NVIDIA GeForce GTX 950A

The NVIDIA GeForce GTX 950A belongs to the GeForce 900A generation and is listed as an end-of-life MXM Module based on the GM107 Maxwell chip, manufactured on TSMC’s 28 nm process. The die contains 1,870 million transistors within 148 mm², giving a transistor density of 12.6M per mm², with clocks set at 993 MHz base and 1124 MHz boost. Its single listed benchmark result is a Geekbench OpenCL score of 10,220, which places it at the 48th percentile of all GPUs in the database. All four nearest rivals sit within 2.2%, making this a tightly packed performance cluster.

Benchmark Performance

The data contains one benchmark entry: Geekbench OpenCL, where the GTX 950A scores 10,220. The 48th percentile placement is the key context: this is a middle-of-the-distribution result, not a low-end or high-end outlier. The compute resources behind that score are 640 shading units, 40 texture mapping units, and 16 ROPs, producing a texture rate of 44.96 GTexel/s and a pixel rate of 17.98 GPixel/s. Peak FP32 throughput is listed at 1,438.7 GFLOPS.

The nearest rival comparison reinforces the median position. Intel Iris Pro Graphics P580 averages 10,189, which is 0.3% lower. NVIDIA Quadro P4000 averages 10,134, or 0.8% lower. AMD Radeon RX 550X averages 10,095, or 1.2% lower. AMD Radeon R9 M375 averages 10,001, or 2.2% lower. No rival in the list is decisively behind; the spread from the GTX 950A to its farthest listed rival is just 2.2%. What does that imply? In the OpenCL workload captured by the benchmark, the GTX 950A and these four rivals are effectively in the same performance bin. A single test cannot establish a wide hierarchy, but it does show that the card competes with a specific set of mid-range parts rather than pulling ahead of them.

The percentile figure tells a similar story. Being at the 48th percentile means roughly half of all GPUs in the database sit above it, and half sit below. That is consistent with the near-tie delta values: the GTX 950A is not a performance leader, but it is also not far off the pace of its closest recorded competitors.

Memory Subsystem

The GTX 950A ships with 2 GB of DDR3 memory on a 128-bit bus, yielding 32.03 GB/s of bandwidth. The memory clock is listed as 1001 MHz with 2 Gbps effective. These numbers matter for how the card handles high-resolution work. The 2 GB frame buffer is a hard capacity limit: once textures, geometry, and render targets exceed that amount, performance degrades. Meanwhile, the 32.03 GB/s bandwidth is a throughput ceiling that limits how quickly the GPU can access textures and write framebuffer data.

The use of DDR3 rather than a faster memory type explains the modest bandwidth figure. For high-resolution scenes, the memory subsystem is likely to be the first bottleneck. A lower-resolution or texture-light workload keeps the card inside the 2 GB capacity and closer to the bandwidth the system can provide. The combination of 2 GB, a 128-bit interface, and DDR3 suggests that the GTX 950A is more comfortable with constrained scene complexity than with massive, high-resolution assets.

Power and Cooling

The listed TDP for the GTX 950A is 75 W. Power connectors are recorded as “None,” which indicates that the MXM slot itself is expected to supply power. The slot width field is “MXM Module,” and the bus interface is MXM-B (3.0). Display outputs are listed as “Portable Device Dependent,” meaning the actual video connectors depend on the host portable system rather than being fixed by the GPU module.

No suggested PSU is provided in the fact pack, and no cooler specifications are recorded. The 75 W TDP and absence of auxiliary power connectors mean power delivery is relatively simple from a system-integration standpoint: the module does not require separate cabling. Because this is an end-of-life product released on 2015-03-12, the power and thermal behavior are tied to the original MXM platform era. The record also notes a predecessor of GeForce 800A, but no successor is listed.

How It Compares

Intel Iris Pro Graphics P580: With an average score of 10,189, the P580 trails the GTX 950A by only 0.3%. That is a near-tie in the recorded OpenCL benchmark. The delta is small enough that the two should be treated as equivalent in this workload.

NVIDIA Quadro P4000: The Quadro P4000 averages 10,134, which is 0.8% below the GTX 950A. Despite the different product naming and intended market positioning implied by the Quadro name, the benchmark data puts the two at essentially the same performance level in this test.

AMD Radeon RX 550X: The RX 550X averages 10,095, or 1.2% lower than the GTX 950A. The gap is still narrow, placing the RX 550X within the same cluster as the GTX 950A and the other rivals.

AMD Radeon R9 M375: The R9 M375 averages 10,001, 2.2% below the GTX 950A. This is the largest delta among the listed nearest rivals, yet it remains a small margin. In practical terms, the benchmark record does not show a meaningful performance separation between the GTX 950A and this card.

Ray Tracing and Feature Set

The fact pack records no rtCores and no tensorCores for the GTX 950A. That means the datasheet does not assign dedicated ray tracing hardware or tensor cores to this part. The architecture is Maxwell, a generation built around traditional rasterization and compute rather than the dedicated acceleration structures found in later designs.

API support is listed as DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. The DirectX entry of “12 (11_0)” is notable: it indicates support for DirectX 12 at feature level 11_0, not a full 12_0 feature level. The core rendering resources are 640 shading units, 40 TMUs, and 16 ROPs, with FP32 performance of 1,438.7 GFLOPS and the pixel and texture rates already noted. For users interested in ray tracing, the data provides no evidence of dedicated RT cores, and the absence of tensor cores also means no listed AI-acceleration hardware.

FAQ

Q: What Geekbench OpenCL score does the GTX 950A achieve?

A: It scores 10,220, which places it at the 48th percentile of all GPUs in the database.

Q: How close are its nearest rivals?

A: The closest listed rival is Intel Iris Pro Graphics P580, which averages 10,189 and is 0.3% lower. The farthest listed rival is AMD Radeon R9 M375, which averages 10,001 and is 2.2% lower.

Q: What is the memory configuration?

A: The GTX 950A has 2 GB of DDR3 memory on a 128-bit bus, with 32.03 GB/s of bandwidth and a 1001 MHz memory clock, listed as 2 Gbps effective.

Q: Does the card require additional power connectors?

A: No. The power connectors field is “None,” and the TDP is 75 W. The slot width is an MXM Module with an MXM-B (3.0) bus interface.

Q: Which APIs are supported?

A: The listed APIs are DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4.

Q: What chip is inside the GTX 950A?

A: The chip is GM107, built on the Maxwell architecture using TSMC’s 28 nm process. It contains 1,870 million transistors on a 148 mm² die, with a transistor density of 12.6M per mm².

Who Should Consider It

The GTX 950A is a candidate for users whose workloads stay within a 2 GB frame buffer and a 32.03 GB/s memory ceiling. The 48th percentile OpenCL score indicates a mid-pack result, so it is not a card for settings that demand maximum texture detail or very large high-resolution render targets. Instead, the data points toward moderate settings and lower-resolution scenarios where the memory capacity and bandwidth constraints can be respected.

The listed lack of RT and tensor cores means ray tracing acceleration is not a reason to select this part. Users focused on traditional rasterization, DX11-era feature levels, or Vulkan-based lighter 3D loads may find the GTX 950A more relevant. Because the display outputs are portable-device dependent and the card is an MXM Module, it is best understood as a component for portable or modular systems that accept MXM-B (3.0) modules. Its 75 W TDP and absence of auxiliary power connectors make integration simpler from a power-delivery perspective, but the overall performance cluster remains firmly in the middle of the field.