producing a 339.4% delta. This is not a marginal victory; it is a complete rout. The discrete GPU’s dedicated GDDR5 memory and 896 shading units overwhelm the integrated part’s 256 shading units and shared memory pool.

In Vulkan, the GTX 1650 scores 33042, while the UHD 750 manages 8138. The 306% delta confirms that the NVIDIA card’s advantage persists across different graphics APIs. Interestingly, the GTX 1650 performs better in Vulkan than in OpenCL relative to its own scores, while the UHD 750 also shows a higher Vulkan score than OpenCL. Both GPUs benefit from Vulkan’s lower overhead, but the absolute gap remains roughly fourfold.

The aggregate benchmark data tells a different story. The GTX 1650’s average score of 7472 comes from a broader set of tests, including 3DMark Steel Nomad DX12 (305), Passmark G3D (7880), and Passmark GPU Compute (3048). The UHD 750’s average of 7441 is derived only from its two Geekbench results. This means the Intel part’s near-parity in average score is an artifact of limited test coverage, not genuine equivalence in real-world workloads.

Looking at the nearest rivals clarifies the picture further. The GTX 1650 is 0.3% ahead of the AMD Radeon HD 8850M, 0.4% ahead of the UHD 750, and 0.6% ahead of the AMD Radeon R7 350, while trailing the Intel Arc A310 by 1%. The UHD 750 is 0.1% behind the HD 8850M, 0.2% ahead of the R7 350, 0.4% behind the GTX 1650, and 1.4% behind the Arc A310. These tiny deltas indicate that the entire cluster of GPUs—including the GTX 1650—performs within a 2% band on average, despite the massive compute disparity between the discrete and integrated parts.

Specification Differences

The most obvious difference is form factor: the GTX 1650 is a dual-slot discrete card, while the UHD 750 is an integrated GPU with no slot width. The NVIDIA part has a TDP of 75 W and a suggested PSU of 250 W, whereas the Intel part draws 15 W and has no suggested PSU listed. Memory configurations are polar opposites—4 GB GDDR5 on a 128-bit bus versus System Shared memory with system-dependent bandwidth.

The GTX 1650’s clock speeds are substantially higher: 1485 MHz base and 1665 MHz boost, compared to 300 MHz base and 1300 MHz boost for the UHD 750. The NVIDIA card’s memory clock is 2001 MHz (8 Gbps effective), while the Intel part has no dedicated memory clock. Shader resources differ by a factor of 3.5 in shading units (896 vs 256), 3.5 in TMUs (56 vs 16), and 4 in ROPs (32 vs 8).

Pixel and texture rates reflect those resource differences, with the GTX 1650 producing 53.28 GPixel/s and 93.24 GTexel/s versus 10.40 GPixel/s and 20.80 GTexel/s for the UHD 750. FP32 compute is 2.984 TFLOPS for the NVIDIA card and 665.6 GFLOPS for the Intel part, a 4.5x gap. FP16 performance shows a similar ratio: 5.967 TFLOPS versus 1,331.2 GFLOPS, both using a 2:1 ratio.

The GTX 1650 has a launch MSRP of 149 USD; the UHD 750 has no launch MSRP listed. Manufacturing nodes differ: 12 nm TSMC for NVIDIA versus 14 nm+++ Intel for the integrated part. The GTX 1650 was released on 2019-04-22 and is end-of-life, while the UHD 750 launched on 2021-03-29 and is also end-of-life. The NVIDIA card’s predecessor is GeForce 10 and successor is GeForce 20; the Intel part lists neither.

The Verdict

The data points to a clear performance hierarchy: the GTX 1650 is categorically faster in compute-heavy workloads, while the UHD 750 is in a different class entirely for power consumption. If the task is 3D rendering, GPU compute, or any workload that stresses the shading units, the GTX 1650’s 339.4% OpenCL advantage and 306% Vulkan advantage make it the only viable choice. The UHD 750’s 15 W TDP and system-shared memory are simply not competitive for those tasks.

However, the average benchmark scores tell a more nuanced story. Both GPUs sit at the 40th percentile, and their average scores differ by just 0.4%. This suggests that in aggregate, across a wide mix of tests, the two perform similarly—likely because the UHD 750’s integrated design excels at lighter, less demanding workloads where the GTX 1650’s discrete memory and compute resources are underutilized. The GTX 1650’s nearest rival list includes the UHD 750 with a deltaPct of 0.4%, and the UHD 750’s nearest rival list includes the GTX 1650 with a deltaPct of -0.4%, confirming they occupy the same tier.

For a system builder, the choice is straightforward. The GTX 1650 is for anyone who needs consistent, predictable discrete GPU performance across a wide range of applications, including DirectX 12 titles and Vulkan games. It offers dedicated GDDR5 memory, a 128-bit bus, and a 40th percentile standing that puts it ahead of many older discrete parts. The UHD 750, by contrast, is for systems where power efficiency and integration matter more than raw throughput. Its 15 W TDP is one-fifth that of the GTX 1650, and it requires no additional cooling or power connectors.

The GTX 1650’s 2.984 TFLOPS FP32 performance and 4 GB memory capacity make it suitable for 1080p gaming at moderate settings. The UHD 750’s 665.6 GFLOPS and shared memory make it more appropriate for basic desktop use, video playback, and light productivity tasks. There is no benchmark in the data where the UHD 750 wins outright, but its power envelope and integrated nature fill a niche that the GTX 1650 cannot address. Users who already own a Rocket Lake processor with UHD 750 graphics may find the performance adequate for non-gaming workloads; those who need more should opt for the GTX 1650, despite its higher power draw and dual-slot footprint.