NVIDIA GeForce GTX 750 vs NVIDIA GeForce GTX 880M Comparison

Where Each One Wins

The benchmark data splits these two GPUs into very different usage profiles. The GeForce GTX 880M dominates in Metal workloads, while the GeForce GTX 750 takes OpenCL. That is not a small margin either way. In the geekbench_metal test, the 880M scores 10458 against the GTX 750's 4274, a 144.7% advantage. For applications that lean on Metal, the choice is clear: the 880M is in another class entirely.

The OpenCL results flip the script. The GTX 750 scores 9315, while the 880M manages only 5622. That is a 39.6% deficit for the larger mobile chip. OpenCL-heavy tasks, such as certain compute workloads or older GPGPU applications, favor the desktop card substantially. The GTX 750's advantage here is not marginal; it is a decisive lead that changes the recommendation depending on the software stack.

The overall averages tell a more nuanced story. The 880M has an average benchmark score of 8040, placing it in the 42nd percentile of all GPUs. The GTX 750 averages 7222, landing in the 40th percentile. The 880M holds a lead of roughly 11% in aggregate scores, but the per-test distribution means that lead is not universal. It wins one test by a landslide and loses the other by a wide margin.

For a builder choosing between these two, the workload determines the winner. Metal-based creative applications, video encoding pipelines, or Apple-centric compute tasks will favor the 880M. OpenCL-based scientific computing, distributed rendering, or legacy GPGPU tasks will favor the GTX 750. There is no all-around winner here; there are two specialized tools.

The nearest rival data adds context for each card's standing. The 880M sits within 0.5% of the NVIDIA Quadro P5000, the GeForce GTX 650 Ti, the GTX 650 Ti Boost, and the GRID K2, with deltas ranging from 0 to -0.5%. That places it in a tight cluster of mid-range performers. The GTX 750, meanwhile, trades blows with the AMD Radeon Vega 8 Mobile (0.3% ahead), the GeForce GTX 560 SE (0.7% ahead), the Intel Iris Pro Graphics P580 (0.7% ahead), and the GeForce GTX 970 (0.9% ahead). The 750 punches above its class in that company, particularly given that the GTX 970 is a much larger desktop card.

Architecture Differences

The two GPUs come from different NVIDIA architectures, and the design choices show up directly in the benchmark results. The GTX 880M uses the GK104 chip, built on the Kepler architecture. The GTX 750 uses the GM107 chip, built on Maxwell. Both are fabricated on a 28 nm process at TSMC, so the manufacturing node is identical. The differences come from the architectural approaches each generation took.

The 880M packs 3,540 million transistors on a 294 mm² die. The GTX 750 has 1,870 million transistors on a 148 mm² die. The transistor density is close: 12.0M per mm² for the 880M versus 12.6M per mm² for the 750. The 880M is simply a much larger chip, with nearly twice the transistors and roughly twice the die area. That size difference translates into substantially more execution resources.

The 880M has 1536 shading units, 128 texture mapping units, and 32 ROPs. The GTX 750 has 512 shading units, 32 TMUs, and 16 ROPs. The 880M has three times the shader count and four times the texture units. That hardware advantage explains its dominance in Metal benchmarks, where the extra shading and texture throughput can be fully utilized.

Memory configurations diverge sharply. The 880M carries 8 GB of GDDR5 on a 256-bit bus, yielding 160.0 GB/s of bandwidth. The GTX 750 has 1024 MB of GDDR5 on a 128-bit bus, yielding 80.19 GB/s. The 880M doubles the memory bus width and delivers twice the bandwidth. For large datasets or high-resolution textures, the 880M has a clear structural edge.

Clock speeds favor the smaller card. The GTX 750 runs at 1020 MHz base and 1085 MHz boost. The 880M runs at 954 MHz base and 993 MHz boost. The 750's higher clocks partially compensate for its fewer cores, but not enough to overcome the 880M's raw resource advantage in Metal. In OpenCL, the 750's higher clocks and Maxwell efficiency appear to matter more.

The compute rates reflect the resource split. The 880M achieves 3.050 TFLOPS of FP32 performance, 127.1 GTexel/s texture fill, and 31.78 GPixel/s pixel fill. The GTX 750 achieves 1,111.0 GFLOPS, 34.72 GTexel/s, and 17.36 GPixel/s. The 880M is roughly 2.7 times faster in FP32 and 3.7 times faster in texture rate. Yet OpenCL results show the 750 winning, which suggests the workload is sensitive to something other than raw throughput, possibly driver optimization or Maxwell's scheduling improvements.

Power and physical design differ completely. The 880M is an MXM Module with a 122 W TDP and no power connectors, designed for laptops. The GTX 750 is a single-slot desktop card with a 55 W TDP, no power connectors, and a 250 W suggested PSU. The 750 draws less than half the power of the 880M. The 880M uses an MXM-B (3.0) interface, while the 750 uses PCIe 3.0 x16.

Head-to-Head Benchmarks

The geekbench_metal test is the 880M's showcase. It scores 10458 against the GTX 750's 4274. The 144.7% delta means the 880M more than doubles the 750's Metal performance. In practical terms, any Metal-accelerated application, such as video transcoding or 3D rendering on compatible frameworks, will run dramatically faster on the 880M. This is the single largest performance gap in the comparison.

The geekbench_opencl test reverses the result. The GTX 750 scores 9315, while the 880M scores 5622. The delta is -39.6% from the 880M's perspective, meaning the 750 outperforms it by roughly two-thirds. For OpenCL compute tasks, the desktop card is the better choice despite having a third of the shading units and a quarter of the texture units. This is a striking result that highlights how differently the two architectures handle the same general-purpose compute API.

The average benchmark score aggregates these two tests. The 880M averages 8040, and the GTX 750 averages 7222. The 880M leads by 818 points, but that lead is entirely attributable to its Metal victory. If a workload uses OpenCL, the 750's advantage is larger in relative terms than the 880M's aggregate lead.

Looking at the nearest rivals for each card puts the scores in perspective. The 880M's 8040 average places it 0% from the Quadro P5000's 8039, -0.2% from the GTX 650 Ti's 8053, -0.3% from the GTX 650 Ti Boost's 8067, and -0.5% from the GRID K2's 8080. These are all within a rounding error of each other. The 880M is effectively tied with a professional workstation GPU and two older desktop cards.

The GTX 750's 7222 average places it 0.3% ahead of the Radeon Vega 8 Mobile's 7203, 0.7% ahead of the GTX 560 SE's 7171, 0.7% ahead of the Iris Pro Graphics P580's 7170, and 0.9% ahead of the GTX 970's 7157. The 750 outperforms a GTX 970 in this aggregate metric, which is notable given the 970's much larger die and higher power envelope. The 750's OpenCL strength drives that result.

The two benchmark types measure different things. Metal is Apple's low-level graphics and compute API, and the 880M's Kepler architecture clearly maps well to it. OpenCL is a cross-platform compute standard, and Maxwell's design appears to handle it more efficiently. The data shows a clean split with no overlap: the 880M wins Metal by a huge margin, the 750 wins OpenCL by a large margin.

The Verdict

Choose the GeForce GTX 880M if your primary workloads are Metal-based. The 144.7% lead in geekbench_metal is the single most decisive result in this comparison. It also offers 8 GB of memory, a 256-bit bus, and 160.0 GB/s of bandwidth, which matters for large textures or datasets. The 880M is a laptop-class MXM module with a 122 W TDP, so it suits mobile workstations that need strong graphics compute in a portable form factor.

Choose the GeForce GTX 750 if OpenCL is your primary API. Its 9315 OpenCL score beats the 880M by 39.6%, and it does so while drawing only 55 W and requiring a 250 W PSU. It is a single-slot desktop card with PCIe 3.0 x16 connectivity, making it trivial to install in almost any desktop systems. For compute workloads that lean on OpenCL, the 750 delivers more performance per watt and is easier to integrate into an existing system.

The aggregate scores favor the 880M at 8040 versus 7222, but that aggregate hides the bimodal distribution. The 880M is not a better GPU overall; it is better for Metal and worse for OpenCL. The GTX 750 is not a worse GPU; it is better for OpenCL and far worse for Metal. Match the card to the software.

For a mixed workload that uses both APIs, the 880M's aggregate lead is modest. The 42nd percentile versus the 40th percentile is a small gap. Neither card dominates the other comprehensively. The decision should rest on which API the target applications use, not on the average score.

FAQ

Q: Which GPU is faster in Metal benchmarks?

A: The GeForce GTX 880M. It scores 10458 in geekbench_metal, which is 144.7% higher than the GTX 750's 4274.

Q: Which GPU is faster in OpenCL benchmarks?

A: The GeForce GTX 750. It scores 9315 in geekbench_opencl, while the 880M scores 5622, a 39.6% deficit for the 880M.

Q: How much memory does each card have?

A: The GTX 880M has 8 GB of GDDR5 on a 256-bit bus. The GTX 750 has 1024 MB of GDDR5 on a 128-bit bus.

Q: What are the power requirements?

A: The GTX 880M has a 122 W TDP and uses an MXM Module form factor. The GTX 750 has a 55 W TDP and requires a 250 W suggested PSU.

Q: Which card has higher clock speeds?

A: The GTX 750 runs at 1020 MHz base and 1085 MHz boost. The GTX 880M runs at 954 MHz base and 993 MHz boost.

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

A: The 880M's average score of 8040 is within 0.5% of the Quadro P5000, GTX 650 Ti, GTX 650 Ti Boost, and GRID K2. The GTX 750's 7222 average is within 0.9% of the Radeon Vega 8 Mobile, GTX 560 SE, Iris Pro Graphics P580, and GTX 970.

Specification Differences

| Field | NVIDIA GeForce GTX 880M | NVIDIA GeForce GTX 750 |

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

| Chip | GK104 (Kepler) | GM107 (Maxwell) |

| Generation | GeForce 800M | GeForce 700 |

| Transistors | 3,540 million | 1,870 million |

| Die Size | 294 mm² | 148 mm² |

| Transistor Density | 12.0M / mm² | 12.6M / mm² |

| Base Clock | 954 MHz | 1020 MHz |

| Boost Clock | 993 MHz | 1085 MHz |

| Memory Size | 8 GB | 1024 MB |

| Memory Bus Width | 256 bit | 128 bit |

| Memory Bandwidth | 160.0 GB/s | 80.19 GB/s |

| Shading Units | 1536 | 512 |

| TMUs | 128 | 32 |

| ROPs | 32 | 16 |

| Pixel Rate | 31.78 GPixel/s | 17.36 GPixel/s |

| Texture Rate | 127.1 GTexel/s | 34.72 GTexel/s |

| FP32 | 3.050 TFLOPS | 1,111.0 GFLOPS |

| TDP | 122 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 |

| Vulkan | 1.2.175 | 1.4 |

| Release Date | 2014-03-11 | 2014-02-17 |

| Predecessor | GeForce 700M | GeForce 600 |

| Successor | GeForce 900M | GeForce 900 |

| Launch MSRP | None | 119 USD |