Intel UHD Graphics P630 Mobile

Intel UHD Graphics P630 Mobile is an integrated graphics processor from Intel, built on the Coffee Lake GT2 chip and using the Generation 9.5 architecture at Intel's 14 nm+++ process. It belongs to the HD Graphics-WM (Coffee Lake) generation and was released on 2018-05-23; the database marks its production status as end-of-life. The slot width is IGP, the bus interface is Ring Bus, and the TDP is 15 W. In this database record, the benchmarks array is empty, the avgBenchmarkScore is 0, and the percentileVsAllGpus is 50. That percentile places the part at the midpoint of all GPUs in the database, but the zero average score means the percentile is not backed by any stored benchmark result.

Benchmark Performance

The absence of benchmark entries is the first finding. There are no scores for this GPU and no nearestRivals entries, so there are no deltaPct values to report against a named product. The only performance-related data are the clock rates and the derived fixed-function throughputs. The base clock is 350 MHz and the boost clock is 1200 MHz; no game clock is listed. The memory clock is System Shared, which indicates that memory timing is not fixed on the GPU itself.

On the compute side, the GPU has 192 shading units, 24 texture mapping units, and 3 render output units. The FP32 throughput is 460.8 GFLOPS, while FP16 throughput is 921.6 GFLOPS, with the fact pack explicitly recording the FP16 ratio as 2:1. This places the FP16 figure at the stated 2:1 ratio relative to FP32, but the data does not list tensor or ray tracing cores, so the FP16 rate should not be interpreted as evidence of specialist AI or RT hardware.

Fixed-function fill rates are also present. The pixel rate is 3.600 GPixel/s and the texture rate is 28.80 GTexel/s. The 3 ROPs directly limit the pixel rate, while the 24 TMUs drive the higher texture throughput. This arrangement implies that fill-heavy workloads will be bounded by the pixel output stage rather than by texturing. The wide difference between base and boost clocks suggests that the part can raise throughput when power and thermals permit, but no sustained clock figure is recorded.

Because no benchmark scores exist, the percentileVsAllGpus value of 50 cannot be verified through measured results. In a database that normally uses averaged scores, a value of 0 is the missing-data marker. Consequently, any statement about this GPU being faster or slower than another product would be unsupported. The data shows a specification profile of a low-throughput integrated GPU, but the relative positioning via nearestRivals is simply absent.

Who Should Consider It

The data suggests this GPU is intended for integrated mobile systems. The slot width is IGP, the TDP is 15 W, and all memory is System Shared, so there is no discrete card installation path and no independent video memory. The intended workloads are therefore those that fit within the fill-rate and compute ceilings above. The 3.600 GPixel/s pixel rate and 3 ROPs are the limiting resources for output resolution and framebuffer fill. Lower resolution and reduced settings are the natural match for those ceilings, although no recorded benchmark verifies a specific settings configuration.

The API support is broad: DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3 are all listed. That means software written to these APIs is theoretically compatible, but compatibility does not imply the performance headroom for high settings. The lack of a game clock also means sustained gaming behavior is not characterized. Users who need basic display output or low-fill-rate 3D workloads may find the feature set sufficient. Users who need high-resolution rendering should note that the pixel rate is the first hard limit. Since Display Outputs are Portable Device Dependent, the actual ports and supported displays are dictated by the host portable device rather than by a fixed board design.

Memory Subsystem

The memory subsystem is entirely shared. The size field is System Shared, the type is System Shared, and the bus width is System Shared. There is no dedicated VRAM capacity, no dedicated memory type, and no fixed bus width. The bandwidth field is System Dependent, so the actual memory bandwidth available to the GPU is set by the host platform's memory system rather than by a specification on this page. The memory clock is also System Shared, meaning there is no separate memory clock value. For high resolutions, this is important because the framebuffer and texture data must be stored in shared memory, competing with other system uses of that memory. Without a fixed bandwidth figure, the data cannot place a lower bound on memory performance. The only fixed throughput values in this record are the pixel rate and texture rate, which are tied to the GPU core clocks, not to memory speed. As a result, the memory subsystem is best understood as a variable-performance element: the same GPU can behave differently depending on the host platform's memory configuration, but the fact pack does not quantify any platform-specific bandwidth.

How It Compares

The nearestRivals array is empty. There are no rival names, no rival scores, and no deltaPct values. This is not a case of a rival being slightly ahead or behind; it is a case of no comparison data being present at all. The only global positioning number is percentileVsAllGpus: 50. On its own, this would suggest a median placement among all GPUs in the database. However, because avgBenchmarkScore is 0 and the benchmark list is empty, the percentile is an unverified value. The database treats the GPU as one without benchmark results, so any paragraph comparing this GPU to a specific rival would require data the fact pack does not provide. The correct analytical statement is that relative performance cannot be established from this record. No rival deltaPct can be cited, and no claim of leadership or deficit is warranted.

Power and Cooling

The TDP is 15 W, which is the only power figure in the record. The suggested PSU field is null, and the power connectors field is null, meaning no external power connector requirement is documented. Because the slot width is IGP, the GPU is not a separate expansion card; it is integrated into a platform. In that form factor, there is no discrete PSU sizing or modular connector to specify. The bus interface is Ring Bus, an internal interconnect rather than a slot-style connection. Cooling is therefore part of the host system's thermal design, not an add-on cooler attachment. The 15 W TDP represents the thermal envelope that the portable device must dissipate. Display outputs are Portable Device Dependent, so physical monitor connections depend on the surrounding device. The data shows a part designed for compact, low-dissipation integration rather than for a high-power add-in environment.