NVIDIA GeForce 945M vs NVIDIA GeForce GTX 750 Comparison

Where Each One Wins

The recorded benchmark data splits cleanly between these two Maxwell-based NVIDIA parts. The GeForce GTX 750 takes the only direct head-to-head comparison available in the database, winning the Geekbench OpenCL test by a margin of 13.1% over the GeForce 945M. The 945M does not log a single benchmark victory in the head-to-head results, which makes the competitive picture straightforward: the desktop GTX 750 is the stronger compute performer in this pairing.

That does not mean the 945M is without a role. The database places the 945M at the 42nd percentile among all GPUs, while the GTX 750 sits at the 40th percentile. In percentile terms, the mobile part is actually slightly better positioned relative to the full GPU field, despite losing the direct comparison. This is because the 945M's average benchmark score of 8099 is built from a single OpenCL result, whereas the GTX 750's average of 7222 is dragged down by its additional Metal and Vulkan runs, which score much lower at 4274 and 8078 respectively.

For use-case planning, the GTX 750 is the choice when OpenCL compute throughput matters most. The 945M, by contrast, is a mobile MXM module with a lower raw compute score but a better overall percentile standing. A builder or system integrator working with MXM-based laptops has no real choice between these two cards, they occupy different physical slots and power envelopes. But for raw performance in a single benchmark metric, the GTX 750 wins outright.

Architecture Differences

Both GPUs are built on the same GM107 chip, using the Maxwell architecture, fabricated by TSMC on a 28 nm process. The transistor count is identical at 1,870 million, and the die size is the same at 148 mm², producing an identical transistor density of 12.6M per square millimeter. The fundamental silicon is therefore shared; the differences come from how NVIDIA configured and clocked that silicon for each market.

The 945M is a GeForce 900M series part, released in late 2015, while the GTX 750 belongs to the GeForce 700 series and launched earlier. The 945M uses 640 shading units, 40 texture mapping units, and 16 ROPs. The GTX 750 uses fewer resources in the shader and TMU departments: 512 shading units and 32 TMUs, but it keeps the same 16 ROPs. Despite having fewer shaders, the GTX 750 runs at higher clocks: a base of 1020 MHz and a boost of 1085 MHz, versus the 945M's 928 MHz base and 1020 MHz boost.

Memory configuration is where the two parts diverge sharply. The 945M comes with 2 GB of DDR3 on a 128-bit bus, running at an effective 1800 Mbps, yielding 28.80 GB/s of bandwidth. The GTX 750 has only 1024 MB of memory, but it is GDDR5 on the same 128-bit bus, running at an effective 5 Gbps, which produces 80.19 GB/s of bandwidth. That is a 2.8x bandwidth advantage for the GTX 750, a decisive factor in memory-bound workloads and a major reason its OpenCL score is higher.

The physical and platform differences are also notable. The 945M is an MXM Module with an MXM-B (3.0) bus interface, no power connectors, and a TDP of 75 W. Display outputs are listed as portable device dependent, meaning the laptop chassis determines connectivity. The GTX 750 is a single-slot desktop card, 145 mm or 5.7 inches long, with a PCIe 3.0 x16 interface, no power connectors, a 55 W TDP, and a suggested PSU of 250 W. It offers 2x DVI and 1x mini-HDMI 1.4a outputs. Both support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4, so API-level feature support is identical.

Head-to-Head Benchmarks

The database contains exactly one head-to-head benchmark between these two GPUs: Geekbench OpenCL. The GeForce GTX 750 scores 9315, while the GeForce 945M scores 8099. The delta is 13.1% in favor of the GTX 750, and the winner is recorded as the GTX 750. This is a substantial margin for two GPUs built on the same chip, and it illustrates how memory bandwidth and clock speed can outweigh raw shader count.

The 945M's higher shading unit count of 640 versus 512 does not translate into a win. The GTX 750 compensates with a 9.9% higher base clock and a 6.4% higher boost clock, plus the massive memory bandwidth advantage. In OpenCL workloads that stress memory throughput, the GDDR5-equipped GTX 750 simply has more data feeding its shaders. The 945M's DDR3 memory at 28.80 GB/s is a severe bottleneck relative to the 80.19 GB/s available on the GTX 750.

Looking at the broader database context, the 945M's single OpenCL score of 8099 puts it in a tight cluster with the NVIDIA GRID K2 at 8080, the AMD Radeon R9 M360 at 8129, the NVIDIA GeForce GTX 650 Ti Boost at 8067, and the NVIDIA GeForce GTX 950M at 8135. The 945M is within 0.4% of the R9 M360 and the GTX 950M, and it leads the GRID K2 by 0.2% and the GTX 650 Ti Boost by 0.4%. This is a razor-thin competitive band, meaning the 945M is essentially performance-equivalent to those four GPUs in the database's measurements.

The GTX 750's average benchmark score of 7222 places it near the AMD Radeon Vega 8 Mobile at 7203, the NVIDIA GeForce GTX 560 SE at 7171, the Intel Iris Pro Graphics P580 at 7170, and the NVIDIA GeForce GTX 970 at 7157. The GTX 750 leads each of these by margins of 0.3% to 0.9%. Interestingly, the GTX 970 appears in this rival list with a lower average score, but that is a consequence of the GTX 750's three benchmark runs being averaged against the GTX 970's own database results, not a statement about their real-world relative performance in gaming.

The Metal score of 4274 for the GTX 750 is notably lower than its OpenCL and Vulkan scores, which reflects the API's different workload characteristics on this architecture. The Vulkan score of 8078 is close to the 945M's OpenCL score of 8099, suggesting that under certain API conditions the two parts are nearly equivalent. But in the only directly comparable test, OpenCL on both, the GTX 750 is clearly ahead.

The Verdict

From the recorded data, the GeForce GTX 750 is the stronger GPU for OpenCL compute tasks. Its 13.1% lead over the 945M in the head-to-head benchmark is decisive, and its memory bandwidth of 80.19 GB/s versus 28.80 GB/s explains why. Anyone choosing between these two for a workload that relies on OpenCL should pick the GTX 750 without hesitation.

The 945M, however, is not without merit. Its percentile rank of 42 versus the GTX 750's 40 indicates that in the database's full GPU distribution, the mobile part is slightly better positioned. Its 2 GB of memory, double the GTX 750's 1024 MB, makes it more suitable for workloads or games that need larger frame buffers, even if the DDR3 bandwidth is limiting. For an MXM-based laptop, the 945M is the only option of the two, as the GTX 750 is a PCIe desktop card.

The GTX 750's higher clocks and GDDR5 memory make it the pick for raw throughput. The 945M's larger memory capacity and better percentile standing make it the pick for mobile systems and capacity-sensitive tasks. There is no scenario in the data where the 945M outperforms the GTX 750 in a direct comparison, so the verdict favors the desktop card for performance, with the mobile part serving a different physical and practical niche.

FAQ

Q: Which GPU wins the direct benchmark comparison?

A: The NVIDIA GeForce GTX 750 wins the only head-to-head test, Geekbench OpenCL, with a score of 9315 against the GeForce 945M's 8099, a margin of 13.1%.

Q: How much memory does each card have, and does it matter?

A: The 945M has 2 GB of DDR3, while the GTX 750 has 1024 MB of GDDR5. The GTX 750's memory bandwidth is 80.19 GB/s versus 28.80 GB/s for the 945M, which is a major factor in its benchmark victory.

Q: Are these GPUs built on the same architecture?

A: Yes, both use the GM107 chip with the Maxwell architecture on a 28 nm TSMC process, with 1,870 million transistors and a 148 mm² die size.

Q: What are the clock speeds of the two cards?

A: The GTX 750 runs at a base of 1020 MHz and a boost of 1085 MHz. The 945M runs at a base of 928 MHz and a boost of 1020 MHz.

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

A: The 945M is within 0.4% of the AMD Radeon R9 M360 and the NVIDIA GeForce GTX 950M, and it leads the NVIDIA GRID K2 by 0.2%. The GTX 750 leads the AMD Radeon Vega 8 Mobile by 0.3% and the NVIDIA GeForce GTX 560 SE by 0.7%.

Q: What is the power draw for each card?

A: The 945M has a TDP of 75 W, while the GTX 750 has a TDP of 55 W and a suggested PSU of 250 W. Neither card requires power connectors.

Specification Differences

| Specification | NVIDIA GeForce 945M | NVIDIA GeForce GTX 750 |

|---|---|---|

| Generation | GeForce 900M | GeForce 700 |

| Shading Units | 640 | 512 |

| TMUs | 40 | 32 |

| ROPs | 16 | 16 |

| Base Clock | 928 MHz | 1020 MHz |

| Boost Clock | 1020 MHz | 1085 MHz |

| Memory Size | 2 GB | 1024 MB |

| Memory Type | DDR3 | GDDR5 |

| Memory Clock | 900 MHz, 1800 Mbps effective | 1253 MHz, 5 Gbps effective |

| Memory Bandwidth | 28.80 GB/s | 80.19 GB/s |

| Pixel Rate | 16.32 GPixel/s | 17.36 GPixel/s |

| Texture Rate | 40.80 GTexel/s | 34.72 GTexel/s |

| FP32 | 1,305.6 GFLOPS | 1,111.0 GFLOPS |

| TDP | 75 W | 55 W |

| Slot Width | MXM Module | Single-slot |

| Bus Interface | MXM-B (3.0) | PCIe 3.0 x16 |

| Display Outputs | Portable Device Dependent | 2x DVI, 1x mini-HDMI 1.4a |

| Length | Not specified | 145 mm, 5.7 inches |

| Suggested PSU | Not specified | 250 W |

| Release Date | 2015-10-26 | 2014-02-17 |

| Launch MSRP | Not specified | 119 USD |

The identical items are the chip (GM107), architecture (Maxwell), process node (28 nm), foundry (TSMC), transistor count (1,870 million), die size (148 mm²), transistor density (12.6M / mm²), ROP count (16), API support (DirectX 12 (11_0), OpenGL 4.6, Vulkan 1.4), and power connector requirement (None). The differences above are the only fields where the two diverge in the database record.