Intel UHD Graphics P630 vs NVIDIA GeForce GTX 980M Comparison

FAQ

Q: How does the Intel UHD Graphics P630 compare to the NVIDIA GeForce GTX 980M in raw compute performance?

A: The GTX 980M is dramatically ahead. In Geekbench OpenCL, the NVIDIA part scores 23,832 versus Intel's 5,111, a 78.6% deficit for the P630. The Vulkan test shows a similar gap: 17,703 versus 5,628, a 68.2% shortfall.

Q: Which GPU has the higher average benchmark score?

A: The two are nearly identical. The P630 averages 5,370 across its benchmark results, while the GTX 980M averages 5,308. The delta is less than 0.2%, placing both GPUs at the 31st percentile of all GPUs.

Q: What is the closest rival to each card according to the data?

A: For the Intel P630, the AMD Radeon R7 M445 is nearest, with an average score of 5,358 versus Intel's 5,370 — a 0.2% difference. For the GTX 980M, the NVIDIA GeForce 930A is closest, scoring 5,317 versus 5,308, a 0.2% gap in the other direction.

Q: Do both GPUs support the same graphics APIs?

A: Both support DirectX 12 (12_1) and OpenGL 4.6. They differ on Vulkan: the Intel P630 supports Vulkan 1.3, while the GTX 980M supports Vulkan 1.4.

Q: What are the memory configurations of these two GPUs?

A: The Intel P630 uses system-shared memory with a system-dependent bandwidth. The GTX 980M has 8 GB of dedicated GDDR5 memory on a 256-bit bus, delivering 160.4 GB/s of bandwidth.

Q: Which GPU was released later?

A: The Intel UHD Graphics P630 launched on May 12, 2020, while the NVIDIA GeForce GTX 980M launched on October 6, 2014. Both are now end-of-life products.

Architecture Differences

The two GPUs come from opposite ends of the design spectrum. The Intel UHD Graphics P630 is built on Intel's Generation 9.5 architecture, specifically the Comet Lake GT2 chip, fabricated on Intel's 14 nm+++ process. It is an integrated graphics processor (IGP) that taps into the host system's memory and relies on a Ring Bus interface. The GTX 980M, in contrast, uses NVIDIA's Maxwell 2.0 architecture with the GM204 chip, manufactured on TSMC's 28 nm process. It is a discrete mobile module with an MXM-B (3.0) interface.

The transistor counts and die sizes reflect the different roles. The GTX 980M packs 5,200 million transistors onto a 398 mm² die, yielding a transistor density of 13.1 million per square millimeter. The P630's transistor count and die size are not specified in the data, but its integrated nature and 14 nm+++ process suggest a much smaller, more power-efficient design.

Clock speeds also diverge sharply. The Intel part runs at a base clock of 350 MHz, boosting to 1200 MHz. The NVIDIA part starts at 1038 MHz and boosts to 1127 MHz. Memory clocks tell a similar story: the P630 uses system shared memory with no dedicated clock, while the GTX 980M runs its GDDR5 at 1253 MHz (5 Gbps effective).

The execution resources are heavily skewed toward the discrete card. The GTX 980M has 1,536 shading units, 96 texture mapping units, and 64 raster output units. The P630 has just 192 shading units, 24 TMUs, and a mere 3 ROPs. This explains the massive throughput differences: the GTX 980M delivers 72.13 GPixel/s pixel fill rate and 108.2 GTexel/s texture rate, versus 3.600 GPixel/s and 28.80 GTexel/s for the Intel part. Floating-point performance is 3.462 TFLOPS (FP32) for NVIDIA versus 460.8 GFLOPS for Intel.

Power and physical format differ as well. The P630 is rated at 15 W TDP and is an IGP, while the GTX 980M's TDP is not listed, but it comes as an MXM Module with no power connectors. Display outputs are motherboard-dependent for Intel and portable-device-dependent for NVIDIA.

Head-to-Head Benchmarks

The head-to-head data contains only two benchmark comparisons, both of which are decisive wins for the NVIDIA GeForce GTX 980M. There are no benchmark tests where the Intel UHD Graphics P630 comes out ahead.

In Geekbench OpenCL, the GTX 980M scores 23,832 against the P630's 5,111. This represents a 78.6% delta in favor of NVIDIA. The gap is so large that the Intel GPU's score is roughly one-fifth of the NVIDIA part's output. This test stresses general-purpose compute workloads, where the GTX 980M's 1,536 shading units and 3.462 TFLOPS of FP32 throughput dominate the P630's 192 shading units and 460.8 GFLOPS.

The Vulkan test narrows the gap slightly but still favors NVIDIA heavily. The GTX 980M scores 17,703, while the P630 scores 5,628 — a 68.2% delta. Vulkan is a lower-overhead API, which may help the integrated Intel GPU relatively, but the raw hardware advantage of the GTX 980M remains overwhelming. The NVIDIA card's Vulkan 1.4 support versus Intel's Vulkan 1.3 may also contribute to the performance difference.

The benchmark summary shows the GTX 980M winning 2 out of 2 head-to-head tests, with 0 wins for the Intel part. However, the average benchmark scores for the two GPUs are nearly identical — 5,370 for Intel and 5,308 for NVIDIA — because the Intel part's benchmark set is limited to two Geekbench tests, while the NVIDIA part is measured across ten tests including DirectX 9, 10, 11, 12, and compute workloads. The GTX 980M's Passmark scores range from 31 (DirectX 12) to 125 (DirectX 9), with a G3D score of 7,338 and a GPU compute score of 2,816. These additional tests pull its average down significantly.

Specification Differences

The two GPUs differ across nearly every specification category:

• Process Node: Intel uses 14 nm+++, NVIDIA uses 28 nm.
• Transistors: GTX 980M has 5,200 million; P630 has no listed value.
• Die Size: GTX 980M is 398 mm²; P630 has no listed value.
• Transistor Density: GTX 980M is 13.1M / mm²; P630 has no listed value.
• Base Clock: P630 is 350 MHz; GTX 980M is 1038 MHz.
• Boost Clock: P630 is 1200 MHz; GTX 980M is 1127 MHz.
• Memory Size: P630 is system shared; GTX 980M is 8 GB GDDR5.
• Memory Bus Width: P630 is system shared; GTX 980M is 256 bit.
• Memory Bandwidth: P630 is system dependent; GTX 980M is 160.4 GB/s.
• Shading Units: P630 has 192; GTX 980M has 1,536.
• TMUs: P630 has 24; GTX 980M has 96.
• ROPs: P630 has 3; GTX 980M has 64.
• Pixel Rate: P630 is 3.600 GPixel/s; GTX 980M is 72.13 GPixel/s.
• Texture Rate: P630 is 28.80 GTexel/s; GTX 980M is 108.2 GTexel/s.
• FP32 Performance: P630 is 460.8 GFLOPS; GTX 980M is 3.462 TFLOPS.
• FP16 Performance: P630 is 921.6 GFLOPS (2:1); GTX 980M has no listed value.
• TDP: P630 is 15 W; GTX 980M has no listed value.
• Slot Width: P630 is IGP; GTX 980M is MXM Module.
• Bus Interface: P630 is Ring Bus; GTX 980M is MXM-B (3.0).
• Vulkan Support: P630 is 1.3; GTX 980M is 1.4.
• Release Date: P630 is 2020-05-12; GTX 980M is 2014-10-06.
• Predecessor/Successor: GTX 980M has predecessor GeForce 800M and successor GeForce 10 Mobile; P630 has neither listed.

Fields that are identical include DirectX 12 (12_1) support, OpenGL 4.6 support, foundry status (both end-of-life), and the fact that neither has a launch MSRP listed.

The Verdict

The benchmark data is unambiguous: the NVIDIA GeForce GTX 980M is the far more capable GPU in every measurable head-to-head test. Its OpenCL score of 23,832 is 4.7 times higher than the Intel P630's 5,111, and its Vulkan score of 17,703 is 3.1 times higher than the P630's 5,628. If the sole criterion is raw graphics or compute performance, the GTX 980M is the clear choice.

However, the story is more nuanced when considering the full context. Both GPUs sit at the 31st percentile of all GPUs, and their average benchmark scores are nearly identical (5,370 versus 5,308). The Intel P630, despite being an integrated part with 192 shading units and a 15 W TDP, holds its own against the discrete GTX 980M when averaged across the NVIDIA card's broader test suite. This is because the GTX 980M's Passmark DirectX scores are surprisingly low — 35 for DirectX 10, 57 for DirectX 11, 31 for DirectX 12, and 125 for DirectX 9 — suggesting the card struggles with legacy or specific driver-optimized workloads.

The GTX 980M is the right pick for users who need maximum compute throughput in OpenCL or Vulkan applications, or who require dedicated VRAM (8 GB GDDR5) and high memory bandwidth (160.4 GB/s). The Intel P630 is the right pick for systems where power efficiency (15 W TDP), integrated simplicity, and modern Vulkan 1.3 support are priorities, and where the workload does not demand the GTX 980M's massive shading unit count.

Where Each One Wins

NVIDIA GeForce GTX 980M wins on:

• Compute-heavy workloads: The OpenCL score of 23,832 versus 5,111 represents a 78.6% advantage, making it the superior choice for GPGPU tasks.
• Vulkan applications: The 17,703 versus 5,628 Vulkan score (68.2% delta) shows a decisive edge in modern, low-overhead graphics APIs.
• Memory bandwidth: With 160.4 GB/s of dedicated GDDR5 bandwidth versus system-shared memory, the GTX 980M handles large textures and datasets far more efficiently.
• Pixel and texture throughput: The 72.13 GPixel/s and 108.2 GTexel/s rates are 20x and 3.8x higher than the P630's respective rates, making it better for high-resolution rendering.

Intel UHD Graphics P630 wins on:

• Power consumption: At 15 W TDP, the P630 is a fraction of the power draw of the MXM-format GTX 980M, making it suitable for thin-and-light systems.
• Average benchmark parity: The P630's average score of 5,370 slightly edges the GTX 980M's 5,308, and its nearest rivals are all low-end discrete or integrated parts, indicating consistent performance across its limited test set.
• FP16 throughput: The P630 delivers 921.6 GFLOPS of FP16 performance, a capability not listed for the GTX 980M, which may matter for specific AI or media workloads.
• Vulkan API version: The P630 supports Vulkan 1.3, while the GTX 980M supports 1.4 — a minor but notable difference for compatibility with older software stacks.
• Integrated convenience: With no power connectors and a motherboard-dependent display output, the P630 requires no additional cooling or power delivery infrastructure, unlike the MXM module format of the GTX 980M.

In short, the GTX 980M is the performance king, but the P630's efficiency and system integration make it a viable option for workloads that do not demand the NVIDIA card's raw throughput. The data shows two GPUs at the same overall percentile, but with vastly different strengths.