Intel UHD Graphics P630 vs NVIDIA GeForce 940MX Comparison

# Intel UHD Graphics P630 vs NVIDIA GeForce 940MX

The data places these two mobile-class graphics processors in different tiers. The Intel UHD P630 wins both recorded head-to-head benchmark comparisons, with the NVIDIA GeForce 940MX failing to take a single test. The margin is not trivial: in the OpenCL workload, the Intel part leads by 3.5 percent, and in the Vulkan test, the gap widens to 18.5 percent. However, the database also shows that both parts sit well below the median of all recorded GPUs, at the 31st and 28th percentiles, respectively. This is a contest between integrated graphics and a low-end discrete part, and the data consistently favors the integrated solution.

The Verdict

The benchmark data in the database is unambiguous for the two workloads measured. The Intel UHD P630 achieves an average score of 5370 across both tests, while the NVIDIA GeForce 940MX averages 4844. That is a 10.9 percent difference in the aggregate, with the Intel part winning both applicable tests. The individual deltas are 3.5 percent in OpenCL and 18.5 percent in Vulkan.

For system builders: the P630 is the stronger part for compute workloads, which is notable given its system-shared memory architecture. The 940MX trails by a consistent margin, and its only claim is compatibility, not performance. For users locked to an NVIDIA ecosystem, the 940MX is sufficient, but the data suggests the Intel solution is faster in every measured workload.

For a laptop buyer: the Intel UHD P630 is the preferable choice when the workload is compute-oriented. The NVIDIA part wins nothing in these tests. For a mobile workstation, the P630 provides 192 shading units, 24 texture mapping units and 3 render output units, which is a full configuration for the Intel architecture.

FAQ

Q: Which part has the higher average benchmark score?

A: The Intel UHD Graphics P630 records an average of 5370 across two benchmarks, the NVIDIA GeForce 940MX averages 4844. The Intel part leads by 10.9 percent on average.

Q: How big is the performance difference, in percentage, for the OpenCL test?

A: The Intel UHD P630 wins the OpenCL workload by 172 points, which equals a 3.5 percent advantage over the GeForce 940MX.

Q: Does the NVIDIA 940MX win any of the head-to-head tests?

A: No. The database lists two head-to-head comparisons, GeForce OpenCL and GeForce Vulkan, and the NVIDIA part loses both. Its best margin is at 18.5 percent in the Vulkan test.

Q: Which GPU is faster in the Vulkan benchmark?

A: The Intel UHD P630 scores 5628 versus 4749, a lead of 18.5 percent, which is the largest margin between these two parts across all workloads.

Q: Where does the GeForce 940MX rank among all tested GPUs?

A: The 940MX sits at the 31st percentile of the database, which places it slightly below the Intel part, and the Intel part is at the 31st percentile, a similar overall position.

Architecture Differences

The two parts use different design philosophies. The Intel UHD P630 is built on the Generation 9.5 graphics architecture, using the Comet Lake GT2 chip. It is fabricated on Intel's 14nm+++ process. The NVIDIA GeForce 940 uses the Maxwell architecture, specifically the GM107 chip, produced by TSMC at 28 nm.

The Intel part integrates 192 shading units, 24 texture mapping units, 3 render output units and no dedicated ray tracing cores. The NVIDIA part has 512 shading units, 32 texture mapping units and 8 render output units, also without ray tracing cores.

Clock behavior differs strongly. The Intel part has a base clock of 350 MHz and a boost of 1200 MHz. The NVIDIA has a higher base of 795 MHz but a lower boost of 861 MHz, which reduces its peak frequency advantage.

Memory configuration shows the largest structural difference. The Intel P630 uses the System Shared memory, with bus width and bandwidth all dependent on the host system. The 940MX has 2 GB of dedicated GDDR5 memory on a 64-bit bus, with a bandwidth of 40.10 GB/s. This gives the discrete part a major bandwidth advantage.

Shader throughput: the Intel part is rated for 460.8 GFLOPS FP32 and 921.6 GFLOPS FP16 (via a 2:1 ratio). NVIDIA has a higher FP32 rating at 881.7 GFLOPS. Pixel throughput also favors NVIDIA: 6.888 GPixel/s versus 3.600 GPixel/s for Intel. Texture fill rate is comparable: 27.55 GTexel/s for NVIDIA versus 28.80 GTexel/s for Intel.

The API support is nearly identical. Both support DirectX 12 (12_1 for NVIDIA, 12_1 for Intel), OpenGL 4.6, and Vulkan 1.4 for Intel, 1.3 for NVIDIA. The bus interface differs: Intel uses a Ring Bus, while NVIDIA uses PCIe 3.0 x8. The NVIDIA is a mobile MXM module, while the Intel is a built-in IGP, i.e., motherboard-dependent.

Specification Differences

| Field | Intel UHD P630 | NVIDIA 940MX |

| --- | --- | --- |

| Architecture | Generation 9.5 (Comet Lake) | Maxwell (GM107) |

| Process | 14nm+++ | 28nm |

| Transistors | Not disclosed | 1,870 million |

| Die Size | Not disclosed | 148 mm² |

| Transistor Density | Not disclosed | 12.6M / mm² |

| Base Clock | 350 MHz | 795 MHz |

| Boost Clock | 1200 MHz | 861 MHz |

| Memory Clock | System Shared | 1253 MHz (5 Gbps effective) |

| Memory Size | System Shared | 2 GB GDDR5 |

| Memory Bus Width | System Shared | 64 bit |

| Memory Bandwidth | System Dependent | 40.10 GB/s |

| Shading Units | 192 | 512 |

| TMUs | 24 | 32 |

| ROPs | 3 | 8 |

| Pixel Rate | 3.600 GPixel/s | 6.888 GPixel/s |

| Texture Rate | 28.80 GTexel/s | 27.55 GTexel/s |

| FP32 | 460.8 GFLOPS | 881.7 GFLOPS |

| FP16 | 921.6 GFLOPS (2:1) | Not listed |

| TDP | 15 W | 23 W |

| Slot Width | IGP (Integrated Graphics) | MXM Module |

| Power Connectors | None | None |

| Interface | Ring Bus | PCIe 3.0 x8 |

| Release | End-of-life | End-of-life |

Head-to-Head Benchmarks

The database contains two relevant head-to-head results. The first, GeForce OpenCL, has the Intel part scoring 5111 versus 4939 for the NVIDIA part, a 3.5 percent advantage. The other test, Vulkan, is more decisive: the Intel UHD P630 scores 562, while the GeForce 940MX scores 4749, making the Intel part 18.5 percent faster.

The overall wins column is 2-0 for the Intel part. The average score across both workloads is 5370 for the Intel part and 4844 for the NVIDIA part. This averages a 10.9 percent overall lead for the integrated graphics solution, and it is consistent: the 950MX does not close the gap in any test.

The biggest single margin is in the Vulkan workload. The database shows Intel leading by 18.5 percent. That is a substantial gap for a low-end mobile part. The OpenCL gap is smaller but still solid. The NVIDIA part does not offer any counterpoint; it is consistently slower in these measured tasks.

The NVIDIA part does have a hardware advantage in raw throughput, with more shading units, more texture mapping units, and more render output units. These are 512, 32, and 8 versus 192, 24, and 3 for the Intel part. The memory bandwidth also is heavily in favor of the NVIDIA: 40.10 GB/s versus the system-shared memory of Intel. Yet in real-world performance as measured by the database, the integrated Intel part is consistently the better performer. This suggests that the Intel drivers and architecture are better suited to these workloads.

The database shows the NVIDIA 940MX at the 31st percentile of all GPUs, and the Intel P630 at the 31st percentile. Both parts are in the lower half of the performance curve. Intel's average score across both tests is 5370, while NVIDIA's average is 4844. The delta is in favor of Intel in both benchmark instances.

For a user who cares about compute workloads, the data is clear. The Intel UHD P630 is the pick of these two. The discrete 940MX has a bandwidth and raw throughput advantage, but the recorded scores show the Intel part converting its resources to performance more effectively.