Intel HD Graphics 630 vs NVIDIA GeForce 920M Comparison

Head-to-Head Benchmarks

The recorded data shows a split decision between these two end-of-life mobile graphics solutions. Each claims one victory in the two available benchmark comparisons, and the margins tell a clear story about their respective strengths.

In Geekbench OpenCL, the NVIDIA GeForce 920M takes the win with a score of 3725 against the Intel HD Graphics 630’s 3587. That is a 3.7 percent advantage for the discrete GPU. The margin is modest, but it is consistent with the 920M’s higher raw compute specifications. The OpenCL test tends to exercise sustained shader throughput, and the 920M’s larger shading engine gives it a measurable edge.

The Vulkan result flips the narrative decisively. Intel’s integrated part scores 3540, while the GeForce 920M manages only 2849. That is a 24.3 percent gap in favor of the HD Graphics 630. This is a substantial margin, not a rounding error. The Intel part’s more modern API feature set, including Vulkan 1.3 support compared to the 920M’s Vulkan 1.2.175, appears to translate directly into better performance on this workload.

Looking at the broader database averages, the Intel HD Graphics 630 carries an average benchmark score of 4075 across all recorded tests, placing it in the 24th percentile of all GPUs. The GeForce 920M’s average sits at 3287, which lands in the 20th percentile. The difference in averages, roughly 24 percent, mirrors the Vulkan result more closely than the OpenCL result. This suggests the Intel part’s overall standing in the database is stronger than the 920M’s, even though the 920M wins the one compute test where both were evaluated head-to-head.

The nearest rival data for each part reinforces this ordering. The HD Graphics 630 sits within 1.8 percent of the NVIDIA Quadro K2100M (score 4151) and within 1.5 percent of the AMD FirePro M4150 (score 4013). It is effectively tied with the AMD Radeon RX 9060 XT 8 GB, which scores 4093, a delta of just 0.4 percent. The 920M, by contrast, is bracketed by the NVIDIA GeForce GT 730M (3316, 0.9 percent faster) and the Intel HD Graphics 530 (3332, 1.4 percent faster). The 920M’s closest rival cluster is meaningfully lower than the HD Graphics 630’s cluster, which confirms that the Intel integrated solution holds the higher performance tier in this comparison.

FAQ

Q: Which GPU wins in OpenCL performance?

A: The NVIDIA GeForce 920M records a score of 3725 in Geekbench OpenCL, compared to 3587 for the Intel HD Graphics 630. That is a 3.7 percent advantage for the NVIDIA part.

Q: Which GPU wins in Vulkan performance?

A: The Intel HD Graphics 630 scores 3540 in Geekbench Vulkan, versus 2849 for the GeForce 920M. The Intel part leads by 24.3 percent, a decisive margin.

Q: How do their overall average benchmark scores compare?

A: The Intel HD Graphics 630 has an average benchmark score of 4075 across all recorded tests, placing it in the 24th percentile of all GPUs. The GeForce 920M averages 3287, placing it in the 20th percentile.

Q: What are the closest rivals for each GPU according to the database?

A: For the Intel HD Graphics 630, the nearest rivals are the AMD Radeon RX 9060 XT 8 GB (4093, 0.4 percent behind), the NVIDIA GeForce GT 755M (4033, 1 percent ahead), the AMD FirePro M4150 (4013, 1.5 percent ahead), and the NVIDIA Quadro K2100M (4151, 1.8 percent behind). For the GeForce 920M, the nearest rivals are the NVIDIA GeForce GT 730M (3316, 0.9 percent ahead), the Intel HD Graphics 530 (3332, 1.4 percent ahead), the NVIDIA GeForce GT 640 (3210, 2.4 percent behind), and the Intel HD Graphics P4600 (3389, 3 percent ahead).

Q: Does the GeForce 920M have a higher pixel fill rate than the Intel part?

A: Yes. The GeForce 920M achieves 7.632 GPixel/s, while the Intel HD Graphics 630 manages 3.000 GPixel/s. The NVIDIA part has a 2.5x advantage in pixel throughput.

Q: Which GPU supports a newer Vulkan version?

A: The Intel HD Graphics 630 supports Vulkan 1.3, while the GeForce 920M supports Vulkan 1.2.175. Both parts support DirectX 12 and OpenGL 4.6, though the Intel part reaches DirectX 12 (12_1) while the NVIDIA part tops out at DirectX 12 (11_0).

Architecture Differences

The two GPUs come from fundamentally different design philosophies. The Intel HD Graphics 630 is built on the Generation 9.5 architecture, specifically the Kaby Lake GT2 chip, fabricated on Intel’s 14 nm++ process. The GeForce 920M uses NVIDIA’s Kepler 2.0 architecture, implemented in the GK208B chip, produced on TSMC’s 28 nm process. The process node gap is significant: 14 nm++ versus 28 nm. This gives Intel a density and efficiency advantage at the transistor level, even though the Intel part is an integrated solution sharing system resources.

The GeForce 920M is a discrete GPU with its own dedicated memory. It packs 1,020 million transistors into an 87 mm² die, yielding a transistor density of 11.7 million transistors per square millimeter. The Intel part’s transistor count and die size are not recorded in the database, but its integrated nature means it shares the CPU package and system memory rather than carrying its own VRAM.

The memory architectures diverge completely. The Intel HD Graphics 630 uses system shared memory, with a bus width and bandwidth that are system dependent. The GeForce 920M has 2 GB of dedicated DDR3 memory on a 64-bit bus, delivering 14.40 GB/s of bandwidth. This dedicated memory gives the NVIDIA part a fixed, predictable memory pipeline, while the Intel part’s performance scales with the host system’s memory configuration.

The compute resources also differ substantially. The GeForce 920M has 384 shading units, 32 texture mapping units, and 8 raster output pipelines. The Intel HD Graphics 630 has 192 shading units, 24 TMUs, and only 3 ROPs. The NVIDIA part has double the shader count and nearly double the TMU count, plus a much higher ROP count. This explains its pixel rate advantage of 7.632 GPixel/s versus 3.000 GPixel/s, and its texture rate of 30.53 GTexel/s versus 24.00 GTexel/s.

However, the Intel part counters with a higher boost clock architecture. The HD Graphics 630 runs at a base clock of 350 MHz and boosts to 1000 MHz. The GeForce 920M has a fixed clock of 954 MHz for both base and boost. The Intel part’s boost clock is slightly higher, but the NVIDIA part’s clock is constant and does not depend on thermal headroom.

The API support reveals another architectural divide. The Intel part supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The NVIDIA part supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The Intel part’s newer feature level for DirectX and Vulkan likely contributes to its strong Vulkan benchmark showing.

Specification Differences

The two parts differ across nearly every measurable specification. The process nodes are distinct: Intel uses 14 nm++ while NVIDIA uses 28 nm. The transistor counts are not directly comparable because the Intel figure is not recorded, but the GeForce 920M lists 1,020 million transistors on an 87 mm² die. The Intel part’s die size and transistor density are absent from the database.

Clock speeds differ significantly. The Intel HD Graphics 630 has a base clock of 350 MHz and a boost clock of 1000 MHz. The GeForce 920M runs at a flat 954 MHz with no boost variation. The memory clocks are not directly comparable: the Intel part uses system shared memory, while the 920M uses DDR3 at 900 MHz, or 1800 Mbps effective.

Memory configuration is a major differentiator. The Intel part has system shared memory with system dependent bandwidth. The GeForce 920M has 2 GB of dedicated DDR3 on a 64-bit bus with 14.40 GB/s bandwidth. The shading units, TMUs, and ROPs all favor NVIDIA: 384 versus 192, 32 versus 24, and 8 versus 3, respectively.

Pixel rate and texture rate both favor NVIDIA. The 920M delivers 7.632 GPixel/s and 30.53 GTexel/s, while the Intel part manages 3.000 GPixel/s and 24.00 GTexel/s. FP32 performance also favors NVIDIA: 732.7 GFLOPS versus 384.0 GFLOPS. The Intel part has FP16 support at 768.0 GFLOPS with a 2:1 ratio, while the NVIDIA part has no recorded FP16 figure.

Thermal design power differs by more than double. The Intel HD Graphics 630 is rated at 15 W, while the GeForce 920M is rated at 33 W. Both are integrated into portable devices, with the Intel part using a Ring Bus interface and the NVIDIA part using PCIe 3.0 x8. The Intel part has no power connectors, and the NVIDIA part also lists none. Display outputs are motherboard dependent for Intel and portable device dependent for NVIDIA.

Release timing also differs. The Intel HD Graphics 630 was released on 2016-08-29, while the GeForce 920M launched on 2015-03-12. Both are marked as end-of-life in the database. The GeForce 920M has recorded predecessors and successors: GeForce 800M and GeForce 10 Mobile, respectively. The Intel part has no recorded lineage.

The Verdict

The data supports a clear split recommendation based on workload priorities. For users who rely on compute-heavy OpenCL workloads, the NVIDIA GeForce 920M holds a measurable edge of 3.7 percent over the Intel HD Graphics 630. Its higher pixel rate, texture rate, and FP32 throughput make it the stronger choice for tasks that stress raw shader output.

For users who prioritize modern API performance, particularly Vulkan, the Intel HD Graphics 630 is the definitive winner. Its 24.3 percent lead in the Vulkan benchmark is not a marginal difference; it represents a generation-level advantage in API efficiency. The Intel part’s support for Vulkan 1.3 and DirectX 12 (12_1) gives it a forward-looking feature set that the Kepler-based 920M cannot match.

The overall database averages reinforce this conclusion. The Intel part’s average score of 4075 places it in the 24th percentile, while the 920M’s average of 3287 places it in the 20th percentile. The Intel part is also bracketed by stronger rivals, sitting within 1.8 percent of the Quadro K2100M, while the 920M’s closest rivals are all clustered below the 3400 mark.

The power envelope is another factor. The Intel HD Graphics 630 draws 15 W, while the GeForce 920M draws 33 W. For portable devices where thermal and battery constraints matter, the Intel part offers competitive performance at less than half the power draw.

The verdict is straightforward. The GeForce 920M wins on raw compute metrics and dedicated memory bandwidth, making it the pick for legacy OpenCL applications or scenarios where consistent memory performance is essential. The Intel HD Graphics 630 wins on modern API support, overall benchmark average, and power efficiency, making it the better choice for Vulkan-based workloads and longer battery life. Neither part is a performance leader in the current landscape, but the data shows the Intel integrated solution holds the stronger overall position in the database.