NVIDIA GeForce 945M vs NVIDIA GeForce GTX 960 Comparison

Head-to-Head Benchmarks

The recorded data includes one direct head-to-head benchmark between the NVIDIA GeForce GTX 960 and the NVIDIA GeForce 945M: Geekbench OpenCL. In this test, the GTX 960 scores 18,925 points, while the 945M scores 8,099 points. This gives the GTX 960 a commanding lead of 133.7% over the 945M, meaning it more than doubles the mobile chip's raw compute throughput in this OpenCL workload. The delta is substantial enough that it represents a generational gap in performance rather than a marginal difference.

Looking at broader database averages, the GTX 960 holds an average benchmark score of 9,273 across all recorded tests, placing it at the 45th percentile among all GPUs in the database. The 945M, by contrast, averages 8,099 and sits at the 42nd percentile. While the percentile gap appears modest, the raw score difference of 1,174 points is meaningful, and the OpenCL head-to-head shows the GTX 960's peak advantage is far larger than the average suggests.

The GTX 960's nearest rivals in the database include the NVIDIA GeForce GTX 465 (average score 9,294, or 0.2% higher), the NVIDIA GeForce GTX 850M (9,302, 0.3% higher), and the AMD Radeon R7 M380 (9,313, 0.4% higher). It also sits just above the AMD Radeon Vega 8 (9,221, 0.6% lower). This places the GTX 960 in a tight cluster where small score differences separate adjacent cards, yet it remains clearly ahead of the 945M, whose nearest rivals include the NVIDIA GRID K2 (8,080, 0.2% higher), the AMD Radeon R9 M360 (8,129, 0.4% lower), the NVIDIA GeForce GTX 650 Ti Boost (8,067, 0.4% higher), and the NVIDIA GeForce GTX 950M (8,135, 0.4% lower).

The takeaway is straightforward: the GTX 960 wins the only shared benchmark by a wide margin, and its average score is higher across the board even when factoring in tests where the 945M has no recorded result.

Architecture Differences

Both GPUs are built by NVIDIA and use a 28 nm process at TSMC, but they diverge significantly in their underlying designs. The GTX 960 uses the GM206 chip based on the Maxwell 2.0 architecture, while the 945M uses the GM107 chip based on the original Maxwell architecture. This architectural split is important: Maxwell 2.0 introduces features and optimizations not present in the first-generation Maxwell implementation found in the 945M.

The transistor counts differ accordingly. The GTX 960 packs 2,940 million transistors on a 228 mm² die, yielding a transistor density of 12.9 million transistors per square millimeter. The 945M is smaller in both absolute and relative terms: 1,870 million transistors on a 148 mm² die, for a density of 12.6 million per square millimeter. The GTX 960's larger chip gives it more room for compute resources, which directly translates into higher performance.

Compute unit counts reinforce this gap. The GTX 960 features 1,024 shading units, 64 texture mapping units, and 32 render output units. The 945M has 640 shading units, 40 TMUs, and only 16 ROPs. That means the GTX 960 has 60% more shading units, 60% more TMUs, and exactly double the ROPs of the 945M. The pixel rate tells the story clearly: the GTX 960 achieves 37.70 GPixel/s, while the 945M manages 16.32 GPixel/s. Texture rate is similarly lopsided at 75.39 GTexel/s versus 40.80 GTexel/s. Floating-point performance (FP32) is 2.413 TFLOPS for the GTX 960 versus 1,305.6 GFLOPS for the 945M, a difference of roughly 85%.

Clock speeds also favor the GTX 960. Its base clock is 1,127 MHz with a boost of 1,178 MHz, while the 945M runs at 928 MHz base and 1,020 MHz boost. Memory is another area of divergence. The GTX 960 uses 2 GB of GDDR5 on a 128-bit bus, with a memory clock of 1,753 MHz (7 Gbps effective) and bandwidth of 112.2 GB/s. The 945M also has 2 GB but uses DDR3 on the same 128-bit bus, with a 900 MHz memory clock (1,800 Mbps effective) and only 28.80 GB/s of bandwidth. This is a 74% reduction in memory bandwidth, which heavily impacts texture-heavy and high-resolution workloads.

Power and physical design also separate the two. The GTX 960 has a TDP of 120 W, requires a single 6-pin power connector, and is a dual-slot card measuring 241 mm (9.5 inches) in length. It connects via PCIe 3.0 x16 and offers display outputs of 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2. The 945M is a mobile MXM module with a 75 W TDP, no power connectors, and display outputs that are portable-device dependent. Its bus interface is MXM-B (3.0). The 945M also supports DirectX 12 at feature level 11_0, while the GTX 960 supports DirectX 12 at feature level 12_1, a meaningful difference for games leveraging higher-tier DirectX 12 features.

The Verdict

The data points to a clear winner for performance: the NVIDIA GeForce GTX 960. It wins the sole head-to-head benchmark by 133.7%, has a higher average score (9,273 versus 8,099), higher percentiles (45th versus 42nd), and superior specifications across every measurable compute category. The GTX 960's larger die, more transistors, higher clock speeds, GDDR5 memory, and doubled ROP count all contribute to its dominance.

However, context matters. The 945M is a mobile part with a 75 W TDP, designed for laptops where power and thermal constraints are tight. The GTX 960 is a desktop card with a 120 W TDP, requiring a 300 W suggested PSU and a dual-slot cooler. If the choice is between a desktop system and a laptop, the GTX 960 is the obvious pick for raw performance. But for users constrained to a laptop form factor, the 945M offers a lower-power alternative, albeit with significantly reduced compute capability.

The GTX 960 also benefits from a higher DirectX feature level (12_1 versus 11_0), which future-proofs it better for modern games that use advanced DirectX 12 features. The 945M, while still DirectX 12 compatible, does not reach the same feature tier.

FAQ

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

A: In the Geekbench OpenCL test, the GTX 960 scores 18,925, while the 945M scores 8,099, giving the GTX 960 a 133.7% advantage.

Q: What are the average benchmark scores for both GPUs?

A: The GTX 960 has an average benchmark score of 9,273, while the 945M averages 8,099.

Q: Which GPU has more memory bandwidth?

A: The GTX 960 has 112.2 GB/s of bandwidth using GDDR5, whereas the 945M has 28.80 GB/s using DDR3. Both have 2 GB of memory on a 128-bit bus.

Q: Do both GPUs support DirectX 12?

A: Yes, both support DirectX 12, but the GTX 960 supports feature level 12_1, while the 945M supports feature level 11_0.

Q: What is the TDP difference between the two?

A: The GTX 960 has a TDP of 120 W, while the 945M has a TDP of 75 W.

Q: Which GPU sits higher in the database percentile rankings?

A: The GTX 960 is at the 45th percentile of all GPUs, while the 945M is at the 42nd percentile.

Where Each One Wins

The GTX 960 wins decisively in all compute and memory metrics recorded in the database. It is the better choice for anyone who needs maximum performance in OpenCL workloads, as evidenced by the 133.7% delta in the head-to-head test. Its GDDR5 memory provides nearly four times the bandwidth of the 945M's DDR3, which matters for high-resolution textures, post-processing effects, and any bandwidth-sensitive application. The GTX 960's higher pixel rate (37.70 GPixel/s versus 16.32 GPixel/s) and texture rate (75.39 GTexel/s versus 40.80 GTexel/s) make it more capable for modern game rendering at higher settings.

The 945M's only advantage is efficiency and form factor. At 75 W, it consumes 37.5% less power than the GTX 960. It requires no external power connectors and ships as an MXM module, making it suitable for laptops where the GTX 960's dual-slot, 241 mm desktop card cannot physically fit. For users who must have a mobile solution and do not need high-end gaming, the 945M is the only one of the two that fits that use case. But within desktop or large-form-factor systems, the 945M offers no measurable benefit according to the data.

Specification Differences

The two GPUs differ in nearly every key specification. The GTX 960 uses the GM206 chip with Maxwell 2.0 architecture, while the 945M uses GM107 with the original Maxwell architecture. Both are 28 nm TSMC parts, but the GTX 960 has 2,940 million transistors on a 228 mm² die, versus 1,870 million transistors on a 148 mm² die for the 945M. Transistor density is nearly identical (12.9M vs 12.6M per mm²).

Clock speeds favor the GTX 960: base 1,127 MHz and boost 1,178 MHz, compared to 928 MHz base and 1,020 MHz boost for the 945M. Memory differs fundamentally: the GTX 960 uses GDDR5 at 1,753 MHz (7 Gbps effective) with 112.2 GB/s bandwidth, while the 945M uses DDR3 at 900 MHz (1,800 Mbps effective) with 28.80 GB/s bandwidth. Both have 2 GB memory and a 128-bit bus.

Compute resources are heavily skewed toward the GTX 960: 1,024 shading units, 64 TMUs, and 32 ROPs versus 640 shading units, 40 TMUs, and 16 ROPs. Pixel rate is 37.70 GPixel/s versus 16.32 GPixel/s. Texture rate is 75.39 GTexel/s versus 40.80 GTexel/s. FP32 performance is 2.413 TFLOPS versus 1,305.6 GFLOPS.

Power and physical specifications also diverge. The GTX 960 has a 120 W TDP, is dual-slot, requires one 6-pin power connector, and has a 300 W suggested PSU. The 945M has a 75 W TDP, is an MXM module, requires no power connectors, and has no suggested PSU listed. The GTX 960 uses PCIe 3.0 x16 and offers DVI, HDMI 2.0, and three DisplayPort 1.2 outputs. The 945M uses MXM-B (3.0) and has portable-device-dependent outputs. DirectX support differs as well: the GTX 960 supports feature level 12_1, while the 945M supports 11_0. Both support OpenGL 4.6 and Vulkan 1.4. The GTX 960 was released on January 21, 2015, with a launch MSRP of 199 USD, while the 945M was released on October 26, 2015. Both are end-of-life products.