AMD Radeon R7 350 vs NVIDIA GeForce GTX 580M Comparison

Head-to-Head Benchmarks

The only direct comparison available in the database is Geekbench OpenCL, and it is a decisive win for the AMD Radeon R7 350. The AMD part scores 7792 against 6389 for the NVIDIA GeForce GTX 580M, a 22% advantage. This is not a marginal gap; it represents a substantial performance lead in general-purpose compute workloads that use OpenCL.

Context from the nearest rivals reinforces the R7 350's standing. Its average benchmark score is 7425, placing it just 0.2% below the Intel UHD Graphics 750 (7441) and 0.3% below the AMD Radeon HD 8850M (7447). The R7 350 also trails the NVIDIA GeForce GTX 1650 by only 0.6% (7472) and the Intel Arc A310 by 1.7% (7550). In practical terms, the R7 350 sits in a tight cluster of GPUs where a few percentage points separate the entire group, and its 22% lead over the GTX 580M is far larger than any delta among its own nearest rivals.

The GTX 580M, by contrast, has an average benchmark score of 6389. Its nearest rivals are a mixed bag: it ties the NVIDIA GeForce GTX 460 SE exactly (0% delta), sits 0.3% below the NVIDIA RTX PRO 5000 72 GB Blackwell (6407), and is 0.9% and 1.0% ahead of the AMD Radeon Pro WX 4100 (6330) and AMD Radeon R7 M350 (6327), respectively. The score profile shows that the GTX 580M is competitive with its immediate peers, but it simply cannot match the R7 350 in the one benchmark where both were measured.

The percentile data tells the same story at a higher level. The R7 350 lands in the 40th percentile of all GPUs in the database, while the GTX 580M sits at the 37th percentile. A three-percentile gap is modest, but combined with the 22% OpenCL delta, it confirms that the AMD card is the stronger performer in this pairing.

Architecture Differences

The two GPUs come from different architectural generations and fundamentally different design philosophies. The AMD Radeon R7 350 uses the Cape Verde chip on GCN 1.0 architecture, built on a 28 nm process at TSMC. The NVIDIA GeForce GTX 580M uses the GF114 chip on Fermi 2.0 architecture, built on a larger 40 nm process, also at TSMC.

The transistor counts and die sizes reflect these process differences. The R7 350 packs 1,500 million transistors into a 123 mm² die, yielding a transistor density of 12.2 million transistors per square millimeter. The GTX 580M has more transistors overall, 1,950 million, but spreads them across a much larger 332 mm² die, resulting in a density of just 5.9 million per square millimeter. The R7 350 is more than twice as dense, a direct consequence of the newer 28 nm process versus the 40 nm node used by the Fermi part.

The compute configurations differ in interesting ways. The R7 350 has 512 shading units, 32 texture mapping units, and 16 ROPs. The GTX 580M has fewer shading units at 384, but more TMUs at 64 and more ROPs at 32. This split means the NVIDIA part has double the texture and pixel throughput capacity relative to its shader count, while the AMD part is more shader-heavy. The measured rates reflect this: the R7 350 achieves 12.80 GPixel/s pixel rate and 25.60 GTexel/s texture rate, while the GTX 580M achieves 9.920 GPixel/s and 39.68 GTexel/s. The R7 350 wins on pixel throughput, the GTX 580M wins on texture throughput.

Memory configurations also diverge. Both cards have 2 GB of GDDR5, but the R7 350 uses a 128-bit bus with 72.00 GB/s of bandwidth, while the GTX 580M uses a 256-bit bus with 96.00 GB/s. The memory clock rates are 1125 MHz (4.5 Gbps effective) for the R7 350 and 750 MHz (3 Gbps effective) for the GTX 580M, but the wider bus on the NVIDIA part gives it the bandwidth advantage.

Floating-point performance is close. The R7 350 delivers 819.2 GFLOPS of FP32, while the GTX 580M delivers 952.3 GFLOPS. Despite the AMD card's newer architecture, the NVIDIA part actually has a 16% raw FP32 advantage, though this does not translate into a benchmark win in the recorded OpenCL test.

Power and physical requirements differ sharply. The R7 350 has a 55 W TDP, is a single-slot card, and requires no power connectors, with a suggested power supply of 250 W. The GTX 580M has a 100 W TDP and uses an MXM module form factor, which means it was designed for laptops and mobile workstations rather than desktop expansion slots. The R7 350 uses PCIe 3.0 x16, while the GTX 580M uses MXM-B (3.0). Display outputs are another distinction: the R7 350 offers 1x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2, while the GTX 580M's outputs are listed as portable device dependent, meaning they vary by the laptop implementation.

API support shows the age difference. The R7 350 supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The GTX 580M supports DirectX 12 (11_0) and OpenGL 4.6, but has no Vulkan support recorded in the database. For modern workloads that rely on Vulkan, the AMD card has a clear feature advantage.

FAQ

Q: Which GPU is faster in Geekbench OpenCL?

A: The AMD Radeon R7 350 scores 7792 compared to 6389 for the NVIDIA GeForce GTX 580M, a 22% lead in favor of the AMD card.

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

A: The GTX 580M ties the NVIDIA GeForce GTX 460 SE at 6389, is 0.3% below the NVIDIA RTX PRO 5000 72 GB Blackwell (6407), and is 0.9% ahead of the AMD Radeon Pro WX 4100 (6330) and 1.0% ahead of the AMD Radeon R7 M350 (6327).

Q: Do both cards support the same modern APIs?

A: No. The R7 350 supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The GTX 580M supports DirectX 12 (11_0) and OpenGL 4.6, but has no Vulkan support in the database.

Q: Which card has more memory bandwidth?

A: The NVIDIA GeForce GTX 580M has 96.00 GB/s of bandwidth thanks to its 256-bit bus, while the AMD Radeon R7 350 has 72.00 GB/s over a 128-bit bus. Both have 2 GB of GDDR5 memory.

Q: What are the power requirements for each card?

A: The R7 350 has a 55 W TDP with no power connectors and a suggested 250 W power supply. The GTX 580M has a 100 W TDP and no power connectors, but no suggested PSU is recorded because it uses an MXM module form factor.

Q: How do their raw compute rates compare?

A: The R7 350 delivers 819.2 GFLOPS FP32, while the GTX 580M delivers 952.3 GFLOPS, making the NVIDIA card about 16% higher in raw FP32 throughput despite losing the OpenCL benchmark.

Specification Differences

| Specification | AMD Radeon R7 350 | NVIDIA GeForce GTX 580M |

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

| Chip | Cape Verde | GF114 |

| Architecture | GCN 1.0 | Fermi 2.0 |

| Generation | Pirate Islands (R7 300) | GeForce 500M |

| Process node | 28 nm | 40 nm |

| Transistors | 1,500 million | 1,950 million |

| Die size | 123 mm² | 332 mm² |

| Transistor density | 12.2M / mm² | 5.9M / mm² |

| Memory clock | 1125 MHz, 4.5 Gbps effective | 750 MHz, 3 Gbps effective |

| Memory bus width | 128 bit | 256 bit |

| Memory bandwidth | 72.00 GB/s | 96.00 GB/s |

| Shading units | 512 | 384 |

| TMUs | 32 | 64 |

| ROPs | 16 | 32 |

| Pixel rate | 12.80 GPixel/s | 9.920 GPixel/s |

| Texture rate | 25.60 GTexel/s | 39.68 GTexel/s |

| FP32 | 819.2 GFLOPS | 952.3 GFLOPS |

| TDP | 55 W | 100 W |

| Slot width | Single-slot | MXM Module |

| Bus interface | PCIe 3.0 x16 | MXM-B (3.0) |

| Display outputs | 1x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 | Portable device dependent |

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

| Vulkan | 1.2.170 | None |

| Release date | 2016-07-05 | 2011-06-27 |

The Verdict

The data points to a clear winner for most use cases: the AMD Radeon R7 350. It wins the only head-to-head benchmark available, posting a 22% higher OpenCL score than the GTX 580M. Its 40th percentile ranking versus the GTX 580M's 37th percentile reinforces that the AMD card is the stronger overall performer in the database's measurements.

The GTX 580M is not without merit. Its 256-bit memory bus delivers 96.00 GB/s of bandwidth, a third more than the R7 350's 72.00 GB/s. Its FP32 output of 952.3 GFLOPS exceeds the R7 350's 819.2 GFLOPS. It also has double the TMUs and double the ROPs, which gives it a 39.68 GTexel/s texture rate that is well ahead of the R7 350's 25.60 GTexel/s. For workloads that are heavily texture-bound, the NVIDIA part has the theoretical edge.

However, the GTX 580M is a 2011-era mobile part built on a 40 nm process with a 100 W TDP. It lacks Vulkan support and uses a form factor that ties it to specific laptops. The R7 350 is a 2016-era desktop card that runs at 55 W, supports Vulkan 1.2.170, and fits in a single PCIe slot. In the recorded OpenCL test, the R7 350's newer architecture translates its efficiency into a decisive win.

The production status of both cards is end-of-life, so this comparison is mainly relevant for used or legacy systems. In that context, the R7 350 is the better pick for general compute and modern API support. The GTX 580M makes sense only if you have a specific need for its higher texture throughput or memory bandwidth, or if you are constrained to an MXM-based laptop platform.

Where Each One Wins

AMD Radeon R7 350 wins on: OpenCL compute performance (7792 vs 6389, 22% ahead), pixel throughput (12.80 GPixel/s vs 9.920 GPixel/s), API support (DirectX 12 11_1 and Vulkan 1.2.170 vs DirectX 12 11_0 and no Vulkan), power efficiency (55 W vs 100 W TDP), and overall percentile ranking (40th vs 37th).

NVIDIA GeForce GTX 580M wins on: Raw FP32 throughput (952.3 GFLOPS vs 819.2 GFLOPS), memory bandwidth (96.00 GB/s vs 72.00 GB/s from its wider 256-bit bus), texture rate (39.68 GTexel/s vs 25.60 GTexel/s, helped by 64 TMUs), ROP count (32 vs 16), and the sheer transistor count (1,950 million vs 1,500 million, though spread over a much larger die).

A practical summary for a builder choosing between these two legacy parts: pick the R7 350 for everything that involves OpenCL compute, modern graphics APIs, lower power draw, and a desktop-friendly single-slot footprint. Pick the GTX 580M only if texture-heavy workloads dominate your usage and you can live with the MXM module limitation and the lack of Vulkan support.