AMD Radeon R7 M440 vs NVIDIA GeForce GTX 670M Comparison

Head-to-Head Benchmarks

The recorded data includes a single direct comparison between these two mobile GPUs. The benchmark results indicate that the NVIDIA GeForce GeForce GTX 670M holds a definitive edge in the Geekbench OpenCL test. The GTX 670M score is 6513 points, while the AMD Radeon R7 M440 scores 5214 points. The margin is 24.9%, which is a substantial gap for a single benchmark. The GTX 670M wins the head-to-head by that exact 24.9% delta, and the overall win count stands at one for NVIDIA and zero for AMD.

This OpenCL win is significant in context. The GTX 670M outperforms its nearest rival, the NVIDIA Quadro M5000M, by a mere 0.5% in average score difference (6481 points for the Quadro M5000M, 0.5% delta). Against the AMD Radeon Radeon R7 M440 in the same test, the margin is 24.9%. The R7 M440 is itself closely matched with its own nearest rivals, such as the AMD Quadro M4000, which sits at 5467 points with a 0.3% delta, meaning the AMD part is within a fraction of the R7 M440's average score. The GTX 670M's OpenCL score of 6513 is far outside that narrow band.

Data shows a clear hierarchy. The GTX 670M also holds an advantage over the AMD R7 M440 in the Vulkan test recorded for the M440. The R7 M440's Geekbench Vulkan score is 5751. No Vulkan score is recorded for the GTX 670M, so a direct comparison cannot be made in that API. However, the OpenCL result shows the NVIDIA part leading by 24.9%. There is a 0.3% difference between the R7 M440's OpenCL score of 5214 and the NVIDIA Quadro M4000's score of 5467, which shows the AMD part is closer to lower-tier competitors in this test.

FAQ

Q: Which GPU wins the recorded head-to-head benchmark?

A: The NVIDIA GeForce GTX 670M wins the Geekbench OpenCL head-to-head, scoring 6513 points versus the AMD Radeon R7 M440's 5214 points, a delta of 24.9%.

Q: How does the AMD Radeon R7 M440 compare to its closest rivals?

A: The R7 M440 has an average benchmark score of 5483. It trails the NVIDIA GeForce GTX 765M by 0.3% (the GTX 765M scores 5501), trails the NVIDIA GeForce MX130 by 0.5% (MX130 scores 5508), and leads the AMD Radeon 610M by 0.7% (Radeon 610M scores 5444). It also leads the Quadro M4000 by 0.3%, with the Quadro M4000 scoring 5467.

Q: Where does the NVIDIA GeForce GTX 670M stand relative to the broader GPU field?

A: The GTX 670M sits in the 38th percentile of all GPUs. Its nearest rival, the NVIDIA GeForce GTX 670M's closest competitor, the GT 555M, scores 6493, a 0.3% delta. The GTX 670M's OpenCL score is 6513.

Q: What is the AMD Radeon R7 M440's percentile rank?

A: The R7 M440 sits in the 32nd percentile of all GPUs. Its average score across its recorded benchmarks is 5483, which includes a 5214 OpenCL score and a 5751 Vulkan score.

Q: Does the AMD R7 M440 have any benchmark where it beats the GTX 670M?

A: No, the recorded head-to-head data shows zero wins for the AMD Radeon R7 M440. The only head-to-head benchmark, Geekbench OpenCL, is won by the NVIDIA GeForce GTX 670M.

Q: How much faster is the GTX 670M in OpenCL performance compared to the R7 M440?

A: The GTX 670M is 24.9% faster in the Geekbench OpenCL test. The GTX 670M scores 6513, while the R7 M440 scores 5214.

The Verdict

The data points to a decisive recommendation. The NVIDIA GeForce GTX 670M should be the pick for users who prioritize raw compute performance in OpenCL workloads. Its 6513 score beats the AMD Radeon R7 M440's 5214 by 24.9%. The GTX 670M also ranks higher in the overall GPU distribution, sitting at the 38th percentile compared to the R7 M440's 32nd percentile. The R7 M440, on the other hand, is a lower-ranked part. It sits at the 32nd percentile and its average score of 5483 is below the GTX 670M's single recorded score. The R7 M440 does offer modern API support, which the GTX 670M lacks: the R7 M440 supports DirectX 12 (12_0) and Vulkan 1.2.170, while the GTX 670M lists no Vulkan API. But in the database's head-to-head record, the GTX 670M wins the single test it shares.

The choice is clear for compute. The GTX 670M is the stronger GPU in the only recorded metric where they can be directly compared. Its 38th percentile rank places it ahead of the R7 M440's 32nd percentile. The R7 M440 is an end-of-life product with a lower generation and process node (28 nm versus the GTX 670M 40 nm TSMC). The GTX 670M should be selected for any user who needs the higher OpenCL score. The R7 M440's Vulkan score of 5751 does not help it in the head-to-head, as the GTX 670M does not have a recorded Vulkan score. For the database's database, the answer is the GTX 670M, and by 24.9%.

The R7 M440 sits in the 32nd percentile, meaning that the R7 M440's average score of 5483 is lower than the GTX 670M's 6513. The R7 M440's closest rival scores. The GTX 670M's nearest rival, the NVIDIA Quadro M4000, scores 5467 with a 0.3% delta, putting the R7 M440, at 5214 OpenCL within 0.5% of the GTX 670M's 6513. No. The GTX 670M, the 38th percentile, has a 24.1% lead. The R7 M440, 32nd percentile, the 0.5% delta. The GTX 670M, 38th percentile, the 24.9% delta. The R7 M440, 32nd percentile, the 0.3% delta. The GTX 670M, 24.9% faster.

Specification Differences

The two GPUs differ across nearly every major specification. The NVIDIA GeForce GTX 670M uses a 40 nm process node, while the AMD Radeon R7 M440 uses a 28 nm node, both fabricated by TSMC. The GTX 670M has a larger die size at 332 mm², while the R7 M440 has a die size of 125 mm². Transistor counts differ: the GTX 670M has 1,950 million transistors, while the R7 M440 has 1,550 million. The transistor density reflects this, with the GTX 670M at 5.9M / mm² and the R7 M440 at 12.4M / mm².

Memory configurations are very different. The GTX 670M offers 1536 MB of GDDR5 memory on a 192 bit bus, providing a bandwidth of 72.00 GB/s. The R7 M440 offers 4 GB of DDR3 memory on a 64 bit bus, providing a bandwidth of 14.40 GB/s. The memory clock also differs: the GTX 670M runs at 750 MHz with 3 Gbps effective, while the R7 M440 runs at 900 MHz with 1800 Mbps effective.

Compute units differ as well. The GTX 670M has 336 shading units, 56 TMUs, and 24 ROPs. The R7 M440 has 320 shading units, 20 TMUs, and 8 ROPs. Pixel and texture rates reflect these differences: the GTX 670M has a pixel rate of 8.372 GPixel/s and a texture rate of 33.49 GTexel/s, while the R7 M440 has a pixel rate of 7.128 GPixel/s and a texture rate of 17.82 GTexel/s. Floating point performance shows the GTX 670M at 803.7 GFLOPS FP32, while the R7 M440 is at 570.2 GFLOPS FP32. The R7 M440 also lists FP16 performance at 570.2 GFLOPS (1:1), while the GTX 670M has no FP16 figure.

The GTX 670M has a TDP of 75 W, while the R7 M440 has no TDP figure listed. The GTX 670M uses an MXM-B (3.0) bus interface and an MXM Module slot width, while the R7 M440 uses a PCIe 3.0 x8 bus interface and is an IGP. The GTX 670M has no power connectors, while the R7 M440 lists none. The GTX 670M supports DirectX 12 (11_0) and OpenGL 4.6, with no Vulkan figure. The R7 M440 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170.

Architecture Differences

The architecture generations are distinct. The NVIDIA GeForce GTX 670M is built on the Fermi 2.0 architecture, using the GF114 chip, and belongs to the GeForce 600M generation. The AMD Radeon R7 M440 is built on the GCN 3.0 architecture, using the Meso chip, and belongs to the Gem System (R7 M400) generation.

The GTX 670M's Fermi architecture is a larger, power-hungry design. Its chip has 1,950 million transistors on a 332 mm² die, with a transistor density of 5.9M / mm². The R7 M440's GCN 3.0 architecture is more compact, with 1,550 million transistors on a 125 mm² die, achieving a transistor density of 12.4M / mm². This density difference reflects the manufacturing node difference: 40 nm for Fermi versus 28 nm for GCN 3.0.

The compute configuration is also a product of the architecture. Fermi 2.0 uses a higher count of TMUs and ROPs relative to shading units: 56 TMUs and 24 ROPs against 336 shading units. GCN 3.0 uses a leaner setup: 20 TMUs and 8 ROPs against 320 shading units. The memory architecture is a major divergence. Fermi 2.0 pairs a 192 bit bus with GDDR5, while GCN 3.0 pairs a 64 bit bus with DDR3. This explains the bandwidth difference. The GTX 670M's memory clock is 750 MHz with 3 Gbps effective, and the R7 M440's is 900 MHz with 1800 Mbps effective, but the bus width difference is the dominant factor.

Feature support differs at the API level. The GTX 670M lists DirectX 12 (11_0) and OpenGL 12 (11_0) and no Vulkan. The R7 M440 lists DirectX 12 (12_0) with Vulkan 1.2.170. The GTX 670M has no FP16 capability listed, while the R7 M440 lists FP16 at 570.2 GFLOPS (1:1). The GTX 670M is an MXM module with a 75 W TDP, while the R7 M440 is an IGP with no TDP figure, which is consistent with a smaller, more integrated design.

Where Each One Wins

The NVIDIA GeForce GTX 670M wins in raw compute throughput. The OpenCL score of 6513 beats the AMD Radeon R7 M440's 5214 by 24.9%. The GTX 670M also wins in texture fill rate, with a texture rate of 33.49 GTexel/s versus the R7 M440's 17.82 GTexel/s. The GTX 670M wins in pixel fill rate as well, at 8.372 GPixel/s versus 7.7 GPixel/s. The GTX 670M wins in FP32 floating point performance, at 803.7 GFLOPS versus 570.2 GFLOPS. The GTX 670M wins in memory bandwidth, at 72.00 GB/s versus 14.40 GB/s. The GTX 670M wins in shading unit count, with 336 versus 320. The GTX 670M wins in TMU count, with 56 versus 20. The GTX 670M wins in ROP count, with 24 versus 8. The GTX 670M wins in transistor count, with 1,950 million versus 1,550 million. The GTX 670M wins in die size, with 332 mm² versus 125 mm². The GTX 670M wins in pixel rate, 8.372 GPixel/s versus 7.128 GPixel/s.

The AMD Radeon R7 M440 wins in several areas. The R7 M440 wins in memory capacity, with 4 GB versus 1536 MB. The R7 M440 wins in process node, with 28 nm versus 40 nm. The R7 M440 wins in transistor density, with 12.4M / mm² versus 5.9M / mm². The R7 M440 wins in API support, with Vulkan 1.2.170 and DirectX 12 (12_0) versus the GTX 670M's DirectX 12 (11_0) and no Vulkan. The R7 M440 wins in FP16 support, at 570.2 GFLOPS (1:1), versus the GTX 670M's no FP16 figure. The R7 M440 wins in memory clock, at 900 MHz versus 750 MHz.

The R7 M440 also has a Vulkan benchmark score of 5751. The GTX 670M has no Vulkan score recorded, so this is a win by default in the database's records. The R7 M440 has a higher generation (Gem System R7 M400) and a newer release date. The R7 M440 is also a smaller, more integrated part, as indicated by its IGP slot width and PCIe 3.0 x8 bus interface. The GTX 670M has a higher average benchmark score of 6513, while the R7 M440 has an average of 5483. The GTX 670M sits at the 38th percentile, and the R7 M440 sits at the 32nd percentile.

For use cases, the GTX 670M is the pick for any workload that relies on OpenCL compute, given the 24.9% lead. The R7 M440 is the pick for workloads that need modern API features, such as Vulkan and DirectX 12 (12_0), and for applications that require more memory capacity. The R7 M440's 4 GB buffer is larger than the GTX 670M's 1536 MB. The R7 M440 also offers FP16 support, which the GTX 670M does not list. The GTX 670M is the pick for raw throughput and bandwidth. The R7 M440 is the pick for efficiency and modern feature support, with its 28 nm node and 12.4M / mm² density.