Intel UHD Graphics 730 vs NVIDIA GeForce 940M Comparison

Intel UHD Graphics 730 and NVIDIA GeForce 940M represent two different approaches to mobile graphics, one as a modern integrated solution and the other as a dedicated discrete part from a previous generation. Benchmark data from the database shows a split decision: the NVIDIA GeForce 940M edges out a narrow victory in OpenCL compute workloads, while the Intel UHD Graphics 730 delivers a decisive win in Vulkan performance. The average benchmark scores tell a broader story, with Intel holding a 12.2% lead overall (5929 versus 5284), which places the integrated part in a higher percentile ranking of all GPUs (33rd versus 31st).

FAQ

Q: Which GPU has the higher average benchmark score?

A: The Intel UHD Graphics 730 scores 5929 on average, which is 12.2% higher than the NVIDIA GeForce 940M's 5284 average.

Q: How do the two GPUs compare in Vulkan performance?

A: The Intel UHD Graphics 730 wins decisively with a score of 5870, which is 29% higher than the NVIDIA GeForce 940M's 4549 in the same test.

Q: Which GPU performs better in OpenCL?

A: The NVIDIA GeForce 940M wins the OpenCL test with a score of 6018, narrowly beating the Intel UHD Graphics 730's 5988 by 0.5%.

Q: What is the memory configuration difference?

A: The Intel UHD Graphics 730 uses System Shared memory with system-dependent bandwidth, while the NVIDIA GeForce 940M has 2 GB of dedicated DDR3 memory on a 64-bit bus with 14.40 GB/s bandwidth.

Q: What are the power consumption figures?

A: The Intel UHD Graphics 730 has a TDP of 15 W, while the NVIDIA GeForce 940M has a TDP of 75 W, a fivefold difference.

Q: Which GPU has a higher pixel fill rate?

A: The NVIDIA GeForce 940M achieves 17.57 GPixel/s, which is 68.9% higher than the Intel UHD Graphics 730's 10.40 GPixel/s.

Architecture Differences

The two GPUs come from fundamentally different design philosophies and manufacturing generations. The Intel UHD Graphics 730 is built on Rocket Lake silicon using Intel's 14 nm+++ process, whereas the NVIDIA GeForce 940M uses the GM107 chip fabricated by TSMC on a 28 nm process. This process node difference is substantial, with Intel's newer node offering improved transistor density characteristics despite the older process name.

The compute architecture diverges sharply. Intel's Generation 12.1 architecture (also labeled HD Graphics for Rocket Lake) employs 192 shading units, 12 texture mapping units, and 8 raster output units. NVIDIA's Maxwell architecture in the GeForce 940M deploys 512 shading units, 32 TMUs, and 16 ROPs. This gives the NVIDIA part a 2.67x advantage in shading units, 2.67x in TMUs, and 2x in ROPs, explaining its higher theoretical fill rates.

Transistor counts and die sizes are only recorded for the NVIDIA part: 1,870 million transistors on a 148 mm² die, yielding a transistor density of 12.6M per mm². The Intel part has no recorded transistor or die size data. Clock behavior also differs: Intel runs at a 300 MHz base with a 1300 MHz boost, while NVIDIA runs much higher at a 1020 MHz base and 1098 MHz boost. Despite the higher base clock, NVIDIA's boost is lower than Intel's, which reflects different power and thermal envelopes.

The memory subsystem is entirely different. Intel uses System Shared memory with system-dependent bandwidth, a hallmark of integrated graphics. NVIDIA uses 2 GB of dedicated DDR3 on a 64-bit bus with 14.40 GB/s bandwidth and an effective 1800 Mbps memory clock. This dedicated memory approach gives NVIDIA predictable performance, whereas Intel depends on the host system's memory configuration.

Where Each One Wins

The Vulkan benchmark is the Intel UHD Graphics 730's strongest territory. Its 5870 score against the GeForce 940M's 4549 represents a 29% advantage, the largest margin in any test recorded. This suggests the newer Intel architecture handles Vulkan's modern API features more efficiently, likely due to its DirectX 12 (12_1) support and Generation 12.1 design. Users running Vulkan-based games or applications should expect noticeably smoother performance from the Intel part.

The NVIDIA GeForce 940M wins in OpenCL, scoring 6018 against Intel's 5988. The margin is slim at 0.5%, but it shows NVIDIA's Maxwell architecture remains competitive in compute workloads. The GeForce 940M also holds advantages in raw throughput metrics: its 1,124.4 GFLOPS FP32 performance is more than double Intel's 499.2 GFLOPS, and its texture rate of 35.14 GTexel/s is 2.25x higher than Intel's 15.60 GTexel/s. These numbers suggest the NVIDIA part would win in texture-heavy workloads and general compute tasks.

The power envelope tells a different story. Intel's 15 W TDP is one-fifth of NVIDIA's 75 W TDP, making the integrated solution far more suitable for thin-and-light laptops where battery life and thermals are priorities. The NVIDIA GeForce 940M, as an MXM Module with no power connectors, targets larger laptops or portable workstations that can accommodate its higher power draw.

Specification Differences

The recorded specifications reveal several clear contrasts between the two GPUs. The process node differs: Intel uses 14 nm+++ while NVIDIA uses 28 nm, with the latter manufactured by TSMC. The NVIDIA chip has documented transistor data (1,870 million on a 148 mm² die), while Intel's chip has no such figures recorded.

Clock speeds show Intel boosting higher (1300 MHz) from a lower base (300 MHz), while NVIDIA runs a more conservative 1020 MHz base to 1098 MHz boost. Memory configurations are fundamentally different: Intel uses System Shared memory with system-dependent bandwidth, while NVIDIA has 2 GB of DDR3 on a 64-bit bus with 14.40 GB/s bandwidth. The NVIDIA memory runs at 900 MHz with 1800 Mbps effective speed.

Compute resources heavily favor NVIDIA: 512 shading units versus 192, 32 TMUs versus 12, and 16 ROPs versus 8. This translates to NVIDIA's pixel rate of 17.57 GPixel/s versus Intel's 10.40 GPixel/s, and texture rate of 35.14 GTexel/s versus 15.60 GTexel/s. FP32 performance is also lopsided: 1,124.4 GFLOPS for NVIDIA versus 499.2 GFLOPS for Intel. Intel does offer FP16 support at 998.4 GFLOPS (2:1 ratio), while NVIDIA has no recorded FP16 figure.

Power and form factor differ significantly: Intel has a 15 W TDP and IGP slot width, while NVIDIA has a 75 W TDP and MXM Module form factor. The bus interface also differs: Intel uses Ring Bus, while NVIDIA uses MXM-B (3.0). API support is similar for DirectX (Intel supports 12_1, NVIDIA supports 11_0) and OpenGL 4.6 both, with Vulkan 1.4 recorded for both.

Head-to-Head Benchmarks

The two recorded benchmarks split the result exactly one win each. In geekbench_opencl, the NVIDIA GeForce 940M scores 6018 against Intel's 5988, a delta of -0.5% favoring NVIDIA. This is an extremely tight margin that falls within typical run-to-run variance, suggesting the two GPUs are effectively equivalent in OpenCL compute performance despite NVIDIA's theoretical FP32 advantage of 2.25x.

The geekbench_vulkan test produces a far more decisive outcome. Intel UHD Graphics 730 scores 5870, while the NVIDIA GeForce 940M manages only 4549. The 29% delta in Intel's favor is the standout result in this comparison. This substantial margin indicates that the Intel architecture extracts far more real-world performance from Vulkan workloads, likely because its Generation 12.1 design was created after Vulkan's specification matured, allowing for more efficient driver and hardware integration.

Looking at the nearest rivals provides additional context. The Intel UHD Graphics 730's average score of 5929 places it within 0.9% of the NVIDIA Quadro K4000 (5982) and within 0.7% of the AMD Radeon HD 8750M (5970). The NVIDIA GeForce 940M's average of 5284 sits close to the GeForce GTX 980M (5308, -0.4%) and GeForce 930A (5317, -0.6%), while running 0.9% ahead of the GeForce GTX 760M (5236). These rival relationships show that the Intel part competes with higher-tier mobile GPUs from the same era, while the NVIDIA part aligns with mid-range options.

The Verdict

The data supports a clear split recommendation. Users prioritizing Vulkan performance should choose the Intel UHD Graphics 730, which delivers 29% higher scores in that API. This advantage, combined with a 12.2% higher average benchmark score and a 33rd percentile ranking versus the 31st percentile for NVIDIA, makes the Intel part the stronger overall performer in the recorded tests.

Users who rely on OpenCL compute workloads should consider the NVIDIA GeForce 940M, which wins that specific benchmark by 0.5%. The NVIDIA part also offers substantially higher theoretical throughput: 2.25x more FP32 GFLOPS, 2.25x higher texture rate, and 1.69x higher pixel rate. For applications that scale with raw shader and texture throughput, the NVIDIA architecture provides more headroom.

Power-sensitive deployments favor the Intel UHD Graphics 730 overwhelmingly. Its 15 W TDP versus 75 W represents a 5x difference, making it suitable for systems where thermal and power budgets are constrained. The integrated nature of the Intel part, with System Shared memory, also simplifies system design. The NVIDIA GeForce 940M, with its dedicated 2 GB DDR3 memory and MXM Module form factor, targets systems that can accommodate discrete graphics and its higher power draw.

The production status of both parts is end-of-life, with Intel releasing in March 2021 and NVIDIA in March 2015. The NVIDIA GeForce 940M has documented predecessors and successors (GeForce 800M and GeForce 10 Mobile), while the Intel part has no recorded lineage. For modern workloads and longevity, the newer Intel architecture shows better API support with DirectX 12 (12_1) versus NVIDIA's 12 (11_0), and its Vulkan 1.4 support matches NVIDIA's. The choice ultimately depends on whether the priority is modern API efficiency and power economy or raw compute throughput and dedicated memory.