AMD Radeon 610M vs NVIDIA GeForce GTX 670MX Comparison

The NVIDIA GeForce GTX 670MX and AMD Radeon 610M are two very different mobile graphics solutions separated by a decade of silicon evolution. Benchmark data shows a 1-1 split in wins, with the older NVIDIA part dominating in OpenCL while the newer AMD integrated processor wins decisively in Vulkan. The GTX 670MX posts an average benchmark score of 5721 against the Radeon 610M’s 5444, a 5.1% margin, but the two chips occupy nearly identical percentile positions (33rd vs 32nd) among all GPUs. This is a matchup where raw compute throughput clashes with modern API efficiency.

Where Each One Wins

The NVIDIA GeForce GTX 670MX is the clear winner for compute-heavy, legacy-style workloads. Its OpenCL score of 6125 is 35.1% higher than the Radeon 610M’s 4535, a massive gap that reflects the GTX 670MX’s brute-force shader count and dedicated memory bandwidth. The data indicates this GPU is better suited for tasks that rely on traditional parallel processing, such as OpenCL-accelerated productivity apps or older game engines that were designed around high shader core counts and GDDR5 memory.

The AMD Radeon 610M takes the crown in Vulkan, a modern low-level API. Its Vulkan score of 6353 beats the GTX 670MX’s 5316 by 16.3%, showing that the RDNA 2.0 architecture’s newer feature set, including DirectX 12 Ultimate support, translates directly into better performance in contemporary graphics workloads. This is a win for newer games and Vulkan-native applications that can exploit the Radeon 610M’s more advanced pipeline design.

The split is clean: the GTX 670MX wins one benchmark by a wide margin, and the Radeon 610M wins the other by a solid but smaller margin. The GTX 670MX has the higher peak score (6125 vs 6353 is actually the AMD’s peak), but the NVIDIA part has the higher average across both tests, 5721 vs 5444. Users prioritizing raw OpenCL throughput should favor the GTX 670MX, while those targeting Vulkan performance should look at the Radeon 610M.

Architecture Differences

The two GPUs are built on fundamentally different architectures from different eras. The NVIDIA GeForce GTX 670MX uses the Kepler architecture with the GK104 chip, manufactured on a 28 nm process at TSMC. It packs 960 shading units, 80 texture mapping units, and 24 ROPs, driven by a 601 MHz base and boost clock. The chip contains 3,540 million transistors on a 294 mm² die, yielding a transistor density of 12.0M per mm².

The AMD Radeon 610M is a modern RDNA 2.0 part with the Mendocino chip, built on a 6 nm TSMC process. It has just 128 shading units, 8 TMUs, and 4 ROPs, but runs at much higher clocks: 1500 MHz base and 1900 MHz boost. The die size is 100 mm², and notably it includes 2 ray tracing cores, a feature completely absent from the Kepler-based NVIDIA part. The AMD chip also supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, while the GTX 670MX is limited to DirectX 12 (11_0) and Vulkan 1.2.175.

Memory systems are radically different. The GTX 670MX has 3 GB of dedicated GDDR5 on a 192-bit bus, delivering 67.20 GB/s of bandwidth and a 2.8 Gbps effective memory clock. The Radeon 610M uses system shared memory with a system-dependent bandwidth and a 15 W TDP, versus the GTX 670MX’s 75 W TDP. The power efficiency gap is enormous: the Radeon 610M delivers its Vulkan win at one-fifth the power draw, thanks to the 6 nm process and RDNA 2.0 design.

Head-to-Head Benchmarks

The Geekbench OpenCL result is the GTX 670MX’s statement win. It scores 6125 against the Radeon 610M’s 4535, a 35.1% advantage. This is consistent with the hardware: the NVIDIA part has 960 shading units and dedicated GDDR5 memory, giving it over 1,153.9 GFLOPS of FP32 throughput versus the AMD’s 486.4 GFLOPS. The GTX 670MX also produces 48.08 GTexel/s and 12.02 GPixel/s, compared to the Radeon 610M’s 15.20 GTexel/s and 7.600 GPixel/s. In raw fill-rate and compute terms, the older GPU is simply faster.

The Vulkan benchmark flips the script. The Radeon 610M scores 6353, beating the GTX 670MX’s 5316 by 16.3%. This is despite the AMD part having far fewer shading units and less raw FP32 power. The win comes from architecture: RDNA 2.0’s modern scheduling, the presence of 2 ray tracing cores, and full DirectX 12 Ultimate support allow it to leverage Vulkan’s low-level access far more effectively than Kepler’s older design. The Radeon 610M also supports FP16 at 972.8 GFLOPS (2:1 rate), which is absent in the GTX 670MX, a feature that modern game engines increasingly use.

The average benchmark scores tell a nuanced story. The GTX 670MX averages 5721, while the Radeon 610M averages 5444. In the nearest-rival comparison, the GTX 670MX sits within 0.2% of the Intel Iris Pro Graphics P6300 and 0.5% of the AMD Radeon HD 8790M, while the Radeon 610M is 0.4% behind the NVIDIA Quadro M4000 and 1% behind the GTX 765M. These deltas confirm that both GPUs are mid-pack performers, but the GTX 670MX’s higher average indicates better consistency across different test types.

The Verdict

From the data, the choice depends entirely on workload. The NVIDIA GeForce GTX 670MX is the pick for users running OpenCL-accelerated applications or older games that rely on high shader counts and dedicated VRAM. Its 35.1% OpenCL lead and higher average benchmark score (5721 vs 5444) make it the stronger all-around compute part, despite its 2012 release date and end-of-life status.

The AMD Radeon 610M is the modern choice for Vulkan-based gaming and DirectX 12 Ultimate titles. Its 16.3% Vulkan advantage, combined with 2 ray tracing cores and a 15 W TDP versus the GTX 670MX’s 75 W, makes it a far more efficient option for thin-and-light laptops. The Radeon 610M’s 6 nm process and 1900 MHz boost clock show that architectural advancements can overcome raw hardware deficits.

A strict score-based recommendation: if the benchmark suite is primarily OpenCL, the GTX 670MX wins. If Vulkan matters more, the Radeon 610M is the clear answer. The GTX 670MX has the higher peak compute and the better average, but the Radeon 610M’s modern API support and power efficiency make it the forward-looking choice. Neither GPU is a top performer — both sit at the 32nd-33rd percentile — but each serves a distinct niche. The GTX 670MX is the legacy compute workhorse; the Radeon 610M is the efficient modern renderer.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 670MX, with an average benchmark score of 5721, compared to the AMD Radeon 610M’s 5444, a 5.1% difference.

Q: How large is the OpenCL performance gap between the two?

A: The GTX 670MX scores 6125 in OpenCL versus the Radeon 610M’s 4535, a 35.1% advantage for the NVIDIA part.

Q: Does the AMD Radeon 610M beat the GTX 670MX in any benchmark?

A: Yes, in Geekbench Vulkan, the Radeon 610M scores 6353 against the GTX 670MX’s 5316, a 16.3% lead for the AMD part.

Q: What is the TDP difference between these two GPUs?

A: The GTX 670MX has a 75 W TDP, while the Radeon 610M is rated at just 15 W, making the AMD part significantly more power-efficient.

Q: Which GPU supports ray tracing?

A: The AMD Radeon 610M includes 2 ray tracing cores as part of its RDNA 2.0 architecture; the NVIDIA GTX 670MX has none.

Q: What DirectX versions do these GPUs support?

A: The GTX 670MX supports DirectX 12 (11_0) and Vulkan 1.2.175, while the Radeon 610M supports DirectX 12 Ultimate (12_2) and Vulkan 1.4.

Specification Differences

| Specification | NVIDIA GeForce GTX 670MX | AMD Radeon 610M |

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

| Architecture | Kepler | RDNA 2.0 |

| Process Node | 28 nm | 6 nm |

| Die Size | 294 mm² | 100 mm² |

| Transistors | 3,540 million | Not listed |

| Base Clock | 601 MHz | 1500 MHz |

| Boost Clock | 601 MHz | 1900 MHz |

| Memory Size | 3 GB | System Shared |

| Memory Type | GDDR5 | System Shared |

| Memory Bus Width | 192 bit | System Shared |

| Memory Bandwidth | 67.20 GB/s | System Dependent |

| Shading Units | 960 | 128 |

| TMUs | 80 | 8 |

| ROPs | 24 | 4 |

| Ray Tracing Cores | None | 2 |

| Pixel Rate | 12.02 GPixel/s | 7.600 GPixel/s |

| Texture Rate | 48.08 GTexel/s | 15.20 GTexel/s |

| FP32 Performance | 1,153.9 GFLOPS | 486.4 GFLOPS |

| FP16 Performance | Not listed | 972.8 GFLOPS (2:1) |

| TDP | 75 W | 15 W |

| Bus Interface | PCIe 3.0 x16 | PCIe 4.0 x8 |

| DirectX Support | 12 (11_0) | 12 Ultimate (12_2) |

| Vulkan Support | 1.2.175 | 1.4 |