AMD Radeon HD 8790M vs Intel Iris Pro Graphics 6200 Comparison

Where Each One Wins

The benchmark data splits cleanly between these two mobile graphics parts. The AMD Radeon HD 8790M takes both recorded head-to-head victories, while the Intel Iris Pro Graphics 6200 has no direct wins in the shared test set. That said, the margin in each case is modest, and the overall picture is closer than the raw win count suggests.

Looking at the individual workloads, the AMD part wins the OpenCL test with a score of 5017 against Intel's 4556, a delta of 9.2 percent in AMD's favor. In the Vulkan test, AMD again comes out ahead, scoring 6365 versus Intel's 6032, a 5.2 percent advantage. These are the only two benchmarks where both GPUs were measured under identical conditions, so the head-to-head story is defined entirely by those two results.

However, the broader database picture tells a more nuanced story. The Intel Iris Pro Graphics 6200 carries an average benchmark score of 6117 across all recorded tests, while the AMD Radeon HD 8790M averages 5691. That is a substantial gap of roughly 426 points, or about 7.5 percent, in Intel's favor when considering all measurements rather than just the shared tests. The Intel part also sits at the 35th percentile among all GPUs in the database, compared to the AMD part's 33rd percentile. So while AMD wins the two tests where both were measured side by side, Intel holds a higher overall standing in the database.

The nearest rival data reinforces this mixed picture. The Intel part's closest competitors are the AMD Radeon HD 8690M (0.3 percent behind), the NVIDIA RTX A400 (0.6 percent ahead), the NVIDIA GeForce MX230 (0.7 percent ahead), and the NVIDIA Quadro P2000 (1.1 percent ahead). The AMD part's nearest rivals are the Intel Iris Pro Graphics P6300 (0.4 percent behind), the NVIDIA GeForce GTX 670MX (0.5 percent behind), the NVIDIA GeForce GTX 550 Ti (0.7 percent behind), and the NVIDIA Quadro M500M (1.6 percent ahead). Both parts sit in a tight cluster of comparable mobile and low-power desktop GPUs, with no dominant separation from their immediate peers.

In terms of use-case wins, the AMD Radeon HD 8790M demonstrates a consistent edge in compute-oriented API workloads, winning both OpenCL and Vulkan. The Intel Iris Pro Graphics 6200, by contrast, appears stronger in the aggregate database metrics, which likely reflect a broader mix of workloads where its architecture performs favorably. The data suggests AMD wins where the specific API path is compute-heavy, while Intel wins the overall average.

Architecture Differences

The two GPUs come from fundamentally different design philosophies and manufacturing processes. The Intel Iris Pro Graphics 6200 is built on Broadwell GT3e silicon using a 14 nm process at Intel's own foundry. It belongs to the Generation 8.0 architecture, part of the HD Graphics (Broadwell) family. The AMD Radeon HD 8790M uses the Mars chip, built on GCN 1.0 architecture, manufactured by TSMC on a 28 nm process. The process node difference is significant: 14 nm versus 28 nm gives Intel a density and efficiency advantage on paper, though the AMD part compensates with a higher base clock.

The compute resources are configured differently despite sharing the same shading unit count. Both parts have 384 shading units, but Intel pairs them with 48 texture mapping units and 6 ROPs, while AMD uses 24 TMUs and 8 ROPs. This means Intel has double the texture throughput hardware but fewer pixel output units. The measured rates reflect this: Intel achieves 52.80 GTexel/s against AMD's 21.60 GTexel/s, a 2.4x advantage in texture fill. AMD counters with 7.200 GPixel/s against Intel's 6.600 GPixel/s, a 9 percent lead in pixel fill.

Clock behavior differs sharply. The Intel part runs at a 300 MHz base clock with a 1100 MHz boost, while the AMD part runs at 850 MHz base and 900 MHz boost. The AMD clock is far more stable, with only a 50 MHz gap between base and boost. The Intel part relies on a much larger boost multiplier to reach its peak performance. Memory configurations are entirely different as well. The Intel GPU uses system shared memory with a system dependent bandwidth, while the AMD part has 2 GB of dedicated GDDR5 memory on a 128 bit bus with 64.00 GB/s of bandwidth. That dedicated memory gives AMD a clear advantage in memory bandwidth consistency.

The transistor budgets reflect the process gap. AMD's Mars chip contains 950 million transistors on a 77 mm² die, yielding a density of 12.3 million transistors per mm². Intel does not report transistor count or die size in the database, but the 14 nm process and embedded DRAM design of the Broadwell GT3e typically imply a larger die. The AMD part also has a documented predecessor (London) and successor (Gem System), while Intel's part does not list either in the database.

Power and physical design differ as well. Intel's Iris Pro Graphics 6200 is rated at 15 W TDP and uses an IGP slot width with a Ring Bus interface. AMD's Radeon HD 8790M does not list a TDP in the database, uses an MXM Module slot width, and connects via MXM-A (3.0). Display outputs follow the platform: Intel's are motherboard dependent, AMD's are portable device dependent. API support shows AMD with a more modern OpenGL version (4.6 versus 4.4) and a newer Vulkan version (1.2.170 versus 1.0), while both support DirectX 12 (11_1).

Head-to-Head Benchmarks

The two shared benchmarks provide the clearest direct comparison. In Geekbench OpenCL, the AMD Radeon HD 8790M scores 5017 against Intel's 4556. That is a 9.2 percent advantage for AMD, the larger of the two head-to-head margins. The result is notable because Intel's part has substantially higher texture throughput (52.80 GTexel/s versus 21.60 GTexel/s) and higher FP32 compute (844.8 GFLOPS versus 691.2 GFLOPS). Despite those theoretical advantages, AMD wins the OpenCL workload, suggesting that memory bandwidth and driver efficiency play a larger role in this particular test.

In Geekbench Vulkan, the AMD part scores 6365 against Intel's 6032, a 5.2 percent margin. Both scores are higher than their respective OpenCL results, which is typical for Vulkan's lower API overhead. AMD's advantage here is smaller than in OpenCL, but still consistent. The Vulkan API version difference is worth noting: AMD supports Vulkan 1.2.170 while Intel supports 1.0. That newer API path may contribute to AMD's edge in this workload.

When the full database is considered, the picture inverts. The Intel part's average benchmark score of 6117 exceeds AMD's 5691 by 426 points. That difference is roughly 7.5 percent. The Intel part also sits at the 35th percentile versus AMD's 33rd. These aggregate numbers include tests beyond the two shared ones, so they reflect a broader set of workloads. The Intel part's higher FP32 rate (844.8 GFLOPS versus 691.2 GFLOPS) and much higher texture rate likely drive its advantage in other benchmarks.

The nearest rival data adds context. The Intel part's nearest competitor, the AMD Radeon HD 8690M, scores 6137, just 0.3 percent above Intel's 6117 average. The AMD Radeon HD 8790M's nearest competitor, the Intel Iris Pro Graphics P6300, scores 5712, just 0.4 percent above AMD's 5691. Both parts are surrounded by a dense field of similarly performing GPUs, with deltas under 2 percent in all cases. Neither part dominates its immediate competitive set.

In the two direct comparisons, the AMD part wins both, but the margins are 9.2 percent and 5.2 percent. The Intel part wins the aggregate comparison by about 7.5 percent. Those numbers suggest that AMD has a targeted advantage in the specific API workloads measured head to head, while Intel has a broader overall performance profile across the database's full test suite.

The Verdict

The data points to a split decision. For users running OpenCL or Vulkan workloads specifically, the AMD Radeon HD 8790M is the stronger choice. It wins both shared benchmarks, with a 9.2 percent lead in OpenCL and a 5.2 percent lead in Vulkan. Its dedicated 2 GB GDDR5 memory with 64.00 GB/s bandwidth likely contributes to these results, as compute workloads often depend heavily on memory throughput. The AMD part also has a higher pixel rate (7.200 GPixel/s versus 6.600 GPixel/s), which benefits rasterization-heavy tasks.

For users considering the broader performance picture, the Intel Iris Pro Graphics 6200 holds the higher average benchmark score (6117 versus 5691) and a better percentile ranking (35th versus 33rd). Its FP32 output of 844.8 GFLOPS exceeds AMD's 691.2 GFLOPS by about 22 percent, and its texture rate of 52.80 GTexel/s more than doubles AMD's 21.60 GTexel/s. These advantages likely translate to better performance in workloads that are not captured by the two shared tests, such as geometry-heavy rendering or texture-bound scenarios.

The integrated nature of the Intel part is a practical consideration. It is an IGP with a 15 W TDP, built into the Broadwell platform, using system shared memory. The AMD part is a discrete MXM module with its own 2 GB of GDDR5 memory. For portable devices, the AMD part offers the benefit of dedicated memory, while the Intel part integrates into the CPU package with lower power draw. The choice between them depends on whether the user prioritizes the two compute workloads where AMD wins, or the broader average where Intel wins.

The tight nearest rival clusters for both parts indicate that neither GPU is dramatically better than its immediate competition. The Intel part's nearest rivals are all within 1.1 percent of its average score. The AMD part's nearest rivals are all within 1.6 percent. This is a segment where small architectural differences and driver maturity matter more than raw specification sheets.

The final recommendation follows the data. Choose the AMD Radeon HD 8790M for OpenCL and Vulkan compute tasks, where it holds a measurable edge. Choose the Intel Iris Pro Graphics 6200 for overall versatility, where its higher average score and better percentile suggest broader competence across a wider range of workloads.

FAQ

Q: Which GPU wins the OpenCL benchmark?

A: The AMD Radeon HD 8790M wins Geekbench OpenCL with a score of 5017, compared to 4556 for the Intel Iris Pro Graphics 6200, a 9.2 percent advantage.

Q: Which GPU has the higher average benchmark score?

A: The Intel Iris Pro Graphics 6200 has an average benchmark score of 6117, while the AMD Radeon HD 8790M averages 5691, giving Intel a lead of about 426 points.

Q: How much memory does each GPU have?

A: The AMD Radeon HD 8790M has 2 GB of dedicated GDDR5 memory on a 128 bit bus with 64.00 GB/s bandwidth. The Intel Iris Pro Graphics 6200 uses system shared memory with system dependent bandwidth.

Q: What are the shading unit counts for both GPUs?

A: Both GPUs have 384 shading units. The Intel part has 48 texture mapping units and 6 ROPs, while the AMD part has 24 TMUs and 8 ROPs.

Q: Which GPU has a higher FP32 compute rate?

A: The Intel Iris Pro Graphics 6200 has an FP32 rate of 844.8 GFLOPS, compared to 691.2 GFLOPS for the AMD Radeon HD 8790M, a difference of roughly 22 percent.

Q: What Vulkan API versions do the two GPUs support?

A: The AMD Radeon HD 8790M supports Vulkan 1.2.170, while the Intel Iris Pro Graphics 6200 supports Vulkan 1.0. Both support DirectX 12 (11_1).

Specification Differences

| Specification | Intel Iris Pro Graphics 6200 | AMD Radeon HD 8790M |

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

| Chip | Broadwell GT3e | Mars |

| Architecture | Generation 8.0 | GCN 1.0 |

| Process Node | 14 nm | 28 nm |

| Foundry | Intel | TSMC |

| Transistors | Not reported | 950 million |

| Die Size | Not reported | 77 mm² |

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

| Base Clock | 300 MHz | 850 MHz |

| Boost Clock | 1100 MHz | 900 MHz |

| Memory Size | System Shared | 2 GB |

| Memory Type | System Shared | GDDR5 |

| Memory Bus Width | System Shared | 128 bit |

| Memory Bandwidth | System Dependent | 64.00 GB/s |

| Texture Mapping Units | 48 | 24 |

| ROPs | 6 | 8 |

| Pixel Rate | 6.600 GPixel/s | 7.200 GPixel/s |

| Texture Rate | 52.80 GTexel/s | 21.60 GTexel/s |

| FP32 | 844.8 GFLOPS | 691.2 GFLOPS |

| TDP | 15 W | Not reported |

| Slot Width | IGP | MXM Module |

| Bus Interface | Ring Bus | MXM-A (3.0) |

| OpenGL | 4.4 | 4.6 |

| Vulkan | 1.0 | 1.2.170 |

| Release Date | 2014-09-04 | 2013-03-31 |

| Predecessor | Not reported | London |

| Successor | Not reported | Gem System |

| Avg Benchmark Score | 6117 | 5691 |

| Percentile vs All GPUs | 35 | 33 |