AMD Radeon HD 8790M vs NVIDIA GeForce GTX 580M Comparison

Head-to-Head Benchmarks

The database contains one direct benchmark comparison between these two mobile graphics processors: Geekbench OpenCL. In that test, the NVIDIA GeForce GTX 580M records a score of 6389, while the AMD Radeon HD 8790M posts 5017. The NVIDIA part wins by 27.3%, a substantial margin that reflects meaningful differences in raw compute throughput and memory subsystem design.

The GTX 580M's OpenCL score places it at the 37th percentile among all GPUs in the database. Its nearest rivals show how tight the mid-pack clustering is: the NVIDIA GeForce GTX 460 SE matches it exactly at 6389 with a 0% delta, while the NVIDIA RTX PRO 5000 72 GB Blackwell sits just 0.3% higher at 6407. On the lower side, the AMD Radeon Pro WX 4100 trails by 0.9% at 6330, and the AMD Radeon R7 M350 is 1% behind at 6327. This means the GTX 580M is essentially locked into a four-way statistical tie with those parts, none of which moves more than a single percentage point away from its score.

The HD 8790M, by contrast, sits at the 33rd percentile overall, a modestly lower standing. Its average benchmark score across both recorded tests is 5691, which is 698 points below the GTX 580M's single OpenCL result. The AMD card's nearest rivals bracket it from both sides. The Intel Iris Pro Graphics P6300 is 0.4% ahead at 5712, the NVIDIA GeForce GTX 670MX is 0.5% ahead at 5721, and the NVIDIA GeForce GTX 550 Ti is 0.7% ahead at 5731. Below it, the NVIDIA Quadro M500M trails by 1.6% at 5604. The HD 8790M is therefore surrounded by competitors within a 2.3% total spread, but none of them approach the GTX 580M's score.

Looking at the head-to-head delta more carefully: the 27.3% advantage for the GTX 580M is not a marginal win. It is a decisive gap that places the two cards in different performance tiers despite both shipping with 384 shading units. The GTX 580M also holds an additional benchmark result in its own right (a second OpenCL run is not listed, but the single score stands), while the HD 8790M only manages to close some ground in Vulkan, where it scores 6365. That Vulkan figure is worth noting because it nearly matches the GTX 580M's OpenCL number, but the database does not record a Vulkan result for the NVIDIA card, so no direct comparison is possible on that API.

The wins tally reflects the data: the GTX 580M wins 1 head-to-head benchmark, the HD 8790M wins 0. No scenario in the recorded measurements favors the AMD part.

FAQ

Q: Which GPU is faster in OpenCL compute?

A: The NVIDIA GeForce GTX 580M scores 6389 in Geekbench OpenCL, while the AMD Radeon HD 8790M scores 5017. The NVIDIA part leads by 27.3%.

Q: Does the AMD card have any benchmark where it beats the NVIDIA card?

A: No. The head-to-head comparison shows only one recorded test, Geekbench OpenCL, and the GTX 580M wins it. The HD 8790M does record a Vulkan score of 6365, but no Vulkan result exists for the GTX 580M, so no direct comparison is possible.

Q: How does the GTX 580M compare to its nearest rivals?

A: The GTX 580M is statistically tied with the NVIDIA GeForce GTX 460 SE (exact match at 6389), sits 0.3% below the NVIDIA RTX PRO 5000 72 GB Blackwell (6407), and sits 0.9% above the AMD Radeon Pro WX 4100 (6330) and 1% above the AMD Radeon R7 M350 (6327).

Q: Where does the HD 8790M rank among all GPUs?

A: The HD 8790M sits at the 33rd percentile of all GPUs in the database, with an average benchmark score of 5691 across its two recorded tests.

Q: What is the HD 8790M's closest competitor?

A: The Intel Iris Pro Graphics P6300 is 0.4% ahead at 5712, and the NVIDIA GeForce GTX 670MX is 0.5% ahead at 5721. The NVIDIA Quadro M500M trails it by 1.6% at 5604.

Q: What is the performance gap between these two cards in percentage terms?

A: The GTX 580M outperforms the HD 8790M by 27.3% in the sole head-to-head OpenCL test.

Architecture Differences

The two GPUs come from different architectural generations and foundry processes, and those differences explain much of the performance gap. The NVIDIA GeForce GTX 580M uses the GF114 chip based on the Fermi 2.0 architecture, built on a 40 nm process at TSMC. The AMD Radeon HD 8790M uses the Mars chip based on GCN 1.0, built on a 28 nm process, also at TSMC.

The transistor counts reflect the generational shift. The GTX 580M packs 1,950 million transistors onto a 332 mm² die, yielding a transistor density of 5.9 million transistors per square millimeter. The HD 8790M, despite being a smaller and more modern part, carries 950 million transistors on a 77 mm² die, which gives it a much higher density of 12.3 million transistors per square millimeter. The AMD chip is therefore more than four times smaller in die area while still containing roughly half the transistor count, a direct consequence of the 28 nm node versus 40 nm.

Both cards use 2 GB of GDDR5 memory, but the memory interfaces diverge sharply. The GTX 580M runs a 256 bit bus with a memory clock of 750 MHz (3 Gbps effective), producing 96.00 GB/s of bandwidth. The HD 8790M uses a 128 bit bus with a memory clock of 1000 MHz (4 Gbps effective), yielding 64.00 GB/s. The NVIDIA part delivers 50% more memory bandwidth despite its lower memory clock, purely because the bus width is doubled. That bandwidth advantage is critical for compute workloads that stream large data sets.

The shading unit count is identical at 384 for both cards, but the rest of the pipeline differs. The GTX 580M has 64 texture mapping units and 32 ROPs, while the HD 8790M has 24 TMUs and 8 ROPs. The GTX 580M's pixel rate is 9.920 GPixel/s versus 7.200 GPixel/s for the AMD part, and its texture rate is 39.68 GTexel/s versus 21.60 GTexel/s. In raw FP32 compute, the GTX 580M reaches 952.3 GFLOPS, while the HD 8790M reaches 691.2 GFLOPS. Every throughput metric favors the NVIDIA card, often by large margins.

Clock speeds tell a different story. The HD 8790M has a base clock of 850 MHz and a boost clock of 900 MHz, while the GTX 580M has no base or boost clock listed in the database. The AMD card's higher clocks partially compensate for its narrower pipeline, but not enough to close the compute gap. The GTX 580M achieves its higher throughput through sheer resource count: more TMUs, more ROPs, and a wider memory bus.

API support also differs. The GTX 580M supports DirectX 12 (11_0) and OpenGL 4.6, with no Vulkan entry recorded. The HD 8790M supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The AMD card is more modern in API coverage, particularly for Vulkan, which aligns with its later release date. The GTX 580M was released on 2011-06-27, while the HD 8790M arrived on 2013-03-31, nearly two years later.

Power and physical specifications differ as well. The GTX 580M has a TDP of 100 W, while the HD 8790M has no TDP listed. Both use MXM modules, but the bus interfaces differ: the GTX 580M uses MXM-B (3.0), and the HD 8790M uses MXM-A (3.0). Neither card requires auxiliary power connectors, and both are marked as end-of-life products.

The Verdict

The data points to a clear hierarchy. The NVIDIA GeForce GTX 580M is the faster GPU in every recorded benchmark, with a 27.3% OpenCL lead over the AMD Radeon HD 8790M. It also holds a higher percentile ranking (37th versus 33rd) and a higher average benchmark score (6389 versus 5691). For any workload that relies on OpenCL compute, the GTX 580M is the objectively stronger choice.

The GTX 580M's advantages are not subtle. Its memory bandwidth is 96.00 GB/s versus 64.00 GB/s, a 50% advantage. Its texture fill rate is 39.68 GTexel/s versus 21.60 GTexel/s, an 84% advantage. Its pixel rate is 9.920 GPixel/s versus 7.200 GPixel/s, a 38% advantage. Its FP32 throughput is 952.3 GFLOPS versus 691.2 GFLOPS, a 38% advantage. These are not rounding differences; they are structural gaps rooted in the NVIDIA card's wider memory bus, higher TMU count, and higher ROP count.

The HD 8790M does have redeeming qualities, but they are not compute-related. It supports Vulkan 1.2.170, which the GTX 580M does not, making it the better option for applications that rely on Vulkan. It also runs at higher clocks (900 MHz boost versus no listed clocks on the NVIDIA part) and uses a more modern 28 nm process with higher transistor density. For a user prioritizing API coverage or power efficiency per transistor, the AMD card has a reasonable argument. But the database does not record any performance test where the HD 8790M wins.

Who should pick the GTX 580M? Anyone whose priority is raw OpenCL compute, memory bandwidth, or texture throughput. The data shows it dominates in all three. Who should pick the HD 8790M? Anyone who needs Vulkan support, prefers a smaller die with higher transistor density, or wants a later product generation. Neither card is currently in production, so availability is a separate question, but on benchmark merit alone, the GTX 580M is the superior mobile GPU.

The verdict is straightforward: the GTX 580M wins the head-to-head, wins the percentile ranking, and wins every throughput metric. The HD 8790M counters with modern API support and a more efficient process node, but those attributes do not translate into benchmark victories in the recorded data.

Specification Differences

| Specification | NVIDIA GeForce GTX 580M | AMD Radeon HD 8790M |

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

| Architecture | Fermi 2.0 | GCN 1.0 |

| Process Node | 40 nm | 28 nm |

| Transistors | 1,950 million | 950 million |

| Die Size | 332 mm² | 77 mm² |

| Transistor Density | 5.9M / mm² | 12.3M / mm² |

| Base Clock | Not listed | 850 MHz |

| Boost Clock | Not listed | 900 MHz |

| Memory Clock | 750 MHz (3 Gbps effective) | 1000 MHz (4 Gbps effective) |

| Memory Bus Width | 256 bit | 128 bit |

| Memory Bandwidth | 96.00 GB/s | 64.00 GB/s |

| TMUs | 64 | 24 |

| ROPs | 32 | 8 |

| Pixel Rate | 9.920 GPixel/s | 7.200 GPixel/s |

| Texture Rate | 39.68 GTexel/s | 21.60 GTexel/s |

| FP32 | 952.3 GFLOPS | 691.2 GFLOPS |

| TDP | 100 W | Not listed |

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

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

| Vulkan Support | Not listed | 1.2.170 |

| Release Date | 2011-06-27 | 2013-03-31 |

| Predecessor | GeForce 400M | London |

| Successor | GeForce 600M | Gem System |

| Geekbench OpenCL | 6389 | 5017 |

| Geekbench Vulkan | Not listed | 6365 |

| Average Benchmark Score | 6389 | 5691 |

| Percentile vs All GPUs | 37 | 33 |

The two cards share several traits: both have 384 shading units, 2 GB of GDDR5 memory, MXM module slot width, no power connectors, portable-device-dependent display outputs, OpenGL 4.6 support, and end-of-life production status. The differences above are where the architectural and performance gaps reside.