NVIDIA GeForce GTX 750 Ti

The NVIDIA GeForce GTX 750 Ti is a Maxwell-generation GPU in the GeForce 700 family, manufactured by TSMC on a 28 nm process. The GM107 die integrates 1,870 million transistors across 148 mm², for a transistor density of 12.6M / mm², and it carries a launch MSRP of 149 USD. It is marked End-of-life, with the GeForce 600 as its predecessor and GeForce 900 as its successor. Its average benchmark score is 3016, placing it at the 18th percentile of all GPUs.

Benchmark Performance

The benchmark data places this card in a tightly grouped performance tier. Its average benchmark score of 3016 is only 1.1% above the GeForce 820A (2983), 1.4% above the GeForce GT 730M (2975), and 1.9% above the GeForce GTX 860M (2959). The one loss among its nearest rivals is to the GeForce GT 640, which scores 3082 and sits 2.1% higher. These margins are narrow enough that the GTX 750 Ti should be treated as a statistical peer of all four cards rather than a decisive winner or loser.

Peak compute rates are modest and consistent with that low-tier position. FP32 throughput is 1,388.8 GFLOPS, texture rate is 43.40 GTexel/s, and pixel rate is 17.36 GPixel/s. In PassMark, the highest API score is DirectX 9 at 60, followed by DirectX 11 at 26, DirectX 12 at 18, and DirectX 10 at 17. The 3DMark Steel Nomad DX12 result is 117. Geekbench results are more varied: OpenCL reaches 11512, Vulkan 10065, and Metal 5182. PassMark G3D is 3903, GPU compute is 1744, and G2D is 528. The API breakdown indicates that the card performs best with older DirectX workloads, while its DirectX 12 result is among its weakest scores.

Memory Subsystem

The memory configuration uses 2 GB of GDDR5 on a 128-bit bus. Memory clocks are 1350 MHz, or 5.4 Gbps effective, yielding a bandwidth of 86.40 GB/s. For high resolutions, the limiting factor is the combination of a small frame buffer and a narrow bus. High-resolution rendering needs both memory capacity and sustained transfer speed; at 2 GB and 86.40 GB/s, the card cannot support heavy high-resolution texture loads. The 128-bit bus keeps memory throughput constrained, even though the memory type is GDDR5.

Ray Tracing and Feature Set

The feature set is built around Maxwell’s conventional shader architecture. There are 640 shading units, 40 texture units, and 16 ROPs. Supported APIs are DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. There are no RT cores and no tensor cores, so hardware ray tracing is not supported. The DirectX 12 listing at the 11_0 feature level further limits modern API feature adoption, despite Vulkan 1.4 being present. The board connects over PCIe 3.0 x16 and offers 2x DVI and 1x mini-HDMI 1.4a outputs.

Who Should Consider It

The data points to a clear audience: users running applications whose demands match the card’s strongest benchmark results. DirectX 9 is the highest PassMark API score at 60, so legacy DirectX 9 titles are the most appropriate workload. The 2 GB GDDR5 frame buffer and 86.40 GB/s bandwidth mean that low resolutions and reduced texture settings are the practical envelope for 3D rendering. For desktop-oriented tasks, the PassMark G2D score of 528 is viable, but the overall 18th percentile position should temper expectations for demanding modern 3D applications. The 3DMark Steel Nomad DX12 score of 117 is low in absolute terms. In short, this is a card for low-demand, legacy-oriented use rather than current high-resolution gaming.

How It Compares

Against the GeForce 820A: The GTX 750 Ti leads with an average score of 3016 versus 2983, a 1.1% margin. The difference is minor, but the 750 Ti still ranks above that rival.

Against the GeForce GT 730M: The lead over the GT 730M is 1.4%, with the GT 730M scoring 2975. The 750 Ti holds a narrow but consistent advantage in this comparison.

Against the GeForce GTX 860M: The 750 Ti is 1.9% ahead of the GTX 860M’s 2959 average score. Despite the GTX 860M’s higher-tier name, the benchmark data puts the 750 Ti on top by less than two percentage points.

Against the GeForce GT 640: This is the only rival with a higher average score. The GeForce GT 640 reaches 3082, placing it 2.1% ahead of the 750 Ti. The margin is still small enough that the two cards effectively share the same performance class.

FAQ

Q: What is the average benchmark score of the GTX 750 Ti and where does it rank?

A: Its average benchmark score is 3016, which places it at the 18th percentile of all GPUs.

Q: What memory configuration does the GTX 750 Ti use?

A: It uses 2 GB of GDDR5 on a 128-bit bus with 86.40 GB/s bandwidth. The memory clock is 1350 MHz, or 5.4 Gbps effective.

Q: Does the GTX 750 Ti require auxiliary power connectors?

A: No. The power connector field is "None", the TDP is 60 W, and the suggested PSU is 250 W.

Q: Does the GTX 750 Ti support hardware ray tracing?

A: No. The RT core and tensor core fields are null, and its DirectX support is listed as 12 (11_0), which does not include hardware ray tracing.

Q: What APIs does the GTX 750 Ti support?

A: It supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4.

Q: Is the GTX 750 Ti still in production?

A: No, it is marked End-of-life. It was released in 2014, with the GeForce 600 as its predecessor and GeForce 900 as its successor.

Power and Cooling

The GTX 750 Ti is rated at a TDP of 60 W. The suggested PSU is 250 W, and no auxiliary power connectors are required. It is a single-slot board with a length of 145 mm, or 5.7 inches. This low-power, compact physical profile keeps cooling requirements modest and makes the card easy to fit into small systems.