AMD Radeon R7 M460 vs NVIDIA GeForce GTX 670MX Comparison

Head-to-Head Benchmarks

The recorded database contains a single shared benchmark between these two mobile graphics processors: GeekBench OpenCL. In that test, the AMD Radeon R7 M460 scores 6612 points, while the NVIDIA GeForce GTX 670MX scores 6125 points. This gives AMD an 8% victory in the only direct comparison available. The margin is meaningful but not overwhelming; it suggests a real performance gap in compute workloads, though the older NVIDIA part is not completely outclassed.

Context from the nearest rivals helps frame this result. The AMD R7 M460 sits at the 38th percentile of all GPUs in the database, with an average benchmark score of 6612. Its closest competitors include the AMD Radeon HD 7730M (6581 points, 0.5% behind), the Intel UHD Graphics P750 (6554 points, 0.9% behind), and the NVIDIA GeForce GT 1010 (6698 points, 1.3% ahead). The NVIDIA GeForce GTX 670M is also nearby, scoring 6513 points, which is 1.5% slower than the R7 M460. These tight margins indicate that the R7 M460 sits in a densely packed performance tier where small architectural differences decide the rankings.

The NVIDIA GeForce GTX 670MX, meanwhile, sits at the 33rd percentile of all GPUs, with an average benchmark score of 5721. Its average combines the OpenCL score of 6125 with a GeekBench Vulkan score of 5316, the latter dragging down its overall average. The nearest rivals for the GTX 670MX include the Intel Iris Pro Graphics P6300 (5712 points, 0.2% behind), the NVIDIA GeForce GTX 550 Ti (5731 points, 0.2% ahead), the AMD Radeon HD 8790M (5691 points, 0.5% behind), and the NVIDIA Quadro M500M (5604 points, 2.1% behind). The GTX 670MX is essentially neck-and-neck with these parts, none of which approach the R7 M460's OpenCL result.

The single head-to-head result shows the AMD part leading by 8%, but the broader data indicates that the GTX 670MX's Vulkan score of 5316 is significantly lower than its OpenCL score. This discrepancy suggests that the Kepler architecture may not handle modern API workloads as efficiently, even if raw OpenCL compute remains respectable. The R7 M460 has no recorded Vulkan score in the database, so a direct comparison on that API is not possible from the recorded data.

Architecture Differences

The two GPUs come from different architectural generations and design philosophies. The AMD Radeon R7 M460 uses the Meso chip based on GCN 3.0 architecture, fabricated on a 28 nm process at TSMC. The NVIDIA GeForce GTX 670MX uses the GK104 chip based on the Kepler architecture, also fabricated on a 28 nm process at TSMC. Both share the same process node and foundry, but their internal designs diverge sharply.

The transistor counts tell a striking story. The AMD chip packs 1,550 million transistors on a 125 mm² die, yielding a transistor density of 12.4 million per square millimeter. The NVIDIA chip contains 3,540 million transistors on a 294 mm² die, with a density of 12.0 million per square millimeter. The GK104 is more than twice the physical size and carries more than double the transistor budget, yet the density figures are nearly identical. This reflects the different scale of the designs: NVIDIA built a much larger GPU with more execution resources, while AMD opted for a smaller, more modestly equipped chip.

The execution resource counts reinforce this divide. The R7 M460 has 384 shading units, 24 texture mapping units, and 8 render output units. The GTX 670MX has 960 shading units, 80 TMUs, and 24 ROPs. That is 2.5 times more shading units, over 3 times more texture units, and 3 times more ROPs on the NVIDIA side. These raw resource differences should theoretically give NVIDIA a substantial edge in pixel throughput and texture-heavy workloads, yet the benchmark data shows AMD winning the OpenCL test. The R7 M460 achieves a pixel rate of 8.192 GPixel/s and a texture rate of 24.58 GTexel/s, while the GTX 670MX reaches 12.02 GPixel/s and 48.08 GTexel/s. NVIDIA's hardware advantages in these specific rates are clear, but they do not translate into a benchmark victory.

Clock speeds complicate the picture. The AMD part runs at a base clock of 730 MHz with a boost of 1024 MHz. The NVIDIA part runs at a fixed 601 MHz for both base and boost, with no dynamic overclocking headroom. The AMD boost clock is 70% higher than NVIDIA's fixed clock, which helps close the gap in raw compute throughput. The FP32 performance figures bear this out: the R7 M460 delivers 786.4 GFLOPS, while the GTX 670MX delivers 1,153.9 GFLOPS. NVIDIA still leads by a healthy margin in theoretical compute, but the 8% OpenCL win for AMD suggests that theoretical peak rates do not always predict real-world benchmark results.

Memory configurations also differ substantially. The AMD R7 M460 has 2 GB of GDDR5 memory on a 64-bit bus, with a memory clock of 1125 MHz (4.5 Gbps effective) and bandwidth of 36.00 GB/s. The NVIDIA GTX 670MX has 3 GB of GDDR5 memory on a 192-bit bus, with a memory clock of 700 MHz (2.8 Gbps effective) and bandwidth of 67.20 GB/s. NVIDIA's memory bandwidth is nearly double that of AMD, which should help in memory-intensive scenarios. The wider bus also gives NVIDIA an advantage in capacity, with 50% more VRAM.

The power envelope differs as well, though the database only records a TDP for the NVIDIA part: 75 W. The AMD part has no recorded TDP. The GTX 670MX also lists no power connectors, suggesting it draws power entirely through the motherboard, and its display outputs are marked as "Portable Device Dependent," indicating laptop-specific implementation.

FAQ

Q: Which GPU wins the only shared benchmark in the database?

A: The AMD Radeon R7 M460 wins the GeekBench OpenCL test with a score of 6612, compared to 6125 for the NVIDIA GeForce GTX 670MX, an 8% difference in AMD's favor.

Q: How does the GTX 670MX's Vulkan performance compare to its OpenCL performance?

A: The GTX 670MX scores 5316 in GeekBench Vulkan, which is notably lower than its OpenCL score of 6125. This difference pulls its average benchmark score down to 5721, and no Vulkan score is recorded for the R7 M460.

Q: What is the transistor density of each chip, and how do they compare?

A: The AMD Meso chip has a transistor density of 12.4 million per square millimeter (1,550 million transistors on 125 mm²), while the NVIDIA GK104 has a density of 12.0 million per square millimeter (3,540 million transistors on 294 mm²). The densities are nearly identical despite the NVIDIA chip being much larger.

Q: What memory specifications differ between the two GPUs?

A: The R7 M460 has 2 GB of GDDR5 on a 64-bit bus with 36.00 GB/s bandwidth, while the GTX 670MX has 3 GB of GDDR5 on a 192-bit bus with 67.20 GB/s bandwidth. NVIDIA's memory bandwidth is roughly 87% higher.

Q: Which GPU has a higher boost clock, and by how much?

A: The AMD R7 M460 has a boost clock of 1024 MHz, while the NVIDIA GTX 670MX is fixed at 601 MHz for both base and boost. AMD's boost clock is about 70% higher than NVIDIA's fixed clock.

Q: How do the nearest rivals of each GPU compare in average benchmark scores?

A: The R7 M460's nearest rival, the NVIDIA GeForce GT 1010, scores 6698, which is 1.3% ahead. The GTX 670MX's nearest rival, the NVIDIA GeForce GTX 550 Ti, scores 5731, which is 0.2% ahead. The R7 M460's rival tier is generally higher-scoring than the GTX 670MX's tier.

Specification Differences

The two parts differ across nearly every major specification category. The process node is the same (28 nm, TSMC), but the chip designs diverge immediately after that.

• Manufacturer and Architecture: AMD uses GCN 3.0 on the Meso chip; NVIDIA uses Kepler on the GK104 chip.
• Transistors: AMD has 1,550 million; NVIDIA has 3,540 million.
• Die Size: AMD measures 125 mm²; NVIDIA measures 294 mm².
• Transistor Density: AMD achieves 12.4M / mm²; NVIDIA achieves 12.0M / mm².
• Base Clock: AMD runs at 730 MHz; NVIDIA at 601 MHz.
• Boost Clock: AMD boosts to 1024 MHz; NVIDIA has no boost (fixed at 601 MHz).
• Memory Clock: AMD runs at 1125 MHz (4.5 Gbps effective); NVIDIA at 700 MHz (2.8 Gbps effective).
• Memory Size: AMD has 2 GB; NVIDIA has 3 GB.
• Memory Bus Width: AMD uses a 64-bit bus; NVIDIA uses a 192-bit bus.
• Memory Bandwidth: AMD delivers 36.00 GB/s; NVIDIA delivers 67.20 GB/s.
• Shading Units: AMD has 384; NVIDIA has 960.
• Texture Mapping Units: AMD has 24; NVIDIA has 80.
• Render Output Units: AMD has 8; NVIDIA has 24.
• Pixel Rate: AMD achieves 8.192 GPixel/s; NVIDIA achieves 12.02 GPixel/s.
• Texture Rate: AMD achieves 24.58 GTexel/s; NVIDIA achieves 48.08 GTexel/s.
• FP32 Performance: AMD delivers 786.4 GFLOPS; NVIDIA delivers 1,153.9 GFLOPS.
• FP16 Performance: AMD lists 786.4 GFLOPS (1:1); NVIDIA lists no FP16 figure.
• TDP: AMD has no recorded TDP; NVIDIA is rated at 75 W.
• Power Connectors: AMD has none recorded; NVIDIA lists "None."
• Bus Interface: AMD uses PCIe 3.0 x8; NVIDIA uses PCIe 3.0 x16.
• Display Outputs: AMD has none recorded; NVIDIA lists "Portable Device Dependent."
• DirectX Support: AMD supports DirectX 12 (12_0); NVIDIA supports DirectX 12 (11_0).
• Vulkan Support: AMD supports Vulkan 1.2.170; NVIDIA supports Vulkan 1.2.175.
• Release Date: AMD launched in May 2016; NVIDIA launched in September 2012.

Where Each One Wins

The AMD Radeon R7 M460 wins in the compute benchmark category. Its OpenCL score of 6612 beats the GTX 670MX by 8%, and its average benchmark score of 6612 places it at the 38th percentile, above the GTX 670MX's 33rd percentile. The R7 M460 also enjoys a significantly higher boost clock (1024 MHz vs. 601 MHz), which likely contributes to its benchmark advantage despite having fewer execution resources. For users running OpenCL compute workloads, the AMD part is the better choice according to the recorded data.

The NVIDIA GeForce GTX 670MX wins in raw hardware specifications. It has 2.5 times more shading units, 3.3 times more texture units, and 3 times more ROPs than the AMD part. Its FP32 throughput of 1,153.9 GFLOPS is about 47% higher than AMD's 786.4 GFLOPS. The GTX 670MX also offers 3 GB of VRAM on a 192-bit bus with 67.20 GB/s bandwidth, nearly double AMD's 36.00 GB/s. For scenarios that demand high memory bandwidth or large frame buffers, such as higher-resolution textures or multi-monitor setups, the NVIDIA part has the hardware headroom. Its pixel rate of 12.02 GPixel/s and texture rate of 48.08 GTexel/s also suggest strength in traditional rasterization workloads, even though the benchmark data does not directly confirm this advantage.

The GTX 670MX also has a recorded Vulkan score of 5316, which the R7 M460 lacks. This means the NVIDIA part has demonstrated performance in a modern API, even if that score is lower than its OpenCL result. The R7 M460 has no Vulkan data in the database, so its modern API performance is unknown. For users who prioritize Vulkan compatibility, the GTX 670MX at least has a measurable track record.

The R7 M460 wins in efficiency of execution per transistor. Despite having 2.3 billion fewer transistors, it outperforms the GTX 670MX in the only shared benchmark. This suggests the GCN 3.0 architecture extracts more useful compute from a smaller silicon budget, at least in the OpenCL workload tested.

The Verdict

The data presents a clear split between benchmark performance and raw specification strength. The AMD Radeon R7 M460 wins the only direct comparison, taking the GeekBench OpenCL test by 8% with a score of 6612 against 6125. Its higher percentile ranking (38th vs. 33rd) and higher average benchmark score (6612 vs. 5721) reinforce this result. For anyone choosing between these two based on recorded compute performance, the R7 M460 is the faster part.

The NVIDIA GeForce GTX 670MX, however, offers substantially more hardware capability. Its 960 shading units, 80 TMUs, and 24 ROPs dwarf the AMD part's 384, 24, and 8 respectively. The 3 GB VRAM and 192-bit memory bus provide 67.20 GB/s of bandwidth, nearly double the AMD's 36.00 GB/s. These specifications would typically suggest better performance in fill-rate-limited or memory-bandwidth-hungry scenarios, but the benchmark data does not capture those strengths. The GTX 670MX also holds a Vulkan score of 5316, giving it a measurable presence in modern API workloads that the R7 M460 lacks.

The release dates add another dimension. The AMD part launched in May 2016, nearly four years after the GTX 670MX's September 2012 debut. The newer AMD chip benefits from architectural refinements in GCN 3.0, while the Kepler-based NVIDIA part is older but larger and more fully featured. Both are marked end-of-life in the database.

The verdict depends on the use case. For compute-oriented tasks measured by OpenCL, the AMD Radeon R7 M460 is the winner. Its benchmark score is higher, its percentile ranking is better, and it achieves this with a smaller, less power-hungry design. For scenarios that demand memory bandwidth, large VRAM capacity, or raw fill rates, the NVIDIA GeForce GTX 670MX has the specifications that matter, even if the recorded benchmarks do not directly prove its superiority in those areas. The GTX 670MX also offers the only Vulkan data point, which may matter for compatibility with newer applications. Users should weigh the measured compute advantage of the R7 M460 against the unmeasured but substantial hardware resources of the GTX 670MX.