NVIDIA GeForce GTX 675MX vs NVIDIA GeForce GTX 680M Comparison

The NVIDIA GeForce GTX 675MX and the NVIDIA GeForce GTX 680M are both mobile graphics solutions from the GeForce 600M generation, built on the same GK104 chip and the Kepler architecture. Despite sharing a common foundation, the recorded benchmark data reveals a distinct performance hierarchy between the two, with the GTX 675MX holding a consistent lead over the GTX 680M in both available tests. This analysis walks through the database records for these two end-of-life mobile GPUs, examining their measured scores, architectural specifications, and the practical implications for users deciding between them.

FAQ

Q: Which GPU has the higher average benchmark score in the database?

A: The NVIDIA GeForce GTX 675MX records an average benchmark score of 8427, while the NVIDIA GeForce GTX 680M records 7023. This places the GTX 675MX in the 43rd percentile of all GPUs, whereas the GTX 680M sits in the 39th percentile.

Q: How do the two GPUs compare in the Geekbench Metal test?

A: In the Geekbench Metal test, the GTX 675MX scores 6131, which is 27.3% higher than the GTX 680M’s score of 4815. The GTX 675MX is the clear winner in this test.

Q: What are the results for the Geekbench OpenCL test?

A: The GTX 675MX scores 10723 in Geekbench OpenCL, while the GTX 680M scores 9230. The GTX 675MX wins this test with a 16.2% advantage.

Q: Are there any tests where the GTX 680M outperforms the GTX 675MX?

A: No. The database records two head-to-head benchmark entries, and the GTX 675MX wins both. The win count stands at 2 for the GTX 675MX and 0 for the GTX 680M.

Q: What are the nearest rivals for each GPU based on average score?

A: For the GTX 675MX, the nearest rivals are the AMD Radeon 880M (average score 8436, 0.1% behind), the NVIDIA GeForce MX330 (8458, 0.4% behind), the AMD Radeon HD 8870M (8462, 0.4% behind), and the AMD Radeon R9 M375X (8325, 1.2% ahead). For the GTX 680M, the nearest rivals are the NVIDIA T600 (7035, 0.2% behind), the AMD Radeon R5 M240 (6975, 0.7% ahead), the NVIDIA GeForce GTX 675M (6946, 1.1% ahead), and the NVIDIA GeForce GTX 970 (7157, 1.9% behind).

Q: Do both GPUs share the same memory bandwidth?

A: Yes. Both the GTX 675MX and the GTX 680M have a 256-bit memory bus and a bandwidth of 115.2 GB/s. However, the GTX 680M has 4 GB of GDDR5 memory, while the GTX 675MX has 2 GB.

The Verdict

Based strictly on the recorded data, the NVIDIA GeForce GTX 675MX is the superior performer. Its average benchmark score of 8427 is 20.0% higher than the GTX 680M’s 7023, and it wins both head-to-head tests decisively. Users seeking the higher measured performance should select the GTX 675MX, as it delivers a 27.3% lead in Metal and a 16.2% lead in OpenCL.

The GTX 680M, however, is not without its own merits. It offers 4 GB of GDDR5 memory, double the capacity of the GTX 675MX’s 2 GB, which could be relevant for workloads that are sensitive to memory capacity rather than raw throughput. Additionally, the GTX 680M’s average score sits within 0.2% of the NVIDIA T600, indicating it is competitive with that professional-grade card in the database’s measurements.

For most users, the GTX 675MX is the clear pick given its higher scores in every recorded benchmark. The GTX 680M remains a viable option only if the larger memory pool is a priority, as the data shows no other advantage for it. The 20-percentage-point gap in average score is substantial and reflects a consistent performance lead across both test types.

Head-to-Head Benchmarks

The database includes two direct comparisons between the GTX 675MX and the GTX 680M, and the results are unambiguous. In the Geekbench Metal test, the GTX 675MX scores 6131 against the GTX 680M’s 4815, a difference of 27.3%. This is the larger of the two margins, indicating that the GTX 675MX has a particularly strong showing in Metal-accelerated workloads.

The Geekbench OpenCL test narrows the gap but still favors the GTX 675MX. Here, the GTX 675MX posts 10723, while the GTX 680M scores 9230, giving the GTX 675MX a 16.2% advantage. While the margin is smaller than in Metal, it remains a solid win that confirms the overall trend.

Looking at the broader context, the GTX 675MX’s average score of 8427 places it just 0.1% behind the AMD Radeon 880M, a modern integrated solution, and 1.2% ahead of the AMD Radeon R9 M375X. The GTX 680M, by contrast, sits 0.2% behind the NVIDIA T600 and 1.1% ahead of the NVIDIA GeForce GTX 675M. These relative positions reinforce the hierarchy: the GTX 675MX competes with a higher tier of GPUs, while the GTX 680M is closer to lower-ranked mobile parts.

The head-to-head results are consistent with the average scores. The GTX 675MX wins both tests, and its percentage leads are roughly in line with the 20.0% gap in average score. There is no recorded test where the GTX 680M pulls ahead, which makes the GTX 675MX the definitive choice in this comparison.

Specification Differences

The two GPUs differ in several key specifications, despite sharing the same chip, architecture, process node, and memory interface. The most notable difference is in the shading unit count. The GTX 680M has 1344 shading units, while the GTX 675MX has 960. This is a 40% difference in raw shader count, yet the benchmark results do not reflect an advantage for the GTX 680M.

Texture mapping units (TMUs) also differ. The GTX 680M has 112 TMUs, while the GTX 675MX has 80. This affects texture fill rates, with the GTX 680M recording 84.90 GTexel/s compared to the GTX 675MX’s 52.32 GTexel/s. The GTX 680M also has a higher pixel rate at 21.22 GPixel/s versus 13.08 GPixel/s for the GTX 675MX.

Clock speeds are another differentiator. The GTX 680M has a base clock of 719 MHz and a boost clock of 758 MHz, while the GTX 675MX has no recorded base or boost clock values. The memory clock is identical at 900 MHz (3.6 Gbps effective), and both share the same 256-bit bus and 115.2 GB/s bandwidth.

Memory capacity is a clear split: the GTX 680M comes with 4 GB of GDDR5, while the GTX 675MX has 2 GB. Floating-point performance also differs, with the GTX 680M rated at 2.038 TFLOPS and the GTX 675MX at 1,255.7 GFLOPS. Both have a TDP of 100 W, use the MXM Module slot width, have no power connectors, and use the MXM-B (3.0) bus interface.

Release dates differ as well. The GTX 680M was released on 2012-06-03, while the GTX 675MX followed on 2012-09-30. Both are end-of-life products, and both share the same predecessor (GeForce 500M) and successor (GeForce 700M). The transistor count (3,540 million), die size (294 mm²), and transistor density (12.0M / mm²) are identical for both.

Architecture Differences

Both GPUs are built on NVIDIA’s Kepler architecture, using the same GK104 chip manufactured by TSMC on a 28 nm process. The die size is 294 mm², and the transistor count is 3,540 million, yielding a transistor density of 12.0M / mm². These are identical figures, meaning the fundamental silicon is the same for both parts.

The architectural differences are therefore in the configuration of that silicon. The GTX 680M enables more of the chip’s resources, with 1344 shading units and 112 TMUs, versus 960 shading units and 80 TMUs for the GTX 675MX. Both have 32 ROPs, so the pixel output stage is equal, but the higher TMU count on the GTX 680M gives it a higher texture rate.

The GTX 680M also operates at higher clocks, with a base of 719 MHz and a boost of 758 MHz, while the GTX 675MX has no recorded clock figures in the database. This, combined with the larger shader and TMU counts, explains the GTX 680M’s higher theoretical throughput numbers, such as 2.038 TFLOPS and 84.90 GTexel/s.

Neither GPU has ray tracing cores or tensor cores, as these are Kepler-generation parts. Both support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The memory subsystem is identical in terms of bus width and bandwidth, but the GTX 680M doubles the capacity to 4 GB. There are no differences in power connectors, slot width, or display outputs, with both being portable-device dependent.

Where Each One Wins

The GTX 675MX wins in every recorded benchmark, making it the stronger choice for general performance. Its Geekbench Metal score of 6131 is 27.3% higher, and its OpenCL score of 10723 is 16.2% higher. This suggests that the GTX 675MX is better suited for compute-oriented tasks, whether they run through Metal or OpenCL APIs.

The GTX 680M’s only clear advantage is memory capacity. With 4 GB of GDDR5, it offers double the memory of the GTX 675MX’s 2 GB. For workloads that require large frame buffers or datasets, such as certain professional applications or high-resolution texture loading, the GTX 680M could have an edge, even though the benchmark data does not capture this directly.

In terms of raw throughput, the GTX 680M has higher theoretical numbers: 2.038 TFLOPS versus 1,255.7 GFLOPS, 84.90 GTexel/s versus 52.32 GTexel/s, and 21.22 GPixel/s versus 13.08 GPixel/s. These figures suggest that the GTX 680M should be faster on paper, but the recorded benchmark results show the opposite. This discrepancy highlights that real-world performance, as measured by Geekbench, does not align with the specification sheet.

For users prioritizing measured performance, the GTX 675MX is the winner without qualification. For users who need more memory and are willing to accept lower benchmark scores, the GTX 680M offers that specific capacity advantage. The data does not show any other scenario where the GTX 680M comes out ahead, so the choice ultimately rests on whether memory size or benchmark performance matters more.