NVIDIA GeForce GTX 670M vs NVIDIA GeForce GTX 750 Comparison

NVIDIA GeForce GTX 750 and NVIDIA GeForce GTX 670M represent two distinct generations of NVIDIA mobile and desktop graphics, separated by a 2012 versus 2014 release window and built on fundamentally different architectures. The data in this analysis draws exclusively from benchmark results, architectural specifications, and comparative metrics, revealing a clear performance hierarchy despite the GTX 670M's larger memory pool and wider memory bus. The GTX 750 emerges as the faster GPU in the single available head-to-head test, but the GTX 670M's unique mobile-oriented feature set and historical context offer a different kind of value for specific use cases.

Head-to-Head Benchmarks

The only direct benchmark comparison in the data is the Geekbench OpenCL test, where the NVIDIA GeForce GTX 750 scores 9315 points against the GTX 670M's 6513 points. This translates to a decisive 43% delta in favor of the GTX 750, a margin that is substantial and consistent with the architectural gulf between the two. In practical terms, the GTX 750 delivers nearly half again as much compute throughput in this OpenCL workload, which is a strong indicator for general-purpose GPU tasks and compute-heavy applications that leverage OpenCL acceleration.

Looking beyond the head-to-head, the GTX 750 also posts additional benchmark scores that contextualize its performance envelope. Its Geekbench Vulkan score reaches 8078, while its Geekbench Metal score is 4274, both of which are absent for the GTX 670M, which only has the single OpenCL result. The GTX 750's average benchmark score across all three tests is 7222, compared to the GTX 670M's average of 6513, reinforcing the overall performance advantage. The GTX 750's percentile ranking among all GPUs is 40, versus the GTX 670M's 38, a modest but meaningful placement difference in the broader competitive landscape.

The nearest rival data further illustrates where these cards sit. The GTX 750's closest competitor is the AMD Radeon Vega 8 Mobile, with an average score of 7203 and a delta of just 0.3%, meaning the GTX 750 is essentially tied with that integrated solution. It also edges out the NVIDIA GeForce GTX 560 SE (7171, 0.7% delta), the Intel Iris Pro Graphics P580 (7170, 0.7% delta), and surprisingly, the NVIDIA GeForce GTX 970 (7157, 0.9% delta) — though that last comparison likely reflects the GTX 970's own benchmark conditions rather than a true performance equivalence. For the GTX 670M, the nearest rival is the Intel UHD Graphics P750 at 6554, which actually scores 0.6% higher, meaning the GTX 670M is slightly behind that integrated part. It also compares closely to the NVIDIA GeForce GT 555M (6493, 0.3% delta), the NVIDIA Quadro M5000M (6481, 0.5% delta), and the AMD Radeon Vega 10 Mobile (6476, 0.6% delta).

The Verdict

From the data, the NVIDIA GeForce GTX 750 is the clear winner for anyone prioritizing raw compute performance. Its 43% OpenCL advantage over the GTX 670M is the single most telling statistic, and its higher average benchmark score (7222 versus 6513) and better percentile ranking (40 versus 38) confirm a consistent edge. The GTX 750 also benefits from a more modern feature set, including Vulkan support (1.4) and a higher boost clock, which the GTX 670M lacks entirely. For users running OpenCL-based workloads, video encoding, or modern game engines that leverage Vulkan, the GTX 750 is the only sensible choice based on the available metrics.

The GTX 670M, however, is not without its justifications. Its 1536 MB of GDDR5 memory exceeds the GTX 750's 1024 MB, and its 192-bit memory bus is wider, though its memory bandwidth is actually lower (72.00 GB/s versus 80.19 GB/s). It also has more texture mapping units (56 versus 32) and more ROPs (24 versus 16), which could theoretically benefit certain texture-heavy or fill-rate-bound scenarios. The GTX 670M was designed as a mobile MXM module, so for legacy laptop upgrades or repairs where compatibility with a specific MXM-B slot is required, it remains the only option. But from a pure performance-per-benchmark perspective, the GTX 750 wins every measurable category. The verdict is unambiguous: choose the GTX 750 for speed, choose the GTX 670M only if physical form factor or memory capacity constraints demand it.

Architecture Differences

The architectural divide between these two GPUs is stark, representing a generational leap in design philosophy. The GTX 750 uses the GM107 chip based on Maxwell architecture, fabricated on a 28 nm process at TSMC. In contrast, the GTX 670M uses the GF114 chip based on the older Fermi 2.0 architecture, fabricated on a 40 nm process at the same foundry. This process shrink alone explains much of the efficiency and performance difference — the Maxwell design achieves higher clock speeds and better compute efficiency per watt, while Fermi 2.0 is a bulkier, less efficient design.

Transistor counts tell a fascinating story. The GTX 750 packs 1,870 million transistors onto a die size of 148 mm², yielding a transistor density of 12.6 million per mm². The GTX 670M, by contrast, has 1,950 million transistors — slightly more in absolute terms — but spread across a massive 332 mm² die, resulting in a density of just 5.9 million per mm². This means the GTX 750 achieves nearly the same transistor count in less than half the silicon area, a direct consequence of the 28 nm process versus 40 nm. The Maxwell architecture's efficiency is further reflected in the shading unit configuration: the GTX 750 has 512 shading units, while the GTX 670M has only 336, despite having more transistors overall.

The clock behavior also diverges sharply. The GTX 750 has a base clock of 1020 MHz and a boost clock of 1085 MHz, delivering a peak FP32 throughput of 1,111.0 GFLOPS. The GTX 670M has no listed base or boost clock, which suggests a fixed or dynamically variable clock that is not documented, but its peak FP32 is just 803.7 GFLOPS. This 307.3 GFLOPS gap underscores how architecture efficiency, not just raw transistor count, drives performance. The GTX 750's pixel rate of 17.36 GPixel/s is more than double the GTX 670M's 8.372 GPixel/s, and its texture rate of 34.72 GTexel/s slightly exceeds the GTX 670M's 33.49 GTexel/s, despite having fewer TMUs.

Specification Differences

The two cards diverge across nearly every measurable specification, with only a few shared traits. Both use GDDR5 memory and support DirectX 12 (11_0) and OpenGL 4.6, but the GTX 750 adds Vulkan 1.4 support while the GTX 670M has no Vulkan entry. Memory configuration differs significantly: the GTX 750 has 1024 MB on a 128-bit bus with 80.19 GB/s bandwidth, while the GTX 670M has 1536 MB on a 192-bit bus with 72.00 GB/s bandwidth. The GTX 670M's larger capacity and wider bus do not compensate for the lower bandwidth, which is a direct result of its slower memory clock (750 MHz versus 1253 MHz).

Power and physical specifications also separate the two. The GTX 750 has a TDP of 55 W and is a single-slot card measuring 145 mm (5.7 inches) in length, with no power connectors and a suggested PSU of 250 W. It connects via PCIe 3.0 x16 and offers two DVI outputs plus one mini-HDMI 1.4a. The GTX 670M, as a mobile part, is an MXM-B (3.0) module with a TDP of 75 W, no suggested PSU, and display outputs that are "Portable Device Dependent." The GTX 750's lower TDP and smaller footprint make it far more adaptable for small-form-factor desktop builds, while the GTX 670M is locked into laptop or mobile workstation chassis. The GTX 750 also has a launch MSRP of 119 USD, while the GTX 670M has no listed MSRP, reflecting its OEM-oriented distribution.

FAQ

Q: Which GPU is faster in OpenCL benchmarks?

A: The NVIDIA GeForce GTX 750 scores 9315 in Geekbench OpenCL, versus 6513 for the GTX 670M, a 43% advantage. This is the only direct head-to-head benchmark available.

Q: Does the GTX 670M have more memory than the GTX 750?

A: Yes, the GTX 670M has 1536 MB of GDDR5 memory, while the GTX 750 has 1024 MB. However, the GTX 750 has higher memory bandwidth (80.19 GB/s versus 72.00 GB/s) due to its faster memory clock.

Q: Which card supports Vulkan?

A: Only the GTX 750 supports Vulkan, with version 1.4 listed. The GTX 670M has no Vulkan support in its API specifications.

Q: What is the TDP difference between the two cards?

A: The GTX 750 has a TDP of 55 W, while the GTX 670M has a TDP of 75 W. This makes the GTX 750 more power-efficient per unit of performance.

Q: How do their percentile rankings compare?

A: The GTX 750 ranks in the 40th percentile among all GPUs, while the GTX 670M ranks in the 38th percentile. This places the GTX 750 slightly higher in the overall performance distribution.

Q: Which card has more shading units?

A: The GTX 750 has 512 shading units, whereas the GTX 670M has 336. This is a significant architectural advantage for the GTX 750 despite the GTX 670M having slightly more transistors overall.

Where Each One Wins

The GTX 750 wins decisively in compute-heavy scenarios. Its 43% OpenCL lead, higher FP32 throughput (1,111.0 GFLOPS versus 803.7 GFLOPS), and double the pixel rate (17.36 GPixel/s versus 8.372 GPixel/s) make it the superior choice for any application that leverages general-purpose GPU computing, including video transcoding, scientific simulations, or machine learning inference. Its Vulkan 1.4 support also positions it for modern game engines and cross-platform graphics APIs, where the GTX 670M cannot participate at all. The GTX 750's lower TDP of 55 W and single-slot form factor mean it can fit into compact desktop cases with minimal power demands, making it a practical upgrade for legacy systems with limited PSU headroom.

The GTX 670M, conversely, wins only in niche scenarios related to its mobile heritage. Its 1536 MB memory capacity and 192-bit bus are the only specifications where it exceeds the GTX 750, which could matter for specific workloads that require larger framebuffers, such as high-resolution texture packs in older games or certain professional 3D rendering tasks. As an MXM-B module, it is the only choice for upgrading a laptop that specifically requires that interface, and its 24 ROPs versus the GTX 750's 16 could offer a slight edge in fill-rate-bound situations, though its lower overall pixel rate contradicts that advantage. For any user with a choice between the two, the GTX 750's benchmark superiority, modern architecture, and lower power draw make it the overwhelmingly better pick. The GTX 670M remains a relic of its era, best suited only for those constrained by a mobile chassis that demands its exact MXM form factor.