NVIDIA GeForce GTX 950 vs NVIDIA GeForce GTX 965M Comparison

Head-to-Head Benchmarks

The benchmark database records two direct comparisons between the NVIDIA GeForce GTX 965M and the NVIDIA GeForce GTX 950. The GTX 950 wins both tests, though the margins differ significantly by workload.

In Geekbench OpenCL, the GTX 950 scores 15,662 against the GTX 965M’s 14,509. That is a 7.4% advantage for the desktop card. The gap is not trivial, but it is modest enough that the mobile part stays competitive in compute-oriented tasks. The GTX 965M’s higher shading unit count and texture rate help it close some of the clock-speed deficit.

The Vulkan test tells a different story. The GTX 950 posts 16,734, while the GTX 965M manages 14,299. That is a 14.6% lead for the GTX 950. Vulkan tends to expose raw throughput and driver efficiency more directly than OpenCL, and the GTX 950’s higher boost clock clearly pays off here. The delta is roughly double what OpenCL showed, which suggests the GTX 950’s architecture scaling under modern API workloads is stronger than the 965M’s.

Looking at the broader database context, the GTX 965M has an average benchmark score of 14,404 across its recorded tests. Its nearest rivals cluster tightly around it: the AMD Radeon RX Vega 11 sits 0.1% ahead at 14,385, the NVIDIA GeForce GTX TITAN is 0.2% behind at 14,373, the AMD Radeon Vega 11 trails by 0.4% at 14,352, and the Intel Iris Xe MAX Graphics is 0.6% back at 14,315. The GTX 965M sits in a dense pack where tiny margins separate competitors, and its 56th percentile among all GPUs reflects that middle-of-the-road positioning.

The GTX 950, by contrast, has an average benchmark score of 13,189, which lands at the 53rd percentile. Its nearest rivals are more spread out. The NVIDIA GeForce GTX 480 is 0.8% ahead at 13,300, the AMD Radeon Pro 555X is 1% ahead at 13,321, and the AMD FirePro M6100 leads by 1.2% at 13,354. Only the NVIDIA Tesla M2090 trails, at 13,075, which is 0.9% behind. The GTX 950’s average is dragged down by the inclusion of a Geekbench Metal score of 7,172, a test the GTX 965M does not have recorded. If Metal were excluded, the GTX 950’s average would be substantially higher, but the database records what it records.

The head-to-head numbers are unambiguous. The GTX 950 wins both direct comparisons, with the Vulkan margin being the decisive statement. The GTX 965M never takes a single recorded victory in these two tests. That said, the OpenCL result shows the 965M is not embarrassed; a 7.4% gap is within reach of driver optimizations or workload variations. The Vulkan gap is the more telling figure for anyone planning to run modern games or compute that leverages low-level APIs.

Architecture Differences

Both GPUs are built on NVIDIA’s Maxwell 2.0 architecture and use TSMC’s 28 nm process node. That is where the similarities in silicon design philosophy end, because the underlying chips differ considerably.

The GTX 965M uses the GM204 chip, the same die family that powers higher-end desktop parts. It packs 5,200 million transistors onto a 398 mm² die, yielding a transistor density of 13.1 million transistors per mm². The GTX 950 uses the smaller GM206 chip, with 2,940 million transistors on a 228 mm² die, for a density of 12.9 million per mm². The 965M’s die is roughly 75% larger by area and carries about 77% more transistors, but that does not translate into a performance win because the mobile part is constrained by clock speeds and power limits.

The GTX 965M fields 1,024 shading units, 64 texture mapping units, and 32 ROPs. The GTX 950 has 768 shading units, 48 TMUs, and 32 ROPs. The 965M has a 33% advantage in shading units and a 33% advantage in TMUs, while ROP counts are equal. In theory, that should give the 965M a clear edge in pixel-heavy and texture-heavy workloads. The benchmark data does not show that, which points to clock speeds and memory bandwidth as the limiting factors.

Clock speeds favor the GTX 950 decisively. The GTX 950 runs a base clock of 1,024 MHz and a boost clock of 1,188 MHz. The GTX 965M runs 924 MHz base and 950 MHz boost. That means the GTX 950 has a 10.8% higher base clock and a 25.1% higher boost clock. The boost clock gap is enormous, and it explains why a chip with fewer cores can still win in most scenarios. The 965M’s boost behavior on mobile platforms is often further limited by thermal and power headroom, though the database does not record those constraints.

Memory configuration is another differentiator. Both cards use 2 GB of GDDR5 on a 128-bit bus. The GTX 950’s memory runs at 1,653 MHz with 6.6 Gbps effective, delivering 105.8 GB/s of bandwidth. The GTX 965M’s memory runs at 1,253 MHz with 5 Gbps effective, delivering 80.19 GB/s. That is a 31.9% bandwidth advantage for the GTX 950, a substantial margin that directly impacts fill-rate-bound workloads and high-resolution texture streaming.

Pixel rates and texture rates reflect these differences in practice. The GTX 950 posts 38.02 GPixel/s and 57.02 GTexel/s, while the GTX 965M manages 30.40 GPixel/s and 60.80 GTexel/s. The 950 leads in pixel rate by 25.1%, matching its clock advantage, while the 965M leads in texture rate by 6.6%, reflecting its higher TMU count. FP32 compute is close: the 965M delivers 1.946 TFLOPS against the 950’s 1.825 TFLOPS, a 6.6% lead for the mobile part.

Other specifications diverge on the platform level. The GTX 965M is an MXM Module with an MXM-B 3.0 interface and no power connectors, designed for laptops with portable-device-dependent display outputs. The GTX 950 is a dual-slot desktop card with a 90 W TDP, a single 6-pin power connector, a suggested PSU of 250 W, and a PCIe 3.0 x16 interface. It measures 202 mm (8 inches) in length and offers 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2 outputs. The 965M’s TDP is not recorded in the database, so no direct power comparison is possible.

Both cards support DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The GTX 950 has a recorded Geekbench Metal score of 7,172, while the 965M has none, which may reflect the target platform rather than capability. The GTX 965M launched on January 8, 2015, and the GTX 950 on August 19, 2015, both now end-of-life. The 965M’s predecessor is GeForce 800M and its successor is GeForce 10 Mobile; the 950’s predecessor is GeForce 700 and its successor is GeForce 10.

Where Each One Wins

The GTX 950 wins on raw benchmark performance in both recorded tests, but the nature of those wins suggests specific strengths. Its Vulkan advantage of 14.6% indicates that low-level API workloads, which include modern game engines and compute shaders, favor the desktop card heavily. The boost clock of 1,188 MHz, combined with 105.8 GB/s of memory bandwidth, gives it a clear edge in scenarios that are latency-sensitive or bandwidth-hungry. Users running Vulkan-based titles or applications that leverage explicit multi-threading should expect the GTX 950 to pull ahead noticeably.

The GTX 950 also excels in pixel throughput. Its 38.02 GPixel/s is 25.1% higher than the 965M’s 30.40 GPixel/s, which matters for high-resolution rendering, anti-aliasing, and any workload that stresses ROP output. The equal 32 ROP counts mean the 950’s clock advantage translates directly into faster pixel processing.

The GTX 965M’s wins are more theoretical than benchmark-proven, but they are real in the right workloads. Its FP32 compute of 1.946 TFLOPS is 6.6% higher than the 950’s 1.825 TFLOPS, which could benefit GPGPU tasks like physics simulation, Folding@Home-style workloads, or OpenCL compute kernels that are shader-bound rather than memory-bound. The 965M’s texture rate of 3% advantage, 60.80 GTexel/s versus 57.02 GTexel/s, is smaller but still meaningful for texture-heavy scenes with heavy anisotropic filtering or detail mapping.

In OpenCL specifically, the 965M loses by 7.4%, a margin small enough that driver tweaks or workload-specific optimizations could flip the result. The 965M’s higher shading unit count of 1,024 versus 768 gives it headroom in compute tasks that scale well with core count, such as rendering, image processing, or simulation code that does not rely on memory bandwidth. The GTX 965M also has an advantage in transistor count and die area, 5,200 million versus 2,940 million, which historically correlates with better sustained performance in shader-heavy workloads that saturate compute units.

For gaming at typical laptop resolutions, the 965M’s 2 GB GDDR5 frame buffer matches the 950’s, so VRAM capacity is not a differentiator. The 965M’s lower memory bandwidth of 80.19 GB/s versus 105.8 GB/s, a 31.9% deficit, will hurt in games that stream large textures or use high-resolution shadow maps. The 950’s higher pixel rate also gives it an edge in games that run at 1080p with heavy post-processing.

Power and platform shape the usage case as well. The GTX 965M is an MXM module with no power connectors, designed for laptops where power efficiency matters more than peak performance. The GTX 950 is a desktop card with a 90 W TDP, a 250 W suggested PSU requirement, and a 6-pin connector, meaning it can sustain its boost clock more reliably. The 965M’s TDP is not recorded, so no direct comparison exists there, but the MXM form factor implies thermal limits that the desktop card does not face.

The Verdict

The data shows a clear performance hierarchy in the recorded benchmarks. The GTX 950 wins both head-to-head tests, with margins of 7.4% in OpenCL and 14.6% in Vulkan. If the only question is which GPU runs faster in measured workloads, the answer is the GTX 950 without qualification. Users building a desktop system with an MXM slot should expect the 950 to outperform the 965M in most scenarios it runs, particularly those using Vulkan-heavy engines.

The GTX 965M is not without a case. Its FP32 compute is 6.6% higher at 1.946 TFLOPS, its texture rate is 6.6% higher at 60.80 GTexel/s, and its shading unit and TMU counts are 33% higher at 1,024 and 64 respectively. For compute-focused users who prioritize raw shader throughput over bandwidth and clock speed, the 965M offers measurable advantages. That is a narrow niche, but it exists.

The GTX 950’s 105.8 GB/s memory bandwidth, 31.9% above the 965M, and its 38.02 GPixel/s pixel rate, 25.1% above, make it the better choice for gaming and real-time graphics. The 14.6% Vulkan delta is the strongest single datapoint in the comparison, and it aligns with the modern API trajectory. Users running current or future titles that use Vulkan or DirectX 12 will see the GTX 950 pull further ahead than older API benchmarks suggest.

The GTX 965M’s 56th percentile versus the GTX 950’s 53rd percentile may seem contradictory given the head-to-head results, but the average benchmark scores include different test sets. The 965M’s average of 14,404 comes from OpenCL and Vulkan only, while the 950’s average of 13,189 includes a Metal score of 7,172 that drags it down. The head-to-head tests are the more reliable comparison for these two specific GPUs, and they favor the 950.

The GTX 950 carries a launch MSRP of 159 USD, which can be stated once as a point of reference, but the database does not record a launch price for the 965M, so no price comparison is possible. Both are end-of-life products, so availability will vary.

For a laptop user who needs a GPU in an MXM slot, the 965M is the only option of the two, and it performs reasonably in OpenCL, sitting within 0.6% of the Intel Iris Xe MAX Graphics and 0.1% of the AMD Radeon RX Vega 11. For a desktop user with a PCIe slot, the GTX 950 is the clear pick from the benchmark data, beating the 965M in both recorded tests and offering higher clocks, more bandwidth, and faster pixel output. Choose the 965M only if the MXM form factor is mandatory and compute-heavy workloads dominate; choose the 950 for everything else.

FAQ

Q: Which GPU wins in the recorded head-to-head benchmarks?

A: The NVIDIA GeForce GTX 950 wins both tests. It scores 15,662 in Geekbench OpenCL against the GTX 965M’s 14,509, a 7.4% lead, and 16,734 in Geekbench Vulkan against 14,299, a 14.6% lead.

Q: How do the clock speeds compare between the two cards?

A: The GTX 950 runs at a base clock of 1,024 MHz and a boost clock of 1,188 MHz. The GTX 965M runs at 924 MHz base and 950 MHz boost, giving the 950 a 10.8% higher base clock and a 25.1% higher boost clock.

Q: Does the GTX 965M have more shading units?

A: Yes, the GTX 965M has 1,024 shading units and 64 TMUs, while the GTX 950 has 768 shading units and 48 TMUs. The 965M leads in FP32 compute at 1.946 TFLOPS against 1.825 TFLOPS.

Q: What memory bandwidth does each GPU offer?

A: The GTX 950 has 105.8 GB/s of memory bandwidth from 2 GB of GDDR5 on a 128-bit bus. The GTX 965M has 80.19 GB/s from the same 2 GB GDDR5 on a 128-bit bus, a 31.9% deficit.

Q: Are the two GPUs built on the same architecture?

A: Both use Maxwell 2.0 architecture on TSMC’s 28 nm process node. The GTX 965M uses the GM204 chip with 5,200 million transistors on a 398 mm² die, while the GTX 950 uses GM206 with 2,940 million transistors on a 228 mm² die.

Q: What is the GTX 950’s launch MSRP?

A: The GTX 950 has a launch MSRP of 159 USD. The GTX 965M has no recorded launch MSRP in the database.