NVIDIA GeForce GTX 950A vs NVIDIA GeForce GTX 960M Comparison

NVIDIA GeForce GTX 950A and NVIDIA GeForce GTX 960M are both end-of-life mobile graphics solutions from the Maxwell generation, built on the same GM107 chip at TSMC's 28 nm process. The data shows they are closely matched, with the GTX 960M taking the single head-to-head benchmark victory, but the GTX 950A holds a higher percentile ranking against all GPUs. The core difference is memory: the GTX 950A uses 2 GB of DDR3, while the GTX 960M uses 4 GB of GDDR5, which explains the performance gap despite identical shader counts.

Where Each One Wins

The benchmark results split cleanly based on memory technology and clock speeds. In the only direct head-to-head test, Geekbench OpenCL, the GTX 960M wins with a score of 11045 against the GTX 950A's 10273, a 7% advantage. This is the sole win for the GTX 960M, but it is the definitive metric for raw compute performance in this comparison.

The GTX 950A, despite losing the head-to-head, wins on relative standing. Its percentile rank of 48 places it higher than the GTX 960M's 46th percentile. This means that while the GTX 960M produces a higher absolute score, the GTX 950A sits slightly better relative to the entire GPU landscape. The GTX 950A also has a higher average benchmark score of 10273 compared to the GTX 960M's 9645, which is dragged down by the GTX 960M's additional Vulkan result of 8245.

For practical use cases, the GTX 960M is the better choice for compute-heavy tasks like OpenCL workloads, where its 7% lead translates to measurable gains. The GTX 950A, however, is not without merit; its higher percentile suggests it handles a broader range of tasks more consistently relative to other GPUs. The GTX 960M's GDDR5 memory and higher clocks make it the stronger performer, but the GTX 950A's position in the 48th percentile versus the 46th indicates it is marginally closer to the median GPU experience.

Architecture Differences

Both GPUs share the same fundamental architecture: Maxwell on a 28 nm TSMC process, with 1,870 million transistors on a 148 mm² die and a transistor density of 12.6M per mm². The shading units, texture mapping units, and render output units are identical at 640, 40, and 16, respectively. Neither GPU features ray tracing or tensor cores, and both support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4.

The differences emerge in clock speeds and memory configuration. The GTX 960M runs at a base clock of 1097 MHz and a boost of 1176 MHz, while the GTX 950A is slower at 993 MHz base and 1124 MHz boost. This clock advantage directly boosts the GTX 960M's pixel rate (18.82 GPixel/s vs 17.98 GPixel/s), texture rate (47.04 GTexel/s vs 44.96 GTexel/s), and FP32 throughput (1.505 TFLOPS vs 1,438.7 GFLOPS).

Memory is the most significant architectural split. The GTX 950A uses 2 GB of DDR3 on a 128-bit bus, yielding 32.03 GB/s of bandwidth. The GTX 960M doubles the capacity to 4 GB and switches to GDDR5, more than doubling bandwidth to 80.19 GB/s. Memory clocks reflect this: the GTX 950A runs at 1001 MHz (2 Gbps effective), while the GTX 960M runs at 1253 MHz (5 Gbps effective). This bandwidth difference is the primary reason the GTX 960M wins the OpenCL test despite both GPUs having the same core configuration.

The power envelope is identical: both are rated at 75 W TDP, use MXM Module slot width, have no power connectors, and connect via MXM-B (3.0). Display outputs are portable device dependent for both. The GTX 960M has a successor in the GeForce 10 Mobile series, while the GTX 950A does not list a successor, and both trace their lineage to the GeForce 800 series (800A and 800M respectively).

The Verdict

From the data, the NVIDIA GeForce GTX 960M is the clear pick for anyone prioritizing raw compute performance. Its 7% lead in Geekbench OpenCL over the GTX 950A is decisive, and its 4 GB of GDDR5 memory with 80.19 GB/s bandwidth versus the GTX 950A's 2 GB DDR3 at 32.03 GB/s makes it the superior choice for memory-intensive workloads. The GTX 960M's higher clocks across the board — base, boost, and memory — translate to better pixel rate, texture rate, and FP32 throughput.

The NVIDIA GeForce GTX 950A, however, should not be dismissed. Its 48th percentile ranking versus the GTX 960M's 46th shows it holds up better against the broader GPU field, and its average benchmark score of 10273 is higher than the GTX 960M's 9645. If you are comparing these two in a vacuum, the GTX 960M wins. If you are looking at which card sits closer to the middle of the pack, the GTX 950A edges ahead.

For a specific use case: if you need OpenCL compute, choose the GTX 960M. If you are evaluating based on percentile standing or average scores across multiple benchmarks, the GTX 950A offers a marginally better relative position. Both are end-of-life MXM modules with 75 W TDP, so thermal and power considerations are identical. The GTX 960M's GDDR5 memory alone justifies its selection for modern workloads that benefit from higher bandwidth.

FAQ

Q: Which GPU has a higher Geekbench OpenCL score?

A: The NVIDIA GeForce GTX 960M scores 11045, while the NVIDIA GeForce GTX 950A scores 10273. The GTX 960M wins by 7%.

Q: Do both GPUs have the same number of shading units?

A: Yes, both the GTX 950A and GTX 960M have 640 shading units, along with 40 texture mapping units and 16 render output units.

Q: What is the memory bandwidth difference between the two?

A: The GTX 960M has 80.19 GB/s of bandwidth with 4 GB of GDDR5, while the GTX 950A has 32.03 GB/s with 2 GB of DDR3. This is more than a 2.5x difference in favor of the GTX 960M.

Q: Which GPU has a higher percentile ranking against all GPUs?

A: The GTX 950A ranks in the 48th percentile, while the GTX 960M ranks in the 46th percentile. The GTX 950A is slightly higher relative to the entire GPU pool.

Q: Are the power requirements the same for both GPUs?

A: Yes, both are rated at 75 W TDP, use MXM Module slot width, have no power connectors, and require no suggested PSU.

Q: What is the average benchmark score for each GPU?

A: The GTX 950A has an average benchmark score of 10273. The GTX 960M has an average of 9645, which includes a Geekbench Vulkan score of 8245 alongside its higher OpenCL score.

Head-to-Head Benchmarks

The single head-to-head benchmark is Geekbench OpenCL, and it is a decisive win for the GTX 960M. The GTX 960M scores 11045 against the GTX 950A's 10273, a delta of -7% from the GTX 950A's perspective. This score difference is consistent with the hardware specifications: the GTX 960M's boost clock of 1176 MHz versus 1124 MHz, and its memory bandwidth of 80.19 GB/s versus 32.03 GB/s, give it a clear compute advantage.

The GTX 960M's FP32 performance of 1.505 TFLOPS versus the GTX 950A's 1,438.7 GFLOPS is a 4.6% lead, which is partially offset by the much larger memory bandwidth advantage. This suggests that the OpenCL test is memory-sensitive, not just compute-bound. The GTX 960M's 4 GB GDDR5 buffer also allows larger datasets to reside on the GPU, reducing reliance on system memory.

The GTX 950A's only counterargument comes from its percentile placement. At 48th percentile, it ranks two points higher than the GTX 960M's 46th, meaning its single OpenCL score of 10273 places it closer to the median GPU than the GTX 960M's mixed benchmark results. The GTX 960M's Vulkan score of 8245 drags its average down, even though its OpenCL score is superior.

Specification Differences

| Specification | NVIDIA GeForce GTX 950A | NVIDIA GeForce GTX 960M |

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

| Generation | GeForce 900A | GeForce 900M |

| Base Clock | 993 MHz | 1097 MHz |

| Boost Clock | 1124 MHz | 1176 MHz |

| Memory Clock | 1001 MHz (2 Gbps effective) | 1253 MHz (5 Gbps effective) |

| Memory Size | 2 GB | 4 GB |

| Memory Type | DDR3 | GDDR5 |

| Memory Bus Width | 128 bit | 128 bit |

| Memory Bandwidth | 32.03 GB/s | 80.19 GB/s |

| Pixel Rate | 17.98 GPixel/s | 18.82 GPixel/s |

| Texture Rate | 44.96 GTexel/s | 47.04 GTexel/s |

| FP32 Performance | 1,438.7 GFLOPS | 1.505 TFLOPS |

| Predecessor | GeForce 800A | GeForce 800M |

| Successor | None | GeForce 10 Mobile |

Both GPUs share identical core counts (640 shading units, 40 TMUs, 16 ROPs), the same 28 nm process, 1,870 million transistors, 148 mm² die size, 75 W TDP, and MXM-B (3.0) bus interface. The GTX 960M is faster in every clocked specification, while the GTX 950A's only advantage is its higher percentile rank. The GDDR5 memory of the GTX 960M is the standout difference, providing 2.5x the bandwidth of the GTX 950A's DDR3, which is the primary driver of its performance lead.