AMD Radeon HD 8870M vs Intel Iris Pro Graphics P580 Comparison

Where Each One Wins

The benchmark data splits cleanly between these two mobile graphics processors, but the split is not what most would expect. The Intel Iris Pro Graphics P580 wins the only direct head-to-head comparison recorded in the database, taking the Geekbench OpenCL test with a score of 9082 against the AMD Radeon HD 8870M’s 8462, a margin of 6.8%. That is the sole benchmark where both chips appear together, so the Intel part claims the only win.

However, the AMD Radeon HD 8870M has its own statistical territory. Its average benchmark score sits at 8462, which places it in the 43rd percentile of all GPUs in the database. The Iris Pro P580, despite winning the head-to-head, has an average score of only 7170, dragged down by its additional Vulkan result of 5258. That places the Intel part in the 39th percentile. So in terms of consistency across multiple API workloads, AMD’s older chip actually holds a higher overall standing.

The use-case split is therefore nuanced. If the workload is OpenCL compute, the Intel Iris Pro P580 is the stronger performer. If the workload is Vulkan, the AMD chip has no recorded result, so the Intel part’s 5258 score is the only data point available. But when looking at the average across all recorded tests, the AMD Radeon HD 8870M comes out ahead by 1292 points, or roughly 15.3%. Neither chip dominates universally; each has a claim depending on which metric the analyst prioritizes.

Architecture Differences

The two GPUs come from fundamentally different design philosophies. The AMD Radeon HD 8870M uses the Venus chip built on GCN 1.0 architecture, manufactured by TSMC on a 28 nm process. It packs 1,500 million transistors into a 123 mm² die, giving a transistor density of 12.2 million per square millimeter. The Intel Iris Pro Graphics P580 uses the Skylake GT4e chip, part of Intel’s Generation 9.0 architecture, built on Intel’s own 14 nm+ process. Intel does not disclose transistor count or die size for this part, so a direct density comparison is impossible from the recorded data.

The compute resources differ in shape. AMD fields 640 shading units, 40 texture mapping units, and 16 ROPs. Intel counters with 576 shading units, 72 TMUs, and only 9 ROPs. That means AMD has 11.1% more shaders, while Intel has 80% more texture units and 43.8% fewer ROPs. These ratios explain why the two chips behave so differently across workloads: AMD’s design leans toward pixel throughput, Intel’s toward texture fill.

Clock speeds tell another story. The AMD chip runs at 725 MHz base and 775 MHz boost, while the Intel part starts at 350 MHz base and ramps to 1000 MHz boost. The Intel boost clock is 29% higher than AMD’s, but its base clock is 51.7% lower. Memory access is entirely different: AMD uses 2 GB of dedicated GDDR5 on a 128-bit bus, delivering 72.00 GB/s of bandwidth. Intel shares system memory, with bandwidth labeled “System Dependent,” meaning performance scales with the host platform’s memory configuration.

The API support also diverges. AMD lists DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. Intel lists DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. Intel’s higher DirectX feature level and newer Vulkan version suggest better modern API compliance, but the recorded benchmarks do not directly test those capabilities.

Head-to-Head Benchmarks

The database contains exactly one head-to-head comparison between these two GPUs: Geekbench OpenCL. The Intel Iris Pro Graphics P580 scores 9082, while the AMD Radeon HD 8870M scores 8462. That is a 6.8% advantage for Intel, or 620 points. In the context of the nearest rivals, this margin is meaningful but not overwhelming.

For context, the AMD Radeon HD 8870M’s 8462 score sits within 0.1% of the NVIDIA GeForce MX330 (8458), within 0.3% of the AMD Radeon 880M (8436), and within 0.4% of the NVIDIA GeForce GTX 675MX (8427). It trails the Intel Arc A380 (8558) by 1.1%. So the AMD chip is clustered tightly with a group of modern and older discrete parts, all within a 1.5% band.

The Intel Iris Pro P580’s 9082 OpenCL score is not directly compared to those same rivals in the database, but its average score of 7170 places it next to a different set. The NVIDIA GeForce GTX 560 SE matches it at 7171 (0.0% delta), the NVIDIA GeForce GTX 970 sits at 7157 (0.2% ahead), the AMD Radeon Vega 8 Mobile is at 7203 (0.5% behind), and the NVIDIA GeForce GTX 750 is at 7222 (0.7% behind). The Intel part’s Vulkan score of 5258 is far lower than its OpenCL result, suggesting that its compute performance is highly API-dependent.

The data implies that the Intel chip’s OpenCL win is real but narrow. A 6.8% margin in one test does not translate to a broad performance advantage. The AMD part’s average score is higher because it does not have a second weak result dragging it down.

FAQ

Q: Which GPU has the higher OpenCL benchmark score?

A: The Intel Iris Pro Graphics P580 scores 9082 in Geekbench OpenCL, while the AMD Radeon HD 8870M scores 8462. Intel leads by 6.8% in this specific test.

Q: Why does the AMD Radeon HD 8870M have a higher average benchmark score than the Intel Iris Pro P580?

A: The AMD chip has only one recorded benchmark result (8462 in OpenCL), while the Intel part has two results: 9082 in OpenCL and 5258 in Vulkan. Averaging Intel’s two scores yields 7170, which is lower than AMD’s single 8462.

Q: How does the AMD Radeon HD 8870M compare to its nearest rivals?

A: Its 8462 average score is 0.1% ahead of the NVIDIA GeForce MX330 (8458), 0.3% ahead of the AMD Radeon 880M (8436), 0.4% ahead of the NVIDIA GeForce GTX 675MX (8427), and 1.1% behind the Intel Arc A380 (8558).

Q: What is the Intel Iris Pro P580’s Vulkan performance?

A: The Intel part scores 5258 in Geekbench Vulkan. This is substantially lower than its OpenCL score of 9082, indicating that Vulkan workloads are not its strength.

Q: Which GPU has more shading units?

A: The AMD Radeon HD 8870M has 640 shading units, while the Intel Iris Pro P580 has 576. AMD has 11.1% more shaders.

Q: Which GPU has more texture mapping units?

A: The Intel Iris Pro P580 has 72 TMUs, while the AMD Radeon HD 8870M has 40. Intel has 80% more texture units.

The Verdict

The data points to a clear but conditional recommendation. For users running OpenCL compute workloads, the Intel Iris Pro Graphics P580 is the better choice, delivering 9082 versus 8462, a 6.8% improvement. This is a real, measurable advantage in the only direct comparison available.

For users who care about consistency across multiple API types or who rely on Vulkan, the picture changes. The Intel part’s Vulkan score of 5258 is dramatically lower than its OpenCL result, pulling its average down to 7170. The AMD Radeon HD 8870M, with no Vulkan result recorded, holds an average of 8462 and a higher percentile ranking (43rd versus 39th). That means the AMD chip is statistically safer for mixed workloads.

The AMD part also benefits from dedicated memory in the form of 2 GB GDDR5 with 72.00 GB/s bandwidth, whereas the Intel chip shares system memory with bandwidth dependent on the host platform. For compute tasks that are bandwidth-sensitive, AMD’s fixed memory pipeline is more predictable.

The Intel part has the edge in texture throughput (72.00 GTexel/s versus 31.00 GTexel/s) and raw FP32 compute (1,152.0 GFLOPS versus 992.0 GFLOPS), plus a higher boost clock of 1000 MHz versus 775 MHz. The AMD part counters with higher pixel rate (12.40 GPixel/s versus 9.000 GPixel/s) and more ROPs (16 versus 9).

There is no universal winner. Choose the Intel Iris Pro P580 if OpenCL is the primary API and texture-heavy workloads dominate. Choose the AMD Radeon HD 8870M if you need a higher average score, better percentile standing, or dedicated video memory.

Specification Differences

| Specification | AMD Radeon HD 8870M | Intel Iris Pro Graphics P580 |

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

| Chip | Venus | Skylake GT4e |

| Architecture | GCN 1.0 | Generation 9.0 |

| Process Node | 28 nm | 14 nm+ |

| Foundry | TSMC | Intel |

| Transistors | 1,500 million | Not disclosed |

| Die Size | 123 mm² | Not disclosed |

| Transistor Density | 12.2M / mm² | Not disclosed |

| Base Clock | 725 MHz | 350 MHz |

| Boost Clock | 775 MHz | 1000 MHz |

| Memory Size | 2 GB | System Shared |

| Memory Type | GDDR5 | System Shared |

| Memory Bus Width | 128 bit | System Shared |

| Memory Bandwidth | 72.00 GB/s | System Dependent |

| Shading Units | 640 | 576 |

| TMUs | 40 | 72 |

| ROPs | 16 | 9 |

| Pixel Rate | 12.40 GPixel/s | 9.000 GPixel/s |

| Texture Rate | 31.00 GTexel/s | 72.00 GTexel/s |

| FP32 Performance | 992.0 GFLOPS | 1,152.0 GFLOPS |

| FP16 Performance | Not disclosed | 2.304 TFLOPS (2:1) |

| TDP | Not disclosed | 15 W |

| Bus Interface | PCIe 3.0 x16 | Ring Bus |

| Display Outputs | Not disclosed | Motherboard Dependent |

| DirectX Support | 12 (11_1) | 12 (12_1) |

| Vulkan Support | 1.2.170 | 1.3 |

| Release Date | 2013-03-31 | 2015-08-31 |

| Production Status | End-of-life | End-of-life |