AMD Radeon HD 8690M vs Intel Iris Pro Graphics 6200 Comparison

The AMD Radeon HD 8690M and Intel Iris Pro Graphics 6200 occupy nearly the same performance tier, with average benchmark scores of 6137 and 6117 respectively. This places both in the 35th–36th percentile of all GPUs, indicating entry-level capability. The data shows a statistical dead heat in overall compute performance, yet the two chips achieve parity through fundamentally different designs, clock strategies, and architectural approaches.

Head-to-Head Benchmarks

The only directly comparable benchmark in the data set is Geekbench OpenCL, and it reveals a clear, if narrow, victory for the AMD Radeon HD 8690M. AMD scores 6137 against Intel’s 4556, a delta of 34.7% in AMD’s favor. That is a substantial margin in a synthetic compute workload, suggesting the Radeon’s dedicated GDDR5 memory and higher sustained clocks deliver a real advantage in OpenCL execution.

However, the picture becomes more nuanced when looking at Intel’s full benchmark suite. The Iris Pro Graphics 6200 posts a Geekbench Metal score of 7764 and a Vulkan score of 6032, both of which exceed its own OpenCL result of 4556. The Metal score, in particular, is 70% higher than its OpenCL result, indicating that the Intel iGPU is heavily optimized for Apple’s Metal API. The AMD part has no equivalent Metal or Vulkan results listed, so cross-API comparisons are impossible from this data.

In terms of average score across all benchmarks, the two are separated by just 0.3%, with AMD at 6137 and Intel at 6117. Looking at the nearest rival lists, both GPUs cluster tightly with NVIDIA RTX A400 (6078) and GeForce MX230 (6077), as well as the older AMD Radeon HD 6950 (6210). The deltaPct values between these rivals range from -1.2% to +1.1%, meaning all five GPUs fall within roughly a 2% performance band. This suggests that in raw compute terms, the HD 8690M and Iris Pro 6200 are interchangeable for most tasks, with the AMD card’s OpenCL win being the only significant differentiator in direct comparison.

The head-to-head table shows AMD wins one test, Intel wins zero. Yet that single win is decisive in OpenCL, a common compute language. Users running OpenCL workloads would see a 34.7% performance uplift on the AMD card, which is not a trivial difference. Conversely, users on macOS or applications that leverage Metal would likely prefer the Intel solution, given its strong Metal showing, though no direct AMD Metal data exists to confirm this.

Architecture Differences

The two GPUs come from opposite ends of the design spectrum. AMD’s Radeon HD 8690M uses the Sun chip built on GCN 1.0 architecture, manufactured by TSMC on a 28 nm process. It packs 690 million transistors into a 56 mm² die, yielding a transistor density of 12.3 million per square millimeter. Intel’s Iris Pro Graphics 6200, by contrast, uses the Broadwell GT3e chip with Generation 8.0 architecture, fabricated in-house on Intel’s 14 nm process. No transistor count or die size is provided for the Intel part, but the process node advantage is clear: 14 nm versus 28 nm.

Clock behavior further separates the pair. AMD runs at a base clock of 925 MHz with a boost up to 975 MHz, while Intel starts much lower at 300 MHz base but boosts aggressively to 1100 MHz. This means the Intel GPU has a 12.8% higher boost clock (1100 vs 975 MHz), but it must ramp up from a very low idle state. The AMD card holds a more consistent clock range, which likely explains its OpenCL advantage in sustained workloads.

Memory configurations are radically different. The AMD Radeon HD 8690M features 2 GB of dedicated GDDR5 memory on a 64-bit bus, delivering 32.00 GB/s of bandwidth. Intel’s Iris Pro uses System Shared memory, with a bus width and bandwidth listed as "System Dependent." This is a fundamental architectural divergence: discrete memory versus shared memory. Dedicated VRAM avoids bandwidth contention with the CPU, while shared memory is more flexible but dependent on system RAM speed and capacity.

Compute resources tell a mixed story. Intel has more shading units (384 vs 320) and significantly more texture mapping units (48 vs 20), giving it a texture rate of 52.80 GTexel/s versus AMD’s 19.50 GTexel/s — a 2.7x advantage in texture throughput. However, AMD has more ROPs (8 vs 6), resulting in a pixel rate of 7.800 GPixel/s versus Intel’s 6.600 GPixel/s. In raw FP32 compute, Intel leads with 844.8 GFLOPS against AMD’s 624.0 GFLOPS, a 35.4% advantage. Yet the OpenCL benchmark favors AMD, indicating that theoretical compute rates do not always translate to real-world API performance.

Power and integration also differ. Intel’s Iris Pro carries a 15 W TDP and is an IGP (integrated graphics processor) using a Ring Bus interface. AMD’s card has no TDP listed and uses PCIe 3.0 x8. Intel’s display outputs are "Motherboard Dependent," while AMD’s are not specified. Both support DirectX 12 (11_1) and OpenGL, but AMD supports Vulkan 1.2.170 versus Intel’s Vulkan 1.0. AMD also supports OpenGL 4.6 compared to Intel’s 4.4.

FAQ

Q: Which GPU is faster in Geekbench OpenCL?

A: The AMD Radeon HD 8690M scores 6137, which is 34.7% higher than the Intel Iris Pro Graphics 6200’s score of 4556 in the same test.

Q: How do their average benchmark scores compare?

A: The AMD card averages 6137, while the Intel card averages 6117. This puts them within 0.3% of each other, effectively a tie.

Q: Does the Intel Iris Pro perform better in any benchmark?

A: Yes. The Intel card scores 7764 in Geekbench Metal and 6032 in Geekbench Vulkan, both higher than its OpenCL score. However, no direct AMD results exist for Metal or Vulkan to compare against.

Q: What are the key memory differences?

A: AMD uses 2 GB of dedicated GDDR5 memory on a 64-bit bus with 32.00 GB/s bandwidth. Intel uses System Shared memory, with bandwidth described as "System Dependent."

Q: Which GPU has higher theoretical compute performance?

A: Intel has higher FP32 compute at 844.8 GFLOPS versus AMD’s 624.0 GFLOPS, a 35.4% advantage. Intel also has more shading units and TMUs.

Q: What process nodes do they use?

A: AMD uses TSMC’s 28 nm process with 690 million transistors on a 56 mm² die. Intel uses its own 14 nm process, though no transistor or die size data is provided.

The Verdict

The data supports different picks depending on workload. For OpenCL compute tasks, the AMD Radeon HD 8690M is the clear choice — its 34.7% lead in that benchmark is the largest performance gap recorded between the two. Users running general-purpose GPU compute, scientific workloads, or OpenCL-accelerated applications should prefer the AMD card.

For users in ecosystems that favor Metal or Vulkan, the Intel Iris Pro Graphics 6200 shows strong results in those APIs, though direct AMD comparisons are absent. The Intel part also offers better theoretical specs in FP32 (844.8 GFLOPS vs 624.0), texture rate (52.80 vs 19.50 GTexel/s), and shading units (384 vs 320). Its 15 W TDP and integrated nature make it suitable for thin-and-light systems, whereas the AMD card is a discrete solution using PCIe 3.0 x8.

The overall average scores are statistically tied, so for general use, either GPU will deliver similar results. The deciding factor is the specific API and workload mix. AMD wins decisively in OpenCL; Intel appears stronger in texture-heavy workloads and has superior Vulkan and Metal support versions. Both are end-of-life products, so the choice is likely driven by existing hardware rather than new purchases. In essence: pick AMD for OpenCL compute, pick Intel for broader API coverage and integrated efficiency.

Specification Differences

| Specification | AMD Radeon HD 8690M | Intel Iris Pro Graphics 6200 |

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

| Manufacturer | AMD | Intel |

| Chip | Sun | Broadwell GT3e |

| Architecture | GCN 1.0 | Generation 8.0 |

| Process Node | 28 nm | 14 nm |

| Foundry | TSMC | Intel |

| Transistors | 690 million | Not specified |

| Die Size | 56 mm² | Not specified |

| Transistor Density | 12.3M / mm² | Not specified |

| Base Clock | 925 MHz | 300 MHz |

| Boost Clock | 975 MHz | 1100 MHz |

| Memory Size | 2 GB | System Shared |

| Memory Type | GDDR5 | System Shared |

| Memory Bus Width | 64 bit | System Shared |

| Memory Bandwidth | 32.00 GB/s | System Dependent |

| Shading Units | 320 | 384 |

| TMUs | 20 | 48 |

| ROPs | 8 | 6 |

| Pixel Rate | 7.800 GPixel/s | 6.600 GPixel/s |

| Texture Rate | 19.50 GTexel/s | 52.80 GTexel/s |

| FP32 | 624.0 GFLOPS | 844.8 GFLOPS |

| TDP | Not specified | 15 W |

| Slot Width | Not specified | IGP |

| Bus Interface | PCIe 3.0 x8 | Ring Bus |

| Display Outputs | Not specified | Motherboard Dependent |

| OpenGL | 4.6 | 4.4 |

| Vulkan | 1.2.170 | 1.0 |

| Release Date | 2013-02-28 | 2014-09-04 |

| Predecessor | London | Not specified |

| Successor | Gem System | Not specified |

| Percentile | 36 | 35 |