NVIDIA GeForce 945M vs NVIDIA GeForce GTX 960M Comparison

Head-to-Head Benchmarks

The database records a single head-to-head benchmark between these two mobile GPUs, and the result is decisive. In the Geekbench OpenCL test, the NVIDIA GeForce GTX 960M scores 11,045 points against the NVIDIA GeForce 945M's 8,099 points. That is a 36.4% advantage for the GTX 960M, a gap large enough to classify these as different performance tiers rather than close competitors. The delta is not a marginal edge; it represents a substantial computing advantage in raw throughput.

Context from the nearest rivals reinforces the separation. The GTX 960M's average benchmark score sits at 9,645, placing it at the 46th percentile among all GPUs in the database. Its closest competitor, the NVIDIA Quadro K5000, averages 9,637, a delta of just 0.1%, meaning the GTX 960M is effectively on par with that workstation card. The AMD Radeon Pro WX 2100 sits at 9,653, only 0.1% above, and the NVIDIA Quadro P4000 at 9,665 is 0.2% ahead. The spread among these rivals is remarkably tight, all within 0.7% of each other, which suggests the GTX 960M occupies a well-populated performance band where small differences dominate.

The 945M, by contrast, averages 8,099 points and sits at the 42nd percentile. Its nearest rivals cluster at similar levels: the NVIDIA GRID K2 scores 8,080 (0.2% above the 945M), the NVIDIA GeForce GTX 650 Ti Boost scores 8,067 (0.4% above), while the AMD Radeon R9 M360 scores 8,129 (0.4% below) and the NVIDIA GeForce GTX 950M scores 8,135 (0.4% below). The 945M is therefore sandwiched in a group of older and lower-tier parts, roughly 1,500 points behind the GTX 960M's average. In percentile terms, the difference is only 4 positions (46 versus 42), but the raw score gap tells the more meaningful story: the GTX 960M delivers roughly 19% higher average performance across all recorded benchmarks than the 945M.

The single OpenCL result is the only direct comparison available, but the average scores corroborate the same direction. The GTX 960M also has a Geekbench Vulkan score of 8,245, which is interesting because that figure nearly matches the 945M's best OpenCL result. In other words, the GTX 960M's Vulkan performance alone is comparable to the 945M's peak OpenCL performance, even before accounting for the GTX 960M's superior OpenCL showing.

The Verdict

The data points to a clear hierarchy: the NVIDIA GeForce GTX 960M is the stronger GPU in every recorded metric. Its OpenCL score leads by 36.4%, its average benchmark score leads by roughly 19%, and its Vulkan result of 8,245 exceeds the 945M's entire benchmark portfolio. The 945M's only recorded benchmark, OpenCL at 8,099, falls short of the GTX 960M's average score of 9,645 by 1,546 points. There is no benchmark category in the database where the 945M wins.

For users choosing between these two, the decision hinges on whether the performance gap justifies the trade-offs. The GTX 960M is the obvious pick for anyone prioritizing compute performance, especially in OpenCL workloads where the 36.4% advantage is decisive. The 945M, with its lower clocks and slower memory, appears positioned for scenarios where power draw and thermal constraints are paramount, though the database lists both at 75 W TDP, so that distinction does not appear in the recorded specifications.

The 945M's closest rival, the GTX 950M, averages 8,135 points, only 0.4% above the 945M. That proximity suggests the 945M is a slow member of the GeForce 900M mobile family. The GTX 960M, meanwhile, sits alongside workstation parts like the Quadro P4000 and K5000, indicating it punches above its mobile-class designation. For gaming or GPU-accelerated tasks, the GTX 960M is the only rational choice from this pair. The 945M would only make sense if the system integration requires a part with lower memory bandwidth demands, but even then, the performance penalty is steep.

FAQ

Q: How much faster is the GTX 960M than the 945M in OpenCL?

A: The GTX 960M scores 11,045 against the 945M's 8,099, a 36.4% advantage.

Q: Does the 945M win any benchmark category?

A: No. The database records only one head-to-head test, Geekbench OpenCL, and the GTX 960M wins it. The 945M has no recorded Vulkan score, while the GTX 960M posts 8,245 in that test.

Q: How do these GPUs compare to their nearest rivals?

A: The GTX 960M averages 9,645, within 0.2% of the Quadro P4000 (9,665) and 0.1% of the Radeon Pro WX 2100 (9,653). The 945M averages 8,099, within 0.4% of the GTX 950M (8,135) and Radeon R9 M360 (8,129).

Q: What is the memory configuration difference?

A: The GTX 960M uses 4 GB of GDDR5 on a 128-bit bus with 80.19 GB/s bandwidth. The 945M uses 2 GB of DDR3 on the same 128-bit bus but only 28.80 GB/s bandwidth.

Q: Are clock speeds different between the two?

A: Yes. The GTX 960M runs at 1097 MHz base and 1176 MHz boost, while the 945M runs at 928 MHz base and 1020 MHz boost. Memory clocks also differ: 1253 MHz (5 Gbps effective) for the GTX 960M versus 900 MHz (1800 Mbps effective) for the 945M.

Q: Which GPU has a higher percentile ranking?

A: The GTX 960M sits at the 46th percentile among all GPUs, while the 945M sits at the 42nd percentile.

Specification Differences

The two GPUs share the same silicon foundation: both use NVIDIA's GM107 chip on a 28 nm process from TSMC, with 1,870 million transistors on a 148 mm² die and a transistor density of 12.6M per mm². Shader counts are identical at 640 shading units, with 40 texture mapping units and 16 raster output units. Both are Maxwell architecture parts from the GeForce 900M generation, both use MXM-B (3.0) interfaces, and both are listed as end-of-life products with no power connectors required at 75 W TDP. Display outputs are portable-device dependent for both.

The differences begin with clocks. The GTX 960M runs at 1097 MHz base and 1176 MHz boost, while the 945M is clocked lower at 928 MHz base and 1020 MHz boost. That 169 MHz base gap and 156 MHz boost gap translate directly into throughput differences. The GTX 960M achieves 18.82 GPixel/s pixel rate and 47.04 GTexel/s texture rate, versus 16.32 GPixel/s and 40.80 GTexel/s for the 945M. Floating-point performance follows suit: 1.505 TFLOPS for the GTX 960M against 1,305.6 GFLOPS for the 945M.

Memory is the other major separation. The GTX 960M carries 4 GB of GDDR5 with a 128-bit bus and 80.19 GB/s bandwidth. The 945M carries 2 GB of DDR3 with the same bus width but only 28.80 GB/s bandwidth, a 64% reduction in memory throughput. Memory clocks also differ: 1253 MHz (5 Gbps effective) versus 900 MHz (1800 Mbps effective). Release dates are apart by roughly seven months: the GTX 960M launched on March 12, 2015, while the 945M launched on October 26, 2015. Both share the same predecessor (GeForce 800M) and successor (GeForce 10 Mobile).

Architecture Differences

Both GPUs are built on NVIDIA's Maxwell architecture using the GM107 chip, which means the core compute architecture is fundamentally the same. The shading units, TMUs, and ROPs are identical in count, and both support DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. The transistor count, die size, and manufacturing process are also identical, confirming that these are two variants of the same physical silicon with different binning and memory configurations.

The architectural difference that matters most is memory type. The GTX 960M pairs GM107 with GDDR5, which provides 80.19 GB/s of bandwidth. The 945M pairs the same chip with DDR3, capping bandwidth at 28.80 GB/s. This is not a minor tweak; it changes the memory subsystem's character entirely. GDDR5 delivers multiple times the bandwidth of DDR3 on the same bus width, which directly affects texture-heavy workloads, frame buffer operations, and any compute task that streams large datasets. The 945M's lower clocks compound this: the memory runs at 900 MHz (1800 Mbps effective) versus 1253 MHz (5 Gbps effective) on the GTX 960M.

The clock difference also reflects binning. Both chips draw from the same 1,870 million transistor pool, but the GTX 960M is configured to run at a 1097 MHz base with a 1176 MHz boost, while the 945M is limited to 928 MHz base and 1020 MHz boost. The 169 MHz base clock gap translates to a 15% difference in raw clock speed, and the boost gap is 15% as well. Combined with the memory bandwidth advantage, the GTX 960M ends up with a 36.4% lead in the recorded OpenCL benchmark, which aligns with the aggregate effect of these architectural differences.

Neither GPU supports ray tracing cores or tensor cores, as these are pre-RTX Maxwell parts. The absence of those features is consistent across both. The 945M's later release date (October 2015 versus March 2015) did not bring any architectural revision; it is the same GM107 silicon with reduced clocks and slower memory. In practical terms, the architecture is identical, but the implementation differs enough to create two distinct performance classes. The GTX 960M is the fully realized version of GM107 in mobile form, while the 945M is a cut-down configuration that sacrifices both clock speed and memory bandwidth. The benchmark data reflects exactly that: a 36.4% performance gap in the only head-to-head test, and a 19% gap in average scores across all recorded benchmarks.