Intel UHD Graphics 730 vs NVIDIA GeForce GTX 675M Comparison

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 675M records an average benchmark score of 6946, while the Intel UHD Graphics 730 averages 5929. This places the NVIDIA part roughly 17% higher in overall performance based on the database's recorded measurements.

Q: How do the two compare in the OpenCL benchmark specifically?

A: In the GeekBench OpenCL test, the GTX 675M scores 6946 against the UHD 730's 5988, giving the NVIDIA GPU a 16% advantage. This is the only head-to-head benchmark recorded in the database, and the GTX 675M wins it outright.

Q: What is the percentile ranking for each GPU?

A: The GTX 675M sits in the 39th percentile among all GPUs in the database, while the UHD 730 sits in the 33rd percentile. The six-point gap indicates the NVIDIA GPU outperforms a larger share of the overall GPU population.

Q: Which GPU supports the Vulkan API?

A: The Intel UHD Graphics 730 lists Vulkan 1.4 support, while the GTX 675M has no recorded Vulkan API entry. Both support DirectX 12 and OpenGL 4.6, though the GTX 675M's DirectX 12 support is at feature level 11_0.

Q: What is the pixel fillrate difference between the two?

A: The Intel UHD Graphics 730 achieves 10.40 GPixel/s, slightly higher than the GTX 675M's 9.920 GPixel/s. Despite the NVIDIA GPU's overall benchmark lead, the Intel part has a small edge in pixel throughput.

Q: How does memory configuration differ?

A: The GTX 675M uses 2 GB of dedicated GDDR5 memory on a 256-bit bus with 96.00 GB/s bandwidth. The UHD 730 uses system-shared memory with bandwidth described as "System Dependent," meaning its performance relies entirely on the host system's RAM.

Architecture Differences

The two GPUs come from fundamentally different design philosophies and eras. The NVIDIA GeForce GTX 675M uses the GF114 chip built on the Fermi 2.0 architecture, fabricated by TSMC at 40 nm. This is a large discrete GPU with 1,950 million transistors packed into a 332 mm² die, giving a transistor density of 5.9 million per square millimeter. The Intel UHD Graphics 730, by contrast, is a Rocket Lake integrated graphics solution built on Generation 12.1 architecture, manufactured by Intel at 14 nm+++. Its transistor count and die size are not recorded in the database, but the integrated nature means it shares silicon with the CPU.

The compute resource allocation differs substantially. The GTX 675M carries 384 shading units, 64 texture mapping units, and 32 raster operation units. The UHD 730 has 192 shading units, 12 TMUs, and 8 ROPs. The NVIDIA GPU therefore fields twice the shader count, over five times the texture units, and four times the ROPs. These structural differences directly explain the benchmark gap.

Clock behavior also diverges. The GTX 675M has no recorded base or boost clock, but its memory runs at 750 MHz (3 Gbps effective). The UHD 730 has a base clock of 300 MHz and a boost clock of 1300 MHz, with system-shared memory. The Intel part compensates for its smaller execution units with a higher boost clock, but the sheer resource advantage of the NVIDIA chip prevails.

The GTX 675M occupies an MXM Module slot with a 256-bit memory bus, while the UHD 730 is an IGP connected via Ring Bus. Display outputs are "Portable Device Dependent" for the NVIDIA GPU versus "Motherboard Dependent" for the Intel. The GTX 675M has no power connectors listed, while the UHD 730's power delivery comes through the motherboard as an integrated component.

API support shows a notable generational split. Both support DirectX 12 and OpenGL 4.6, but the GTX 675M's DirectX 12 is at feature level 11_0, while the UHD 730 reaches 12_1. The Intel part also supports Vulkan 1.4, which the NVIDIA GPU lacks entirely. This gives the UHD 730 a forward-looking API advantage despite its lower raw performance.

Head-to-Head Benchmarks

The database records one direct comparison between these GPUs: GeekBench OpenCL. The GTX 675M scores 6946, and the UHD 730 scores 5988. The delta is 16% in favor of NVIDIA. This is a decisive margin in a compute-oriented benchmark, reflecting the GTX 675M's larger shader count and dedicated memory subsystem.

Looking at the rival context reinforces the gap. The GTX 675M's nearest rivals are all within a tight band: the AMD Radeon R5 M240 scores 6975 (-0.4%), the NVIDIA GeForce GTX 680M scores 7023 (-1.1%), the NVIDIA T600 scores 7035 (-1.3%), and the AMD FirePro M5100 scores 6830 (+1.7%). The GTX 675M essentially trades blows with these mid-range discrete parts, sitting right in the middle of that cluster.

The UHD 730's nearest rivals cluster even more tightly around its own score. The AMD Radeon HD 8730M scores 5955 (-0.4%), the NVIDIA Quadro K620M scores 5957 (-0.5%), the AMD Radeon HD 8750M scores 5970 (-0.7%), and the NVIDIA Quadro K4000 scores 5982 (-0.9%). Every rival is within one point-percent of the Intel GPU, indicating that the UHD 730 lands squarely in a well-populated performance tier.

The practical interpretation: the GTX 675M outperforms the UHD 730 by roughly the same margin that separates the UHD 730 from a slightly weaker integrated solution. The 16% delta in OpenCL is consistent with the average benchmark gap of about 17%. There is no benchmark in the database where the UHD 730 wins.

Specification Differences

The two GPUs differ across nearly every recorded specification field. The GTX 675M uses a 40 nm TSMC process; the UHD 730 uses Intel's 14 nm+++. The NVIDIA chip contains 1,950 million transistors on a 332 mm² die; the Intel part's transistor count and die size are not recorded. The GTX 675M's transistor density is 5.9M per mm²; no density figure exists for the UHD 730.

Memory represents the starkest contrast. The GTX 675M has 2 GB of GDDR5 on a 256-bit bus with 96.00 GB/s bandwidth. The UHD 730 uses system-shared memory with no dedicated capacity, no fixed bus width, and bandwidth listed as system-dependent. This means the NVIDIA GPU's memory performance is consistent and self-contained, while the Intel GPU's memory performance varies with the host system's RAM speed and configuration.

Shading resources: 384 versus 192 units. Texture units: 64 versus 12. ROPs: 32 versus 8. Pixel rate: 9.920 GPixel/s versus 10.40 GPixel/s. Texture rate: 39.68 GTexel/s versus 15.60 GTexel/s. FP32 compute: 952.3 GFLOPS versus 499.2 GFLOPS. The UHD 730 lists FP16 at 998.4 GFLOPS (2:1), while the GTX 675M has no FP16 entry.

Power draw differs by a factor of nearly seven: 100 W for the GTX 675M versus 15 W for the UHD 730. The NVIDIA GPU uses an MXM Module slot; the Intel part is an IGP. Bus interface: MXM-B (3.0) versus Ring Bus. Display outputs: portable-device-dependent versus motherboard-dependent.

Release dates are nine years apart: March 2012 for the GTX 675M and March 2021 for the UHD 730. Both are listed as end-of-life. The GTX 675M's predecessor is the GeForce 500M and its successor is the GeForce 700M; the UHD 730 has neither predecessor nor successor recorded. Both have no launch MSRP in the database.

Where Each One Wins

The NVIDIA GeForce GTX 675M wins every recorded benchmark comparison. It takes the OpenCL test by 16% and holds the higher average score by 17%. Its dedicated GDDR5 memory with 96.00 GB/s bandwidth gives it a structural advantage in memory-bound workloads, since it does not compete with the CPU for system RAM. The 384 shading units and 64 TMUs make it substantially stronger in compute-heavy and texture-heavy tasks, as reflected in its 952.3 GFLOPS FP32 throughput versus the UHD 730's 499.2 GFLOPS.

The Intel UHD Graphics 730 wins in specific non-benchmark dimensions. It has a higher pixel rate at 10.40 GPixel/s, meaning it can fill more pixels per second despite its lower overall compute. It supports newer API features: DirectX 12_1 versus 11_0, and Vulkan 1.4 versus none. Its 15 W TDP makes it far more power-efficient than the GTX 675M's 100 W, and as an IGP it requires no separate graphics card slot or power connector. Its 1300 MHz boost clock provides a higher peak frequency than the GTX 675M's unrecorded clocks.

For gaming use cases, the GTX 675M is the clear choice based on benchmark data: its OpenCL score is 16% higher, and its nearest rivals are all discrete mobile GPUs, indicating it sits in a performance tier above integrated graphics. The UHD 730's nearest rivals include older discrete parts like the Radeon HD 8730M and Quadro K620M, but its integrated nature means it will appeal to systems where power draw and physical space are constrained.

The Verdict

The data points to a straightforward conclusion: the NVIDIA GeForce GTX 675M is the faster GPU. Its 6946 OpenCL score against the UHD 730's 5988, its 16% head-to-head win, its 39th versus 33rd percentile ranking, and its 952.3 GFLOPS FP32 versus 499.2 GFLOPS all confirm this. Anyone selecting a GPU purely for compute performance should choose the GTX 675M.

However, the UHD 730 is not without merit. Its 15 W TDP makes it suitable for low-power systems where the GTX 675M's 100 W draw is impractical. Its Vulkan 1.4 support and DirectX 12_1 feature level give it modern API compatibility that the older Fermi-based GPU cannot match. Its higher pixel rate of 10.40 GPixel/s offers a small advantage in fill-rate-bound scenarios.

The verdict depends on the use case. For gaming, rendering, or any compute-heavy workload recorded in the database, the GTX 675M wins outright. For a system that prioritizes power efficiency, modern API support, and integrated simplicity, the UHD 730 makes sense despite its lower scores. The benchmark database records one win for the GTX 675M and zero for the UHD 730; the performance story is unambiguous, while the efficiency story favors Intel.