NVIDIA GeForce GTX 670MX vs NVIDIA GeForce MX130 Comparison

The NVIDIA GeForce GTX 670MX and the NVIDIA GeForce MX130 represent two very different eras of mobile graphics, yet their benchmark scores place them in surprisingly close proximity. The data shows a clear, if narrow, victory for the older Kepler-based part, with the GTX 670MX winning both head-to-head benchmark comparisons. In Geekbench OpenCL, the GTX 670MX scores 6125 against the MX130’s 6102, a marginal 0.4% lead. The gap widens significantly in the Vulkan test, where the GTX 670MX posts 5316 points versus the MX130’s 4914, an 8.2% advantage. Despite the MX130’s newer Maxwell architecture and much higher clock speeds, the GTX 670MX’s larger memory bus and greater raw compute resources carry it to victory, though the overall average benchmark scores remain close.

Head-to-Head Benchmarks

The most decisive result in this comparison comes from the Vulkan API workload. The GTX 670MX achieves a score of 5316, which is 8.2% higher than the MX130’s 4914. This is a substantial margin that suggests the older card handles low-level graphics API tasks with considerably more efficiency. The advantage likely stems from the GTX 670MX’s wider 192-bit memory interface, which provides 67.20 GB/s of bandwidth compared to the MX130’s 64-bit bus and 40.10 GB/s. In Vulkan, which often stresses memory throughput and draw-call processing, this bandwidth advantage proves decisive.

The OpenCL result tells a different story, however. Here the GTX 670MX scores 6125, edging out the MX130’s 6102 by just 0.4%. This near-tie indicates that in general-purpose compute workloads, the two GPUs are effectively interchangeable. The MX130’s higher boost clock of 1189 MHz (versus the GTX 670MX’s fixed 601 MHz) helps it close the gap against the older card’s superior core configuration. The GTX 670MX fields 960 shading units, 80 TMUs, and 24 ROPs, while the MX130 is limited to 384 shading units, 24 TMUs, and 8 ROPs. Yet the MX130’s clock advantage nearly compensates for this 2.5x deficit in shader count in OpenCL tasks.

When looking at the broader competitive landscape, both cards sit in similar performance tiers. The GTX 670MX’s average benchmark score of 5721 places it at the 33rd percentile of all GPUs, while the MX130’s 5508 average lands at the 32nd percentile. The GTX 670MX’s nearest rivals include the Intel Iris Pro Graphics P6300 (delta of 0.2%), the NVIDIA GeForce GTX 550 Ti (delta of -0.2%), and the AMD Radeon HD 8790M (delta of 0.5%). The MX130’s nearest rivals are the NVIDIA GeForce GTX 765M (delta of 0.1%), the AMD Radeon R7 M440 (delta of 0.5%), and the AMD FirePro M4000 (delta of -0.5%). These tight deltas confirm that neither card dominates its peer group; both sit squarely in the mid-range mobile graphics tier.

The texture and pixel throughput figures further illustrate the GTX 670MX’s theoretical advantages. It delivers 48.08 GTexel/s and 12.02 GPixel/s, compared to the MX130’s 28.54 GTexel/s and 9.512 GPixel/s. The GTX 670MX also produces 1,153.9 GFLOPS of FP32 compute, versus 913.2 GFLOPS for the MX130. These numbers suggest the older card should be faster in most scenarios, yet the real-world benchmark scores show a much closer contest, indicating that the MX130’s architecture extracts more efficiency per unit of raw compute.

The Verdict

The benchmark data points to the GTX 670MX as the stronger performer, but the margin is far from overwhelming. With two wins out of two head-to-head tests, the GTX 670MX is the clear victor in this comparison. The 8.2% lead in Vulkan is the most meaningful differentiator, as it represents a significant advantage in modern graphics workloads. The 0.4% OpenCL lead is essentially negligible and falls within typical run-to-run variance.

However, the MX130 is not without its own merits. It achieves this near-parity while consuming only 30 W of power, compared to the GTX 670MX’s 75 W TDP. The MX130 also benefits from a much more compact design, with an IGP slot width and no power connectors required, whereas the GTX 670MX is also listed as having no power connectors but carries a much higher power draw. For users prioritizing battery life and thermal efficiency in a thin-and-light laptop, the MX130’s 30 W power envelope is a substantial practical advantage.

The release dates tell a story of technological progression: the GTX 670MX launched on 2012-09-30, while the MX130 arrived much later on 2017-11-16. The MX130’s Maxwell architecture (chip GM108S) is a generation newer than the GTX 670MX’s Kepler (chip GK104), yet the older card still wins on raw performance. This suggests that NVIDIA’s product segmentation for the MX series prioritized power efficiency over outright speed.

Where Each One Wins

The GTX 670MX wins decisively in Vulkan-based applications and games. Its 8.2% advantage in the Vulkan benchmark indicates that titles leveraging this modern API will run noticeably smoother on the older card. The 192-bit memory bus and 67.20 GB/s bandwidth give it a clear edge in scenarios that demand high memory throughput, such as high-resolution textures and complex geometry. The GTX 670MX’s 3 GB of GDDR5 memory (versus the MX130’s 2 GB) also provides more headroom for memory-intensive workloads.

The MX130 wins in efficiency and thermal management. Its 30 W TDP is less than half of the GTX 670MX’s 75 W, making it far better suited for ultraportable laptops where cooling is limited. The MX130’s higher clock speeds (1109 MHz base, 1189 MHz boost) and newer Maxwell architecture deliver competitive compute performance per watt. Its 913.2 GFLOPS of FP32 performance from just 384 shading units demonstrates the architectural efficiency gains NVIDIA achieved between the Kepler and Maxwell generations.

For general compute workloads as measured by OpenCL, the two cards are effectively tied. The 0.4% difference between 6125 and 6102 is statistically insignificant. Users running OpenCL-accelerated applications should expect virtually identical performance from either card, making other factors like power consumption and driver support the deciding considerations.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 670MX has the higher average benchmark score at 5721, compared to the MX130’s 5508. This places the GTX 670MX at the 33rd percentile of all GPUs, while the MX130 sits at the 32nd percentile.

Q: How large is the performance gap in Vulkan workloads?

A: The GTX 670MX leads by 8.2% in the Geekbench Vulkan test, scoring 5316 against the MX130’s 4914. This is the largest performance difference between the two cards in any benchmark.

Q: Is the MX130 more power-efficient than the GTX 670MX?

A: Yes. The MX130 has a TDP of 30 W, while the GTX 670MX is rated at 75 W. The MX130 achieves this efficiency with a smaller die size of 77 mm² and 1,020 million transistors, compared to the GTX 670MX’s 294 mm² die and 3,540 million transistors.

Q: What memory configuration does each card use?

A: The GTX 670MX uses 3 GB of GDDR5 on a 192-bit bus, providing 67.20 GB/s of bandwidth. The MX130 uses 2 GB of GDDR5 on a 64-bit bus, providing 40.10 GB/s of bandwidth.

Q: How do these cards compare to their nearest rivals?

A: The GTX 670MX’s closest rival is the NVIDIA GeForce GTX 550 Ti, with a delta of -0.2%, meaning the GTX 670MX is 0.2% behind it. The MX130’s closest rival is the NVIDIA GeForce GTX 765M, with a delta of 0.1%, meaning the MX130 is 0.1% ahead.

Q: Which card has more shading units?

A: The GTX 670MX has 960 shading units, while the MX130 has 384. The GTX 670MX also has 80 TMUs and 24 ROPs, compared to the MX130’s 24 TMUs and 8 ROPs.

Architecture Differences

The GTX 670MX is built on NVIDIA’s Kepler architecture, utilizing the GK104 chip manufactured on a 28 nm process at TSMC. This is a large die measuring 294 mm² and containing 3,540 million transistors, resulting in a transistor density of 12.0M per mm². The Kepler architecture was designed to deliver high performance with improved power efficiency over its predecessor, and the GTX 670MX represents the higher end of the GeForce 600M generation.

The MX130, in contrast, uses the Maxwell architecture with the GM108S chip, also on a 28 nm process at TSMC. However, the die is dramatically smaller at just 77 mm², housing 1,020 million transistors. This yields a higher transistor density of 13.2M per mm², demonstrating the density improvements achieved with the newer architecture. The Maxwell architecture focuses on even greater power efficiency than Kepler, which is reflected in the MX130’s 30 W TDP.

The memory subsystems differ substantially. The GTX 670MX employs a 192-bit memory bus with 3 GB of GDDR5 running at 2.8 Gbps effective, delivering 67.20 GB/s of bandwidth. The MX130 uses a much narrower 64-bit bus with 2 GB of GDDR5 at 5 Gbps effective, achieving 40.10 GB/s. Despite the MX130’s faster memory clock, the GTX 670MX’s wider bus provides 67.7% more bandwidth.

Clock speeds tell the opposite story. The GTX 670MX runs at a fixed 601 MHz for both base and boost clocks, with memory at 700 MHz. The MX130 runs at 1109 MHz base and 1189 MHz boost, with memory at 1253 MHz. This nearly 2x clock advantage helps the MX130 compensate for its much smaller core configuration.

Both cards support DirectX 12 (11_0) and OpenGL 4.6, but their Vulkan support differs. The GTX 670MX supports Vulkan 1.2.175, while the MX130 supports Vulkan 1.4. The MX130 also has a narrower PCIe interface at PCIe 3.0 x4, compared to the GTX 670MX’s PCIe 3.0 x16, though this has minimal impact on mobile workloads. The production status for both is end-of-life, with the GTX 670MX succeeding the GeForce 500M series and preceding the GeForce 700M series, while the MX130 has no listed predecessor or successor.