GPU Comparison

The NVIDIA GeForce MX110 and NVIDIA GeForce GT 740 represent two distinct approaches to discrete graphics from the same manufacturer, separated by three years of architectural evolution. The MX110 is a late-2017 mobile-class part built on Maxwell, while the GT 740 is a 2014 desktop card based on Kepler. Benchmark data shows a split decision: the MX110 wins one of two head-to-head tests, and the GT 740 wins the other, with the margin of victory differing substantially between them.

Head-to-Head Benchmarks

The two available head-to-head benchmark results paint a clear picture of complementary strengths. In Geekbench OpenCL, the MX110 scores 4255 against the GT 740’s 3847, giving the MX110 a 10.6% advantage. This is a decisive win for the newer chip, despite its smaller physical footprint and lower power envelope. The OpenCL test tends to reward memory bandwidth efficiency and compute throughput per watt, both areas where the Maxwell architecture in the MX110 shows its maturity.

The Geekbench Vulkan test flips the result, and it does so emphatically. The GT 740 scores 4083, while the MX110 manages only 3413. That is a 16.4% deficit for the MX110, meaning the GT 740 outperforms its younger rival by a wider margin in Vulkan than the MX110’s OpenCL lead. Vulkan is a low-level API that stresses raw geometry throughput and driver overhead handling; the GT 740’s larger memory bus and higher texture fill rate appear to matter more in this workload.

Looking at the wider benchmark context reinforces this split. The MX110’s average benchmark score across all tests is 3834, which places it at the 23rd percentile of all GPUs. The GT 740’s average is lower at 3431, putting it at the 21st percentile. However, the GT 740’s Vulkan score of 4083 is its strongest result, while the MX110’s OpenCL score of 4255 is its peak. Each card leads in the API where its architectural strengths are most relevant.

The deltaPct values in the head-to-head data are worth examining closely. The MX110’s 10.6% OpenCL win is solid but not overwhelming; the GT 740’s 16.4% Vulkan victory is a more pronounced gap. In practical terms, this means the MX110 is the better choice for compute-oriented OpenCL workloads, while the GT 740 holds a clear edge in Vulkan-based rendering. Neither card dominates the other across both tests, making the choice dependent on the specific application.

Architecture Differences

The two GPUs come from different architectural generations. The MX110 uses the GM108S chip built on the Maxwell architecture, while the GT 740 uses GK107 on the older Kepler design. Both are manufactured on a 28 nm process at TSMC, so the node is identical. The difference lies in how each architecture allocates its transistor budget.

The GT 740 packs 1,270 million transistors into a 118 mm² die, resulting in a transistor density of 10.8 million per square millimeter. The MX110 is smaller in both absolute and relative terms: 1,020 million transistors on a 77 mm² die, for a higher density of 13.2 million per square millimeter. This density advantage reflects Maxwell’s more efficient use of silicon per transistor, a hallmark of that generation’s design goals.

Shader resources differ significantly. The GT 740 has 384 shading units, 32 texture mapping units, and 16 raster operation pipelines. The MX110 counters with 256 shading units, 16 TMUs, and 8 ROPs. Despite having half the TMUs and ROPs, the MX110 achieves a slightly higher pixel rate of 8.048 GPixel/s versus the GT 740’s 7.944 GPixel/s. The texture rate tells a different story: the GT 740’s 31.78 GTexel/s more than doubles the MX110’s 16.10 GTexel/s.

Compute throughput follows the shader count. The GT 740 delivers 762.6 GFLOPS of FP32 performance, while the MX110 produces 515.1 GFLOPS. That is a 47.8% advantage for the GT 740 in raw floating-point math, which explains its Vulkan lead. However, the MX110’s Maxwell architecture is more efficient per clock, allowing it to close the gap in memory-bound tasks.

Memory subsystems are a major differentiator. The MX110 uses 2 GB of GDDR5 on a 64-bit bus, yielding 40.10 GB/s of bandwidth. The GT 740 has only 1024 MB of GDDR5 but on a 128-bit bus, doubling bandwidth to 80.19 GB/s. Both run memory at 1253 MHz (5 Gbps effective), so the entire bandwidth difference comes from bus width. The GT 740’s wider bus is a decisive factor in Vulkan’s data-heavy workload patterns.

Other structural differences include the bus interface. The MX110 connects via PCIe 3.0 x4, while the GT 740 uses PCIe 3.0 x16. The GT 740 also requires a single-slot cooler with a 6-pin power connector and a 250 W suggested PSU, whereas the MX110 is an integrated-style part with no power connectors and a 30 W TDP. The GT 740’s TDP is 64 W, more than double the MX110’s.

Where Each One Wins

The MX110 wins in OpenCL compute tasks, as shown by its 4255 score against the GT 740’s 3847. This makes it the better option for applications that leverage OpenCL for general-purpose GPU computing, such as physics simulations or image processing filters. Its higher transistor density suggests Maxwell’s scheduler handles divergent workloads more efficiently than Kepler’s.

The GT 740 dominates Vulkan rendering, posting 4083 versus the MX110’s 3413. This advantage stems from its 128-bit memory bus providing 80.19 GB/s of bandwidth, exactly double the MX110’s 40.10 GB/s. Vulkan’s explicit memory management rewards this bandwidth headroom, allowing the GT 740 to feed its 384 shading units more effectively. The GT 740 also has double the TMUs and ROPs, which helps in geometry-heavy scenes.

For legacy DirectX 12 (11_0) workloads, both cards support the same feature level, so neither has an API advantage there. However, the GT 740 offers Vulkan 1.2.175 support versus the MX110’s Vulkan 1.4, giving the newer MX110 a higher Vulkan API version. In practice, the GT 740 still wins the Vulkan benchmark, so the MX110’s newer API revision does not translate to better performance in this specific test.

Power-constrained environments favor the MX110. Its 30 W TDP and lack of power connectors make it suitable for systems without auxiliary PCIe power. The GT 740’s 64 W TDP and 6-pin connector require a more robust power delivery setup. For compact or low-power builds, the MX110 is the practical choice, provided the workload is OpenCL-oriented.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The MX110 has an average benchmark score of 3834, while the GT 740 averages 3431. The MX110 also ranks at the 23rd percentile of all GPUs, versus the GT 740’s 21st percentile.

Q: How do the two cards compare in memory bandwidth?

A: The GT 740 provides 80.19 GB/s of bandwidth via a 128-bit bus, which is exactly double the MX110’s 40.10 GB/s from a 64-bit bus. Both use GDDR5 memory running at 1253 MHz (5 Gbps effective).

Q: What explains the GT 740’s Vulkan victory?

A: The GT 740 scores 4083 in Geekbench Vulkan versus the MX110’s 3413, a 16.4% difference. Its 128-bit memory bus and double the TMUs (32 vs 16) and ROPs (16 vs 8) give it a structural advantage in Vulkan’s bandwidth-intensive rendering path.

Q: Is the MX110 faster in any benchmark?

A: Yes, the MX110 wins Geekbench OpenCL with 4255 against the GT 740’s 3847, a 10.6% margin. Its Maxwell architecture achieves higher transistor density (13.2M/mm² vs 10.8M/mm²) and a slightly better pixel rate.

Q: What are the power requirements for each card?

A: The MX110 has a 30 W TDP and requires no power connectors. The GT 740 has a 64 W TDP, needs a single 6-pin power connector, and carries a suggested PSU rating of 250 W.

Q: How do their nearest rivals compare?

A: The MX110’s closest rival is the GTX 650 with a 0.3% deltaPct, while the GT 740’s nearest is the Quadro 2000M at -0.1% deltaPct. The GT 740 also trails the GeForce 920MX by 2.7% and the HD Graphics 530 by 3%.

Specification Differences

| Specification | MX110 | GT 740 |

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

| Architecture | Maxwell | Kepler |

| Chip | GM108S | GK107 |

| Transistors | 1,020 million | 1,270 million |

| Die Size | 77 mm² | 118 mm² |

| Transistor Density | 13.2M / mm² | 10.8M / mm² |

| Memory Size | 2 GB | 1024 MB |

| Memory Bus Width | 64 bit | 128 bit |

| Memory Bandwidth | 40.10 GB/s | 80.19 GB/s |

| Shading Units | 256 | 384 |

| TMUs | 16 | 32 |

| ROPs | 8 | 16 |

| Pixel Rate | 8.048 GPixel/s | 7.944 GPixel/s |

| Texture Rate | 16.10 GTexel/s | 31.78 GTexel/s |

| FP32 | 515.1 GFLOPS | 762.6 GFLOPS |

| TDP | 30 W | 64 W |

| Slot Width | IGP | Single-slot |

| Power Connectors | None | 1x 6-pin |

| Suggested PSU | None | 250 W |

| Bus Interface | PCIe 3.0 x4 | PCIe 3.0 x16 |

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

| Vulkan Version | 1.4 | 1.2.175 |

| Length | None | 145 mm (5.7 inches) |

| Release Date | 2017-11-16 | 2014-05-28 |

| Launch MSRP | None | 89 USD |

The Verdict

The data supports a straightforward split based on workload. For OpenCL compute tasks, the MX110 is the better performer, delivering 4255 points versus the GT 740’s 3847. Its 10.6% lead in that test, combined with a 30 W TDP and no power connector requirement, makes it suitable for low-power systems where OpenCL acceleration matters more than raw rendering throughput.

For Vulkan-based rendering, the GT 740 is the clear choice. Its 4083 score beats the MX110 by 16.4%, and its 80.19 GB/s memory bandwidth provides double the data throughput of the MX110. The GT 740 also offers more shading units (384 vs 256), higher FP32 performance (762.6 GFLOPS vs 515.1 GFLOPS), and double the TMUs and ROPs. These hardware advantages make it the stronger GPU for modern graphics APIs, despite its older Kepler architecture and higher 64 W power draw.

Users with a desktop system that can supply a 6-pin power connector and 250 W PSU should favor the GT 740 for gaming or Vulkan applications. Users needing a low-power, connector-free solution for OpenCL compute in a compact or portable chassis should select the MX110. The GT 740’s higher raw specs do not translate to overall dominance because the MX110’s Maxwell architecture proves more efficient in compute-oriented tests. Ultimately, the GT 740 wins on rendering performance, while the MX110 wins on efficiency and OpenCL throughput.