GPU Comparison

The AMD Radeon HD 8790M and NVIDIA GeForce 930A are two end-of-life mobile graphics processors that occupy similar performance tiers but arrive from different architectural generations. The data shows a narrow victory for the NVIDIA part in the only shared benchmark, yet the underlying specifications reveal divergent design philosophies that affect real-world capability beyond raw compute scores.

Head-to-Head Benchmarks

The sole direct comparison available is the Geekbench OpenCL test, where the NVIDIA GeForce 930A scores 5,317 points against the AMD Radeon HD 8790M’s 5,017 points. This represents a 5.6% advantage for the NVIDIA part, a margin that is meaningful but not overwhelming in practical terms. The delta places the 930A in a position to win compute-heavy workloads, yet the gap is small enough that driver optimizations or specific application characteristics could narrow it considerably.

Looking at the broader benchmark landscape, the AMD part’s average benchmark score of 5,691 is actually higher than its OpenCL result, suggesting the Vulkan workload is more favorable to its architecture. The Radeon HD 8790M scores 6,365 in Geekbench Vulkan, which is 26.9% higher than its OpenCL score and substantially above the 930A’s OpenCL result. This indicates that while NVIDIA wins the single cross-platform test, AMD’s part is capable of significantly better performance when utilizing modern graphics APIs.

The percentile rankings reinforce the closeness of this pairing. The AMD Radeon HD 8790M sits at the 33rd percentile of all GPUs, while the NVIDIA GeForce 930A sits at the 31st percentile. This 2-percentile gap is trivial, placing both parts in the same performance class. The nearest rivals for each part confirm this positioning: the AMD part’s closest competitor is the Intel Iris Pro Graphics P6300 (5,712 average score, 0.4% faster), while the NVIDIA part’s nearest rival is the GeForce 840M (5,322 average score, 0.1% faster). Both parts are effectively interchangeable in performance terms, with the 930A’s 5.6% OpenCL win being the only notable differentiator.

Architecture Differences

The architectural split between these two GPUs is substantial despite their similar performance. The AMD Radeon HD 8790M is built on the Graphics Core Next 1.0 architecture (chip codename Mars) from the Solar System generation of HD 8700M parts. In contrast, the NVIDIA GeForce 930A uses the Maxwell architecture (chip codename GM108) from the GeForce 900A generation. Both are fabricated on a 28 nm process at TSMC, and both have a die size of 77 mm², but the transistor counts differ slightly.

AMD packs 950 million transistors into its 77 mm² die, yielding a transistor density of 12.3 million per mm². NVIDIA fits 1,020 million transistors into the same die area, achieving 13.2 million per mm². This 7.4% higher density on the NVIDIA side reflects a more compact design, likely contributing to its higher clock speeds. The AMD part runs at a base clock of 850 MHz and a boost of 900 MHz, while the NVIDIA part runs at 928 MHz base and 941 MHz boost. That is a 9.2% higher base clock and 4.6% higher boost clock for NVIDIA.

The memory subsystems present the starkest contrast. The AMD Radeon HD 8790M uses 2 GB of GDDR5 memory on a 128-bit bus, delivering 64.00 GB/s of bandwidth. The NVIDIA GeForce 930A uses 2 GB of DDR3 memory on a 64-bit bus, delivering only 16.02 GB/s of bandwidth. This is a 4x difference in memory bandwidth in favor of AMD, which is a critical factor for texture-heavy workloads and higher resolutions. The memory clock speeds reflect this: AMD runs at 1000 MHz with 4 Gbps effective data rate, while NVIDIA runs at 1001 MHz but only 2 Gbps effective. The NVIDIA part’s memory is clocked at a similar base frequency but transfers half the data per cycle due to the DDR3 type and narrower bus.

Compute resources are identical in count but not in execution. Both parts have 384 shading units, 24 texture mapping units, and 8 render output units. The pixel rate is nearly identical at 7.200 GPixel/s for AMD and 7.528 GPixel/s for NVIDIA, a 4.6% advantage for the latter. The texture rate follows a similar pattern: 21.60 GTexel/s for AMD versus 22.58 GTexel/s for NVIDIA, a 4.5% advantage. The FP32 compute throughput is 691.2 GFLOPS for AMD and 722.7 GFLOPS for NVIDIA, a 4.6% difference. These near-identical rates are a direct consequence of the same core counts and similar clock speeds.

API support differs in notable ways. The AMD part supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The NVIDIA part supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. While both support DirectX 12, the feature levels differ, and NVIDIA’s newer Vulkan version (1.4 versus 1.2.170) may offer better support for recent extensions. The AMD part’s higher DirectX feature level (11_1 versus 11_0) could matter for specific legacy titles.

FAQ

Q: Which GPU has higher raw compute performance?

A: The NVIDIA GeForce 930A has a slight edge in FP32 throughput at 722.7 GFLOPS versus 691.2 GFLOPS for the AMD Radeon HD 8790M, a 4.6% advantage. This is reflected in the Geekbench OpenCL score where NVIDIA leads by 5.6%.

Q: Why does the AMD part have such a large lead in memory bandwidth?

A: The AMD Radeon HD 8790M uses GDDR5 memory on a 128-bit bus, yielding 64.00 GB/s. The NVIDIA GeForce 930A uses DDR3 on a 64-bit bus, yielding 16.02 GB/s. The 4x bandwidth advantage for AMD comes from both a wider bus and faster memory type.

Q: Are these GPUs suitable for modern graphics APIs?

A: Both support DirectX 12, OpenGL 4.6, and Vulkan. The AMD part supports DirectX 12 (11_1) and Vulkan 1.2.170, while the NVIDIA part supports DirectX 12 (11_0) and Vulkan 1.4. The NVIDIA part has a newer Vulkan version, but the AMD part has a higher DirectX feature level.

Q: How do these parts compare to their nearest competitors?

A: The AMD Radeon HD 8790M is 1.6% faster than the NVIDIA Quadro M500M and 0.4% slower than the Intel Iris Pro Graphics P6300. The NVIDIA GeForce 930A is 0.6% faster than the GeForce 940M and 0.2% faster than the GeForce GTX 980M.

Q: What is the power consumption of the NVIDIA part?

A: The NVIDIA GeForce 930A has a TDP of 33 W. No TDP is listed for the AMD Radeon HD 8790M in the data.

Q: Which GPU has better Vulkan performance?

A: The AMD Radeon HD 8790M scores 6,365 in Geekbench Vulkan, which is the only Vulkan result available. The NVIDIA GeForce 930A has no Vulkan benchmark score listed, but it supports a newer Vulkan version (1.4 versus 1.2.170).

The Verdict

The benchmark data indicates that the NVIDIA GeForce 930A is the better choice for raw OpenCL compute performance, winning the only head-to-head test by 5.6%. Its higher clock speeds (928 MHz base versus 850 MHz) and slightly higher transistor density (13.2M/mm² versus 12.3M/mm²) contribute to a modest but consistent lead in pixel rate, texture rate, and FP32 throughput. For users prioritizing compute workloads that rely on OpenCL, the 930A is the data-supported pick.

However, the AMD Radeon HD 8790M offers a compelling counterargument through its memory subsystem. The 64.00 GB/s of bandwidth versus 16.02 GB/s is a 4x advantage that could be decisive in memory-bound scenarios, even if the compute core is slightly slower. The AMD part’s Vulkan score of 6,365 is 19.7% higher than its OpenCL score, suggesting that the GCN architecture scales better with modern APIs. This could make the 8790M the better choice for users running Vulkan-based applications or games, despite the NVIDIA part’s newer Vulkan version support.

The production status and release dates favor the NVIDIA part, which was released on March 12, 2015, compared to the AMD part’s March 31, 2013 release. Both are end-of-life, but the NVIDIA part is from a newer generation. The physical form factors differ: the AMD part uses an MXM Module with a 3.0 bus interface, while the NVIDIA part is an integrated graphics processor (IGP) on PCIe 3.0 x8. This makes the AMD part more suitable for upgradeable laptops, while the NVIDIA part is likely soldered to the motherboard.

For users who need maximum compute throughput in OpenCL workloads, the NVIDIA GeForce 930A is the safer choice based on the single head-to-head result. For users who expect to use Vulkan or who need high memory bandwidth for texture-heavy tasks, the AMD Radeon HD 8790M presents a stronger case, especially considering its 27% higher Vulkan score over its own OpenCL result. The overall performance class is nearly identical, so the decision should hinge on which API and workload characteristics matter most.

Specification Differences

| Specification | AMD Radeon HD 8790M | NVIDIA GeForce 930A |

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

| Architecture | GCN 1.0 | Maxwell |

| Chip | Mars | GM108 |

| Transistors | 950 million | 1,020 million |

| Transistor Density | 12.3M / mm² | 13.2M / mm² |

| Base Clock | 850 MHz | 928 MHz |

| Boost Clock | 900 MHz | 941 MHz |

| Memory Type | GDDR5 | DDR3 |

| Memory Bus Width | 128 bit | 64 bit |

| Memory Bandwidth | 64.00 GB/s | 16.02 GB/s |

| Memory Effective Rate | 4 Gbps | 2 Gbps |

| Pixel Rate | 7.200 GPixel/s | 7.528 GPixel/s |

| Texture Rate | 21.60 GTexel/s | 22.58 GTexel/s |

| FP32 | 691.2 GFLOPS | 722.7 GFLOPS |

| DirectX Support | 12 (11_1) | 12 (11_0) |

| Vulkan Support | 1.2.170 | 1.4 |

| TDP | Not listed | 33 W |

| Slot Width | MXM Module | IGP |

| Bus Interface | MXM-A (3.0) | PCIe 3.0 x8 |

| Release Date | 2013-03-31 | 2015-03-12 |

| Geekbench OpenCL | 5,017 | 5,317 |