NVIDIA GeForce GTX 960A

The NVIDIA GeForce GTX 960A is an end-of-life mobile graphics module in the GeForce 900A generation, built on the GM107 chip and Maxwell architecture, and released on 2015-03-12. Its Geekbench OpenCL score is 12002, which places it at the 50th percentile of all GPUs in the database. Its nearest rivals are tightly packed: the GTX 960A trails the GeForce GTX 960 by 0.4%, sits 0.1% behind the GeForce GTX 670, leads the Radeon Pro 5500M by 0.1%, and is 0.9% ahead of the Radeon RX 6500 XT. This is a mid-tier part whose overall position is defined by margins that top out at 0.9%.

Benchmark Performance

The only benchmark result stored for the GTX 960A is Geekbench OpenCL, where it scores 12002. The strongest listed rival is the GeForce GTX 960, with an average score of 12045; that is 0.4% higher than the 960A. The GeForce GTX 670 scores 12014, putting the 960A 0.1% behind. The AMD Radeon Pro 5500M scores 11986, and the AMD Radeon RX 6500 XT scores 11897, placing the 960A 0.1% and 0.9% ahead respectively. Across the entire rival group, the largest gap in either direction is only 0.9%, so the 960A is not separated from its nearest competitors by any meaningful score gap.

The hardware behind that score is a 28 nm TSMC chip with 1,870 million transistors on a 148 mm² die, at a transistor density of 12.6M / mm². The GPU contains 640 shading units, 40 texture units, and 16 ROPs. Clocks are listed at 1097 MHz base and 1176 MHz boost, producing a pixel rate of 18.82 GPixel/s, a texture rate of 47.04 GTexel/s, and 1.505 TFLOPS of FP32 compute. These numbers explain the performance band: a small Maxwell chip with enough shading capacity to match the mid-range mobile competition, but without enough raw throughput to climb into the upper tier of the database. The 50th-percentile ranking matches that interpretation exactly.

Who Should Consider It

The GTX 960A is aimed at portable systems that need moderate 3D performance in a compact MXM form factor. Its slot width is listed as MXM Module, and its bus interface is MXM-B (3.0), so it belongs in a laptop or mobile workstation built around that standard rather than in a desktop tower. Display outputs are listed as Portable Device Dependent, meaning the host system controls the actual display connection; the GPU has no native fixed outputs of its own.

The 2 GB GDDR5 frame buffer and 80.19 GB/s bandwidth make this a card for restrained graphical settings. Users should expect to lower texture quality and shadow detail in demanding scenes, because the 128-bit memory bus and limited VRAM capacity will not sustain high-detail asset loads for long. The 50th-percentile OpenCL score reinforces that guidance: this is a middle-of-the-pack mobile part, not a high-end solution. It is most appropriate for users maintaining an older MXM-based system who need a functional replacement or a modest upgrade within the same power envelope.

How It Compares

NVIDIA GeForce GTX 670. The GeForce GTX 670 averages 12014, only 0.1% above the GTX 960A. In the recorded Geekbench OpenCL data, the two are effectively level; neither card holds a meaningful advantage.

AMD Radeon Pro 5500M. The Radeon Pro 5500M averages 11986, which puts the GTX 960A 0.1% ahead. The margin is so small that the two should be treated as equivalent in this benchmark, despite coming from different vendors and product lines.

NVIDIA GeForce GTX 960. The GeForce GTX 960 averages 12045, 0.4% above the GTX 960A. This is the largest deficit the 960A faces among its listed rivals, but it is still a narrow gap. The 960A is essentially a near-identical performer in this database's comparison.

AMD Radeon RX 6500 XT. The Radeon RX 6500 XT averages 11897. The GTX 960A is 0.9% faster in this test, which is the largest advantage the 960A holds over any nearest rival. Even so, the difference is minimal in practical terms.

Power and Cooling — TDP, PSU recommendation, connector requirements

The GTX 960A is rated at 75 W TDP. It has no power connectors, and the slot width is listed as MXM Module, so the module is powered through the host system's MXM-B (3.0) interface rather than through a supplementary cable. The fact pack lists no suggested PSU; because this is a mobile MXM part, any power-delivery considerations belong to the host laptop rather than to a desktop power supply. Cooling details are not listed, but the 75 W TDP keeps the thermal envelope relatively modest for a portable module.

FAQ

Q: What chip and architecture does the GeForce GTX 960A use?

A: It uses the GM107 chip with the Maxwell architecture, fabricated by TSMC on a 28 nm process. The die contains 1,870 million transistors and measures 148 mm², for a transistor density of 12.6M / mm².

Q: How much memory does it have, and what is the memory bandwidth?

A: It has 2 GB of GDDR5 on a 128-bit bus, with a memory clock of 1253 MHz and 5 Gbps effective signaling, yielding 80.19 GB/s of bandwidth.

Q: Does the GTX 960A have ray tracing or tensor cores?

A: The fact pack lists no RT cores and no tensor cores. Its API support is DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4, so there is no dedicated ray tracing or tensor compute hardware listed for this Maxwell part.

Q: What power connectors does the GTX 960A need?

A: None. The TDP is 75 W, and the module uses the MXM-B (3.0) bus interface with an MXM Module slot width. The fact pack does not list a suggested PSU.

Q: How does the GTX 960A compare to the GeForce GTX 670 in the recorded benchmark?

A: The GeForce GTX 670 has an average score of 12014, while the GTX 960A scores 12002. The delta is -0.1%, so the two GPUs are essentially tied.

Q: What is the production status of the GTX 960A?

A: It is listed as end-of-life. It was released on 2015-03-12, and its predecessor is listed as the GeForce 800A. No successor is listed.

Memory Subsystem

The GTX 960A's memory subsystem is centered on 2 GB of GDDR5, a 128-bit bus, and 80.19 GB/s of bandwidth. The memory clock is 1253 MHz, with an effective signaling rate of 5 Gbps. For high-resolution workloads, the two limiting factors are capacity and bus width: 2 GB is a moderate buffer, and the 128-bit bus restricts how much data can move per clock. The pixel rate of 18.82 GPixel/s and texture rate of 47.04 GTexel/s further indicate that heavily textured scenes will strain the render backend. The data suggests a card that can handle moderate-quality scenes comfortably but will need reduced texture resolution and filtering loads when frame-buffer pressure rises.

Ray Tracing and Feature Set — RT/tensor cores, API support from facts

The fact pack for the GTX 960A lists no RT cores and no tensor cores, so the data contains no dedicated ray tracing hardware or tensor compute acceleration. The architecture is Maxwell, and the API support is DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. That means the GPU is exposed to modern graphics APIs at the 11_0 feature level for DirectX 12. Without dedicated RT or tensor cores, any ray tracing or machine learning workloads would have to run through the general-purpose shading units, of which there are 640, alongside 40 TMUs and 16 ROPs. Feature support is therefore conventional for a mobile GPU from its release period, with API compatibility but no specialized acceleration for newer ray-traced or tensor-based workloads.