NVIDIA GeForce GTX 650 Ti Boost vs NVIDIA GeForce GTX 675MX Comparison

The Verdict

The data splits these two Kepler-era NVIDIA cards along a clear line. The NVIDIA GeForce GTX 675MX wins every head-to-head benchmark recorded, taking both the Geekbench Metal and OpenCL tests. Its average benchmark score of 8427 sits 4.5% above the GTX 650 Ti Boost’s 8067. In percentile terms, the 675MX lands at the 43rd percentile of all GPUs, while the 650 Ti Boost sits just one point lower at the 42nd percentile. The 675MX is the pick for compute-oriented workloads and for anyone running Geekbench Metal or OpenCL tasks. The GTX 650 Ti Boost, however, is the only card in this pair with a Vulkan benchmark score (10041), making it the choice when Vulkan compute matters — the 675MX has no recorded Vulkan result in the data. For raw performance in the tests both share, the 675MX leads decisively; for a desktop card with a wider API footprint and a conventional PCIe form factor, the 650 Ti Boost holds its own.

Architecture Differences

Both GPUs are built on NVIDIA’s Kepler architecture, fabricated by TSMC on a 28 nm process. The similarities end there. The 675MX uses the GK104 chip, a 3,540 million-transistor die measuring 294 mm², with a transistor density of 12.0M per mm². The 650 Ti Boost uses the smaller GK106 chip, packing 2,540 million transistors into 221 mm², for a density of 11.5M per mm². That is a 1,000 million-transistor gap and a 73 mm² die-size difference, giving the 675MX a substantial silicon advantage.

The 675MX fields 960 shading units, 80 texture mapping units, and 32 ROPs. The 650 Ti Boost counters with 768 shading units, 64 TMUs, and 24 ROPs — a 192-shader, 16-TMU, and 8-ROP deficit. Clock speeds tell a different story. The 650 Ti Boost runs a 980 MHz base clock with a 1032 MHz boost, while the 675MX lists no base or boost clock in the data. Memory clocks also favor the desktop card: the 650 Ti Boost runs its GDDR5 at 1502 MHz (6 Gbps effective), versus the 675MX’s 900 MHz (3.6 Gbps effective). That clock advantage shows in bandwidth — the 650 Ti Boost delivers 144.2 GB/s across a 192-bit bus, while the 675MX manages 115.2 GB/s over a wider 256-bit bus. Both cards carry 2 GB of GDDR5.

The output configurations diverge sharply. The 675MX is an MXM Module with a bus interface of MXM-B (3.0), no power connectors, and display outputs described as “Portable Device Dependent.” The 650 Ti Boost is a dual-slot desktop card with a PCIe 3.0 x16 interface, one 6-pin power connector, a 300 W suggested PSU, and fixed outputs: 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. It also measures 241 mm (9.5 inches) in length. The 675MX draws 100 W TDP; the 650 Ti Boost draws 134 W.

Where Each One Wins

The 675MX wins outright in both shared benchmarks. In Geekbench Metal, it scores 6131 against the 650 Ti Boost’s 4858, a 26.2% advantage. In Geekbench OpenCL, it scores 10723 versus 9302, a 15.3% lead. That dominance tracks its larger chip: more shading units (960 vs 768), more TMUs (80 vs 64), more ROPs (32 vs 24), and a wider 256-bit memory bus. The 675MX also posts a higher average benchmark score overall (8427 vs 8067) and a stronger percentile rank (43 vs 42).

The 650 Ti Boost wins in areas the 675MX cannot contest. It has a Vulkan benchmark score of 10041; the 675MX has no Vulkan result recorded, meaning the desktop card is the only one of the two with evidence of Vulkan compute capability in this dataset. The 650 Ti Boost also offers higher memory bandwidth (144.2 GB/s vs 115.2 GB/s), a higher pixel rate (16.51 GPixel/s vs 13.08 GPixel/s), and a higher texture rate (66.05 GTexel/s vs 52.32 GTexel/s), all driven by its faster 980 MHz base and 1032 MHz boost clocks. Its FP32 throughput of 1.585 TFLOPS also exceeds the 675MX’s 1,255.7 GFLOPS — a 26% raw compute advantage on paper, even though the 675MX wins the actual compute benchmarks. The 650 Ti Boost’s desktop form factor, with fixed display outputs and a PCIe slot, makes it the practical choice for a stationary system.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 675MX, with an average score of 8427, compared to the GTX 650 Ti Boost’s 8067 — a 4.5% gap.

Q: Does the GTX 650 Ti Boost beat the 675MX in any benchmark?

A: In the recorded head-to-head tests, no. The 675MX wins both Geekbench Metal (6131 vs 4858) and Geekbench OpenCL (10723 vs 9302). However, the 650 Ti Boost has a Vulkan score of 10041, which the 675MX lacks entirely.

Q: How do their memory subsystems compare?

A: The 675MX uses a 256-bit bus with 2 GB GDDR5 at 900 MHz (3.6 Gbps effective), yielding 115.2 GB/s bandwidth. The 650 Ti Boost uses a narrower 192-bit bus with the same 2 GB GDDR5 but at 1502 MHz (6 Gbps effective), yielding 144.2 GB/s — 25% more bandwidth.

Q: What are the physical form factor differences?

A: The 675MX is an MXM Module with an MXM-B (3.0) interface and no power connectors, designed for portable devices. The 650 Ti Boost is a dual-slot desktop card, 241 mm long, using PCIe 3.0 x16, with one 6-pin power connector and a 300 W suggested PSU.

Q: Which card has more shading units?

A: The 675MX has 960 shading units, compared to 768 on the 650 Ti Boost — a 192-unit advantage. It also leads in TMUs (80 vs 64) and ROPs (32 vs 24).

Q: Do both cards support the same APIs?

A: Yes, both list DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175 in their API support. The difference is that the 650 Ti Boost has a recorded Vulkan benchmark score (10041), while the 675MX does not.

Head-to-Head Benchmarks

The single largest margin comes in Geekbench Metal. The 675MX scores 6131, the 650 Ti Boost 4858, and the delta is 26.2% in favor of the 675MX. That is a lead of 1273 points, more than a quarter of the runner-up’s total. The 675MX’s 960 shading units and 32 ROPs likely drive this gap, though the data does not break down per-test hardware utilization.

Geekbench OpenCL shows a smaller but still decisive margin. The 675MX scores 10723, the 650 Ti Boost 9302, for a 15.3% advantage — 1421 points. Notably, the 675MX’s OpenCL score is nearly double its Metal score, while the 650 Ti Boost’s OpenCL score is roughly 1.9x its Metal score. Both scale similarly, but the 675MX scales from a higher base.

The 650 Ti Boost’s only recorded win is outside the head-to-head set — its Geekbench Vulkan score of 10041. That number exceeds both cards’ Metal scores and comes close to the 675MX’s OpenCL result, but with no 675MX Vulkan score to compare, it cannot be called a head-to-head victory. Within the defined head-to-head benchmarks, the 675MX wins 2, and the 650 Ti Boost wins 0.

Specification Differences

| Specification | NVIDIA GeForce GTX 675MX | NVIDIA GeForce GTX 650 Ti Boost |

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

| Chip | GK104 | GK106 |

| Generation | GeForce 600M | GeForce 600 |

| Transistors | 3,540 million | 2,540 million |

| Die Size | 294 mm² | 221 mm² |

| Transistor Density | 12.0M / mm² | 11.5M / mm² |

| Base Clock | — | 980 MHz |

| Boost Clock | — | 1032 MHz |

| Memory Clock | 900 MHz (3.6 Gbps effective) | 1502 MHz (6 Gbps effective) |

| Memory Bus Width | 256 bit | 192 bit |

| Memory Bandwidth | 115.2 GB/s | 144.2 GB/s |

| Shading Units | 960 | 768 |

| TMUs | 80 | 64 |

| ROPs | 32 | 24 |

| Pixel Rate | 13.08 GPixel/s | 16.51 GPixel/s |

| Texture Rate | 52.32 GTexel/s | 66.05 GTexel/s |

| FP32 | 1,255.7 GFLOPS | 1.585 TFLOPS |

| TDP | 100 W | 134 W |

| Slot Width | MXM Module | Dual-slot |

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

| Suggested PSU | — | 300 W |

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

| Display Outputs | Portable Device Dependent | 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 |

| Length | — | 241 mm (9.5 inches) |

| Release Date | 2012-09-30 | 2013-03-25 |

| Launch MSRP | — | 169 USD |

| Predecessor | GeForce 500M | GeForce 500 |

| Successor | GeForce 700M | GeForce 700 |

| Avg Benchmark Score | 8427 | 8067 |

| Percentile vs All GPUs | 43 | 42 |

The 675MX leads in transistor count, die size, shading units, TMUs, ROPs, and memory bus width. The 650 Ti Boost leads in clock speeds, memory bandwidth, pixel rate, texture rate, FP32 throughput, and TDP. Both share 2 GB GDDR5, the 28 nm TSMC process, and identical API support. The 675MX released roughly six months earlier, and the 650 Ti Boost carries a launch MSRP of 169 USD. The data’s verdict is unambiguous: the 675MX wins both recorded benchmarks, but the 650 Ti Boost offers a desktop form factor, Vulkan evidence, and higher theoretical throughput rates.