AMD Radeon HD 8790M vs Intel UHD Graphics P630 Comparison

# AMD Radeon HD 8790M vs Intel UHD Graphics P630

The benchmark data splits this matchup nearly down the middle: the AMD Radeon HD 8790M wins the Vulkan test by a decisive 13.1% margin, while the Intel UHD Graphics P630 takes the OpenCL test by a slim 1.8%. The AMD part posts a higher average benchmark score of 5691 versus 5370 for Intel, but both cards sit in the bottom third of all GPUs — the AMD at the 33rd percentile and the Intel at the 31st. For gaming workloads that lean on Vulkan, the AMD Radeon HD 8790M is the clear choice; for general compute tasks using OpenCL, the Intel UHD Graphics P630 edges ahead, though the difference is within noise territory.

The Verdict

The AMD Radeon HD 8790M is the stronger overall performer when averaging the two benchmark tests, with a 5.6% higher average score (5691 vs 5370). Its Vulkan advantage is substantial — 6365 versus 5628, a 13.1% lead — which matters for modern game engines and compute APIs. The Intel UHD Graphics P630 counters with a marginal OpenCL win (5111 vs 5017), but that 1.8% gap is small enough to be considered essentially a tie.

Pick the AMD Radeon HD 8790M if your workload prioritizes Vulkan performance, as the data shows a clear and meaningful advantage in that API. The AMD part also delivers higher raw compute throughput at 691.2 GFLOPS FP32 versus 460.8 GFLOPS for Intel, and it has double the shading units (384 vs 192). Pick the Intel UHD Graphics P630 if you need a low-power integrated solution — its 15 W TDP compares favorably to the AMD's MXM module form factor, and its OpenCL score is marginally better. However, the Intel part's 3 ROPs are a severe limitation for pixel-heavy rendering, and its memory bandwidth is listed as "System Dependent," meaning performance scales with the host system's RAM configuration.

For users who already have a laptop with an MXM slot, the AMD Radeon HD 8790M offers the better benchmark pedigree. For those building or buying a system with an integrated GPU, the Intel UHD Graphics P630 provides adequate compute performance but trails in Vulkan workloads by a significant margin.

Architecture Differences

The AMD Radeon HD 8790M is built on TSMC's 28 nm process using the GCN 1.0 architecture with the Mars chip. It packs 950 million transistors into a 77 mm² die, yielding a transistor density of 12.3 million per square millimeter. The Intel UHD Graphics P630 uses Intel's 14 nm+++ process with the Generation 9.5 architecture and the Comet Lake GT2 chip. Intel does not disclose transistor count or die size for this part, but the process node advantage (14 nm+++ vs 28 nm) is notable.

Core configuration differs substantially. The AMD part features 384 shading units, 24 texture mapping units, and 8 ROPs. The Intel part has exactly half the shading units at 192, but matches the TMU count at 24, and drops to just 3 ROPs. This ROP disparity — 8 vs 3 — explains why the AMD part achieves a pixel rate of 7.200 GPixel/s versus 3.600 GPixel/s for Intel, exactly double.

Clock speeds tell another story. The AMD Radeon HD 8790M runs at a base clock of 850 MHz with a boost of 900 MHz, while the Intel UHD Graphics P630 runs at a much lower base of 350 MHz but boosts to a higher 1200 MHz. The texture rate reflects this: Intel achieves 28.80 GTexel/s versus 21.60 GTexel/s for AMD, a 33% advantage despite fewer shading units, because the higher boost clock compensates.

Memory architecture is fundamentally different. The AMD part uses 2 GB of dedicated GDDR5 memory on a 128-bit bus, delivering 64.00 GB/s of bandwidth. The Intel part uses system-shared memory with a system-dependent bus width and bandwidth — this is a critical architectural difference because dedicated VRAM provides consistent performance, while shared memory performance varies with the host system's memory configuration and speed.

API support also diverges. The AMD part supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The Intel part supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The newer Vulkan 1.3 support on Intel is a feature advantage, though the benchmark scores show AMD winning the Vulkan test anyway.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon HD 8790M has an average benchmark score of 5691, which is 5.6% higher than the Intel UHD Graphics P630's 5370.

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

A: The AMD Radeon HD 8790M scores 6365 in the Geekbench Vulkan test, while the Intel UHD Graphics P630 scores 5628 — a 13.1% advantage for AMD.

Q: Which GPU has more shading units?

A: The AMD Radeon HD 8790M has 384 shading units, exactly double the 192 shading units in the Intel UHD Graphics P630.

Q: What is the memory configuration difference?

A: The AMD Radeon HD 8790M uses 2 GB of dedicated GDDR5 memory with a 128-bit bus and 64.00 GB/s bandwidth. The Intel UHD Graphics P630 uses system-shared memory with system-dependent bandwidth.

Q: Does the Intel GPU have a TDP advantage?

A: Yes, the Intel UHD Graphics P630 has a 15 W TDP, whereas the AMD Radeon HD 8790M has no listed TDP but comes as an MXM module, which typically requires more power.

Q: Which GPU has better OpenCL performance?

A: The Intel UHD Graphics P630 scores 5111 in Geekbench OpenCL, slightly ahead of the AMD Radeon HD 8790M's 5017, a 1.8% difference.

Specification Differences

The two GPUs differ on nearly every major specification. The AMD Radeon HD 8790M uses a 28 nm TSMC process, while the Intel UHD Graphics P630 uses Intel's 14 nm+++ process. The AMD chip is "Mars" based on GCN 1.0 architecture, while Intel's is "Comet Lake GT2" based on Generation 9.5. Transistor count is 950 million for AMD, with Intel not disclosing this figure. Die size is 77 mm² for AMD, again undisclosed for Intel.

Clock speeds differ: AMD runs at 850 MHz base and 900 MHz boost, while Intel runs at 350 MHz base and 1200 MHz boost. Memory configuration is completely different: AMD uses 2 GB GDDR5 with a 128-bit bus and 64.00 GB/s bandwidth; Intel uses system-shared memory with system-dependent bandwidth. Shading units are 384 for AMD versus 192 for Intel. TMUs are equal at 24 each, but ROPs differ drastically: 8 for AMD versus 3 for Intel. Pixel rate is 7.200 GPixel/s for AMD versus 3.600 GPixel/s for Intel. Texture rate is 21.60 GTexel/s for AMD versus 28.80 GTexel/s for Intel. FP32 compute is 691.2 GFLOPS for AMD versus 460.8 GFLOPS for Intel. Intel additionally lists FP16 at 921.6 GFLOPS (2:1), while AMD has no FP16 listing.

Form factor and connectivity also differ: AMD uses an MXM Module with MXM-A (3.0) bus interface and no power connectors; Intel is an IGP with a Ring Bus interface. Display outputs are "Portable Device Dependent" for AMD and "Motherboard Dependent" for Intel. API support differs in DirectX version (12 (11_1) for AMD vs 12 (12_1) for Intel) and Vulkan version (1.2.170 for AMD vs 1.3 for Intel). The AMD part was released in 2013, while the Intel part came in 2020.

Head-to-Head Benchmarks

The Geekbench OpenCL test shows the Intel UHD Graphics P630 winning with a score of 5111 against the AMD Radeon HD 8790M's 5017, a delta of -1.8% from AMD's perspective. This is a narrow margin — the kind of difference that could be attributed to driver optimization or system memory configuration. The Intel part's higher boost clock of 1200 MHz and higher texture rate of 28.80 GTexel/s likely contribute to this win, as OpenCL workloads often scale with texture throughput.

The Geekbench Vulkan test is a different story entirely. The AMD Radeon HD 8790M scores 6365, versus 5628 for the Intel UHD Graphics P630, a 13.1% advantage. This is a substantial, decisive win. The AMD part's dedicated GDDR5 memory with 64.00 GB/s bandwidth versus Intel's system-shared memory likely plays a significant role here — Vulkan workloads are sensitive to memory latency and bandwidth, and dedicated VRAM provides a consistent advantage. Additionally, AMD's 384 shading units provide more parallel compute capacity for Vulkan's explicit programming model.

The wins split evenly at one each, but the magnitude matters. The AMD Vulkan win is 737 points, while the Intel OpenCL win is only 94 points. In percentage terms, AMD's margin is 13.1% versus Intel's 1.8%. The average benchmark scores reflect this asymmetry: AMD's average of 5691 is 6.0% higher than Intel's 5370. Looking at nearest rivals, the AMD Radeon HD 8790M sits between the Intel Iris Pro Graphics P6300 (5712, -0.4%) and the NVIDIA Quadro M500M (5604, +1.6%). The Intel UHD Graphics P630 sits between the AMD Radeon R7 M445 (5358, +0.2%) and the AMD Radeon R7 M365X (5416, -0.8%). Both GPUs occupy similar performance tiers, but AMD's is slightly higher.

Where Each One Wins

The AMD Radeon HD 8790M wins decisively in Vulkan-based workloads. The 13.1% advantage in the Vulkan benchmark makes it the better choice for modern gaming APIs and compute applications that use Vulkan. Its dedicated 2 GB GDDR5 memory with 64.00 GB/s bandwidth provides predictable performance that doesn't depend on system RAM, and its 384 shading units deliver 691.2 GFLOPS of FP32 compute — 50% more than the Intel part. The AMD GPU also doubles the pixel rate (7.200 vs 3.600 GPixel/s), making it better suited for resolution-heavy rendering tasks that stress ROP throughput.

The Intel UHD Graphics P630 wins in OpenCL workloads, albeit marginally. Its 5111 OpenCL score edges out AMD's 5017, and its 28.80 GTexel/s texture rate is 33% higher than AMD's 21.60 GTexel/s. The Intel part also has a significant power advantage with its 15 W TDP, making it suitable for thin-and-light systems where thermal and power constraints are paramount. Its newer Vulkan 1.3 API support and DirectX 12 (12_1) are also forward-looking features, though the benchmark data shows AMD still wins in Vulkan despite older API support.

For users weighing these two, the decision hinges on workload priority. Vulkan-heavy workloads strongly favor the AMD Radeon HD 8790M, while OpenCL compute slightly favors the Intel UHD Graphics P630. The AMD part's higher average score and bigger win margin give it the overall performance crown, but the Intel part's efficiency and system integration make it a viable choice for low-power builds. The data shows a clear tradeoff: raw performance and memory bandwidth versus power efficiency and integration.