NVIDIA GeForce GTX 675MX vs NVIDIA GeForce GTX 880M Comparison

# Head-to-Head Benchmarks

The benchmark data presents a split decision between these two Kepler-based mobile GPUs. Each card claims one decisive victory, and the margin of those victories tells a compelling story about their respective strengths.

In the Geekbench Metal test, the NVIDIA GeForce GTX 880M posts a score of 10,458, which is 41.4% higher than the GTX 675MX's 6,131. This is not a marginal difference; the GTX 880M nearly doubles the performance gap that separates the GTX 675MX from parity. The deltaPct of -41.4% (expressed from the GTX 675MX's perspective) indicates that the older card trails by a substantial margin in this API-specific workload. For context, the GTX 880M's Metal score alone places it in a different performance class, as its nearest rivals in the aggregate benchmark list — the NVIDIA Quadro P5000, GeForce GTX 650 Ti, GeForce GTX 650 Ti Boost, and GRID K2 — all cluster within a 0.5% band around its average score of 8,040.

The Geekbench OpenCL test flips the script entirely. Here, the GTX 675MX delivers 10,723 points against the GTX 880M's 5,622, a 90.7% advantage. That is nearly double the score, and it is the single largest margin of victory in either direction across the head-to-head data. The GTX 675MX's OpenCL result is so dominant that it drags its average benchmark score up to 8,427, which actually exceeds the GTX 880M's average of 8,040 despite the latter's much stronger Metal showing. This inversion — where one card wins one API test by 41.4% and the other wins the second test by 90.7% — suggests that architectural efficiency in specific compute paths, rather than raw hardware capability alone, drives these outcomes.

Looking at the aggregate percentiles, the GTX 675MX sits at the 43rd percentile among all GPUs, while the GTX 880M is at the 42nd. These are nearly identical positions, underscoring that the two cards occupy the same overall performance tier despite their divergent API-specific results. The average benchmark scores reflect this parity: 8,427 for the GTX 675MX versus 8,040 for the GTX 880M, a gap of roughly 4.8% in favor of the older part. When placed against their nearest rivals, the GTX 675MX's competition includes the AMD Radeon 880M (deltaPct -0.1%), NVIDIA GeForce MX330 (-0.4%), AMD Radeon HD 8870M (-0.4%), and AMD Radeon R9 M375X (+1.2%). The GTX 880M's rivals are all NVIDIA parts — Quadro P5000 (0%), GTX 650 Ti (-0.2%), GTX 650 Ti Boost (-0.3%), and GRID K2 (-0.5%) — indicating that the GTX 880M's aggregate score is tightly bracketed by a homogeneous group.

# Architecture Differences

Both GPUs share the same fundamental silicon: the GK104 chip built on TSMC's 28 nm process node, with 3,540 million transistors on a 294 mm² die. The transistor density is identical at 12.0M per mm². This means the underlying hardware capabilities are constrained by the same physical limits, and any performance divergence must stem from configuration choices rather than fabrication differences.

The GTX 675MX, from the GeForce 600M generation, ships with 960 shading units, 80 texture mapping units, and 32 ROPs. The GTX 880M, from the GeForce 800M generation, doubles down on compute resources with 1,536 shading units and 128 TMUs, while keeping the same 32 ROPs. That 60% increase in shading units and 60% increase in TMUs is the primary architectural differentiator. The GTX 880M's pixel rate of 31.78 GPixel/s is more than double the GTX 675MX's 13.08 GPixel/s, and its texture rate of 127.1 GTexel/s dwarfs the older card's 52.32 GTexel/s. FP32 compute follows the same pattern: 3.050 TFLOPS for the GTX 880M versus 1,255.7 GFLOPS for the GTX 675MX, a 143% uplift.

Clock speeds tell a complementary story. The GTX 880M runs at a 954 MHz base clock with a 993 MHz boost, while the GTX 675MX's base and boost clocks are not listed in the data pack. The memory clock also differs: the GTX 880M operates at 1250 MHz (5 Gbps effective), whereas the GTX 675MX runs at 900 MHz (3.6 Gbps effective). This feeds into memory bandwidth, where the GTX 880M achieves 160.0 GB/s against the GTX 675MX's 115.2 GB/s — a 39% advantage. Both use GDDR5 across a 256-bit bus, but the GTX 880M's higher clock rate extracts more throughput from the same interface width.

Memory capacity is a major split. The GTX 675MX comes with 2 GB of GDDR5, while the GTX 880M ships with 8 GB. That 4x capacity difference has implications for texture-heavy workloads and large frame buffers, even if the bus width remains constant. The power envelope also scales: the GTX 675MX is rated at 100 W TDP, while the GTX 880M draws 122 W. Both are MXM modules with MXM-B (3.0) interfaces and no power connectors, and both rely on portable-device-dependent display outputs.

Neither card features RT cores or tensor cores, as both predate those technologies. They share identical API support: DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The production status for both is end-of-life, with the GTX 675MX released on September 30, 2012, and the GTX 880M following on March 11, 2014. The GTX 675MX's predecessor is GeForce 500M and its successor is GeForce 700M; the GTX 880M's predecessor is GeForce 700M and its successor is GeForce 900M.

# FAQ

Q: Which GPU has the higher raw compute throughput?

A: The GTX 880M. Its FP32 rating of 3.050 TFLOPS is more than double the GTX 675MX's 1,255.7 GFLOPS, and its texture rate of 127.1 GTexel/s far exceeds the 52.32 GTexel/s of the older card.

Q: Why does the GTX 675MX win the OpenCL benchmark despite having fewer shading units?

A: The data shows the GTX 675MX scoring 10,723 in Geekbench OpenCL versus 5,622 for the GTX 880M, a 90.7% advantage. This suggests that the specific OpenCL workload in the test favors the GTX 675MX's configuration, even though the GTX 880M has 1,536 shading units against the GTX 675MX's 960.

Q: How much memory bandwidth difference is there between the two cards?

A: The GTX 880M delivers 160.0 GB/s, which is 39% higher than the GTX 675MX's 115.2 GB/s. Both use GDDR5 on a 256-bit bus, but the GTX 880M's 1250 MHz memory clock (5 Gbps effective) exceeds the GTX 675MX's 900 MHz clock (3.6 Gbps effective).

Q: Do these cards support the same graphics APIs?

A: Yes. Both support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. Neither has RT cores or tensor cores.

Q: Which GPU sits higher in the overall performance percentile?

A: The GTX 675MX is at the 43rd percentile among all GPUs, while the GTX 880M is at the 42nd. Their average benchmark scores are 8,427 and 8,040 respectively, making the GTX 675MX the slightly higher-ranked part in aggregate.

Q: What is the transistor count and die size for these chips?

A: Both use the GK104 chip with 3,540 million transistors on a 294 mm² die, fabricated by TSMC on a 28 nm process. The transistor density is 12.0M per mm² for both.

# Specification Differences

| Specification | GTX 675MX | GTX 880M |

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

| Generation | GeForce 600M | GeForce 800M |

| Base Clock | Not listed | 954 MHz |

| Boost Clock | Not listed | 993 MHz |

| Memory Clock | 900 MHz (3.6 Gbps effective) | 1250 MHz (5 Gbps effective) |

| Memory Size | 2 GB | 8 GB |

| Memory Bandwidth | 115.2 GB/s | 160.0 GB/s |

| Shading Units | 960 | 1,536 |

| TMUs | 80 | 128 |

| Pixel Rate | 13.08 GPixel/s | 31.78 GPixel/s |

| Texture Rate | 52.32 GTexel/s | 127.1 GTexel/s |

| FP32 | 1,255.7 GFLOPS | 3.050 TFLOPS |

| TDP | 100 W | 122 W |

| Release Date | 2012-09-30 | 2014-03-11 |

| Predecessor | GeForce 500M | GeForce 700M |

| Successor | GeForce 700M | GeForce 900M |

Identical specifications include the GK104 chip, 28 nm TSMC process, 3,540 million transistors, 294 mm² die size, 32 ROPs, 256-bit memory bus, GDDR5 memory type, MXM-B (3.0) bus interface, MXM Module slot width, no power connectors, portable-device-dependent display outputs, and the same API set (DirectX 12 (11_0), OpenGL 4.6, Vulkan 1.2.175). Neither card lists a launch MSRP.

# The Verdict

The data paints a nuanced picture. In raw hardware specifications, the GTX 880M is overwhelmingly superior: it has 60% more shading units, 60% more TMUs, 143% higher FP32 compute, 39% more memory bandwidth, and four times the memory capacity. Its pixel rate and texture rate are more than double those of the GTX 675MX. On paper, this should translate to a universal performance lead.

Yet the benchmark results contradict that expectation. The GTX 880M wins Geekbench Metal by 41.4%, but the GTX 675MX wins Geekbench OpenCL by 90.7%. The aggregate scores favor the GTX 675MX (8,427 versus 8,040), and its 43rd percentile ranking edges out the GTX 880M's 42nd. This means that for users whose workloads align with OpenCL compute paths, the older card is not just competitive — it is decisively faster. For Metal-centric tasks, the GTX 880M is the clear choice.

The release timeline matters here. The GTX 675MX arrived on September 30, 2012; the GTX 880M came on March 11, 2014. Both are end-of-life, but the GTX 880M is a newer part with a higher TDP (122 W versus 100 W) that leverages its additional hardware to dominate in one of the two benchmark categories. The GTX 675MX, despite being older and less capable on paper, demonstrates that driver maturity or workload-specific optimization can offset hardware deficits in certain compute scenarios.

# Where Each One Wins

The GTX 880M is the winner for Metal-based workloads. Its 10,458 Metal score versus the GTX 675MX's 6,131 is a 41.4% margin, and this aligns with its superior shading unit count and memory bandwidth. Applications that stress pixel throughput or rely heavily on texture operations will benefit from the GTX 880M's 31.78 GPixel/s pixel rate and 127.1 GTexel/s texture rate. The 8 GB frame buffer also provides headroom for large datasets or high-resolution textures that would exhaust the GTX 675MX's 2 GB allocation.

The GTX 675MX is the winner for OpenCL compute. Its 10,723 score against the GTX 880M's 5,622 represents a 90.7% advantage, and this result drives its higher average benchmark score of 8,427. For users running OpenCL-accelerated applications — scientific computing, certain rendering tasks, or general-purpose GPU compute — the GTX 675MX is the data-backed pick. Its lower TDP of 100 W also makes it a more power-efficient option for portable systems, though the benchmark results alone justify the choice without that consideration.

The tie in wins (1-1) and the near-identical percentiles (43rd versus 42nd) suggest that the selection hinges entirely on the target API. There is no universal winner here. The GTX 880M's hardware advantages are real and measurable, but they do not translate into across-the-board benchmark superiority. Conversely, the GTX 675MX's OpenCL dominance is so pronounced that it cannot be dismissed as a statistical fluke. For a buyer or system integrator deciding between these two end-of-life MXM modules, the decision rests on whether the primary workload is Metal-centric or OpenCL-centric. The data provides a clear answer for each scenario.