AMD Radeon R7 M460 vs NVIDIA GeForce GTX 670M Comparison

The AMD Radeon R7 M460 and NVIDIA GeForce GTX 670M are two end-of-life mobile graphics processors separated by four years of architectural evolution. The benchmark data shows a statistical dead heat in raw compute performance, with the AMD part edging out a 1.5% lead in the single Geekbench OpenCL test. However, the underlying specifications reveal two fundamentally different design philosophies: the R7 M460 is a small, power-efficient chip built on a modern 28 nm process, while the GTX 670M is a large, older 40 nm die with a much wider memory bus. This analysis breaks down where each GPU wins, how their architectures diverge, and what the numbers actually mean for potential use cases.

Where Each One Wins

The head-to-head benchmark results are unambiguous: the AMD Radeon R7 M460 wins the only test conducted, scoring 6612 in Geekbench OpenCL against the GTX 670M’s 6513. The delta is a slim 1.5%, which places both cards in the same performance tier. The R7 M460’s win is narrow enough that it would be imperceptible in real-world workloads, but it does represent a consistent advantage in compute-heavy tasks that leverage OpenCL.

Where the GTX 670M would win is in memory-intensive scenarios, based on its specification sheet. The NVIDIA part has a 192-bit memory bus delivering 72.00 GB/s of bandwidth—exactly double the R7 M460’s 36.00 GB/s. For texture-heavy workloads, the GTX 670M also holds a commanding lead in texture fill rate at 33.49 GTexel/s versus the AMD part’s 24.58 GTexel/s. The pixel rates are nearly identical (8.372 GPixel/s vs 8.192 GPixel/s), so neither card dominates rasterization throughput.

The AMD card wins on architectural modernity and feature support. It supports DirectX 12 (12_0), while the GTX 670M is limited to DirectX 12 (11_0). The R7 M460 also supports Vulkan 1.2.170; the GTX 670M has no Vulkan support listed at all. For any software that leverages modern graphics APIs, the R7 M460 is the clear choice despite its raw compute parity.

Architecture Differences

The two GPUs come from different eras and design philosophies. The AMD Radeon R7 M460 uses the Meso chip built on GCN 3.0 architecture, fabricated on a 28 nm process at TSMC. It packs 1,550 million transistors into a 125 mm² die, achieving a transistor density of 12.4M per mm². The NVIDIA GeForce GTX 670M uses the GF114 chip based on the older Fermi 2.0 architecture, also fabricated by TSMC but on a 40 nm process. It contains 1,950 million transistors spread across a much larger 332 mm² die, yielding only 5.9M transistors per mm².

The transistor counts tell a story of different design goals. The GTX 670M uses its larger transistor budget for a wider memory interface and more texture units: it has 56 TMUs and 24 ROPs, compared to the R7 M460’s 24 TMUs and 8 ROPs. The AMD part has more shading units (384 vs 336) but fewer texture and pixel units. Memory configurations differ substantially: the R7 M460 has 2 GB of GDDR5 on a 64-bit bus, while the GTX 670M has 1536 MB of GDDR5 on a 192-bit bus.

Clock behavior also diverges. The R7 M460 has a defined base clock of 730 MHz and a boost clock of 1024 MHz, with memory running at 1125 MHz (4.5 Gbps effective). The GTX 670M lists no base or boost clock, but its memory runs at 750 MHz (3 Gbps effective). The AMD card’s higher memory clock partially compensates for its narrower bus, but the math still favors NVIDIA on bandwidth.

Power characteristics differ markedly. The GTX 670M has a TDP of 75 W and uses an MXM Module form factor with no power connectors. The R7 M460 lists no TDP or slot width, but its smaller die and newer process strongly suggest lower power draw. The GTX 670M’s bus interface is MXM-B (3.0), while the R7 M460 uses PCIe 3.0 x8—a sign that the AMD part is designed for more flexible system integration.

FAQ

Q: Which GPU has higher raw compute performance in OpenCL?

A: The AMD Radeon R7 M460 scores 6612 in Geekbench OpenCL, which is 1.5% higher than the GTX 670M’s 6513. This places the AMD card ahead by a margin that is technically measurable but practically negligible.

Q: Does the GTX 670M have any specification advantage over the R7 M460?

A: Yes. The GTX 670M has double the memory bandwidth (72.00 GB/s vs 36.00 GB/s), a wider 192-bit memory bus versus 64-bit, more TMUs (56 vs 24), more ROPs (24 vs 8), and a higher texture fill rate (33.49 GTexel/s vs 24.58 GTexel/s).

Q: Which GPU supports newer graphics APIs?

A: The AMD Radeon R7 M460 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170. The GTX 670M supports DirectX 12 (11_0) and OpenGL 4.6, but has no Vulkan support listed.

Q: How do these two cards compare to other GPUs in their respective performance brackets?

A: The R7 M460’s nearest rival is the NVIDIA GeForce GTX 670M itself, with a 1.5% delta. The GTX 670M’s closest competitor is the NVIDIA GeForce GT 555M, which scores 6493 (0.3% lower). Both cards sit at the 38th percentile of all GPUs.

Q: What is the transistor density difference between the two chips?

A: The R7 M460’s Meso chip achieves 12.4M transistors per mm² on a 28 nm process. The GTX 670M’s GF114 chip achieves 5.9M transistors per mm² on a 40 nm process—roughly half the density.

Q: Which GPU has more shading units?

A: The AMD Radeon R7 M460 has 384 shading units, while the NVIDIA GeForce GTX 670M has 336. Despite having fewer shading units, the GTX 670M achieves similar FP32 performance (803.7 GFLOPS vs 786.4 GFLOPS) due to its higher clock speeds.

Specification Differences

| Specification | AMD Radeon R7 M460 | NVIDIA GeForce GTX 670M |

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

| Architecture | GCN 3.0 | Fermi 2.0 |

| Process Node | 28 nm | 40 nm |

| Transistors | 1,550 million | 1,950 million |

| Die Size | 125 mm² | 332 mm² |

| Transistor Density | 12.4M / mm² | 5.9M / mm² |

| Memory Size | 2 GB | 1536 MB |

| Memory Bus Width | 64 bit | 192 bit |

| Memory Bandwidth | 36.00 GB/s | 72.00 GB/s |

| Shading Units | 384 | 336 |

| TMUs | 24 | 56 |

| ROPs | 8 | 24 |

| Pixel Rate | 8.192 GPixel/s | 8.372 GPixel/s |

| Texture Rate | 24.58 GTexel/s | 33.49 GTexel/s |

| FP32 Performance | 786.4 GFLOPS | 803.7 GFLOPS |

| TDP | Not listed | 75 W |

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

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

| Vulkan Support | 1.2.170 | Not listed |

| Release Date | 2016-05-14 | 2012-03-21 |

Head-to-Head Benchmarks

The single head-to-head benchmark, Geekbench OpenCL, produces a narrow victory for the AMD Radeon R7 M460. The AMD card scores 6612, while the GTX 670M scores 6513—a delta of 1.5%. This result is consistent with the FP32 compute figures, where the GTX 670M’s 803.7 GFLOPS edges out the R7 M460’s 786.4 GFLOPS, but the AMD card’s architectural efficiencies in GCN 3.0 apparently compensate for the raw throughput deficit.

The benchmark margin is so small that it falls within the noise of typical run-to-run variation. The R7 M460’s nearest rival list shows it beating the GTX 670M by 1.5%, but also losing to the NVIDIA GeForce GT 1010 by 1.3%. The GTX 670M’s nearest rival list shows it trailing the Intel UHD Graphics P750 by 0.6% and beating the AMD Radeon Vega 10 Mobile by 0.6%. Both cards occupy the same crowded performance band, where the difference between the best and worst performers in their peer group is less than 4%.

The real differentiators are not the benchmark scores but the feature sets. The R7 M460’s Vulkan 1.2.170 support and DirectX 12 (12_0) compliance make it the more future-proof choice for modern software titles, despite its memory bandwidth disadvantage. The GTX 670M’s double memory bandwidth and superior texture rate give it an edge in legacy workloads that are fill-rate bound. The data suggests that for compute tasks, the two cards are interchangeable; for memory-heavy rendering, the GTX 670M has a clear specification advantage; and for modern API adoption, the R7 M460 is the only option with full support.