NVIDIA GeForce GTX 960M vs NVIDIA GeForce MX330 Comparison

Where Each One Wins

The benchmark split between the NVIDIA GeForce GTX 960M and the NVIDIA GeForce MX330 is a clean one, with each card taking a single victory in the two recorded tests. The GTX 960M dominates in OpenCL compute workloads, while the MX330 takes the Vulkan graphics test. This is not a case of one card being broadly superior; it is a case of two different workloads favoring two different designs.

In Geekbench OpenCL, the GTX 960M records a score of 11045 against 7896 for the MX330, a 39.9% advantage. That is a substantial gap, and it reflects the GTX 960M's larger shader array and wider memory bus. For any application that leans on raw compute throughput, such as GPU-accelerated rendering, video encoding, or general-purpose compute, the GTX 960M is the clear choice. The data shows a decisive win in raw compute performance.

The MX330 wins the Vulkan test, scoring 9019 against 8245 for the GTX 960M, an 8.6% advantage. Vulkan is a lower-level API that allows games and applications to manage GPU resources more directly, and the MX330's newer Pascal architecture, released five years later, handles it better. For modern games or applications that use Vulkan, the MX330 is the stronger pick. It is not a massive win, but it is a consistent one in that specific API.

The average benchmark score across both tests puts the GTX 960M at 9645, which places it in the 46th percentile of all GPUs in the database. The MX330 averages 8458, sitting in the 43rd percentile. So on average, the GTX 960M is the higher-performing card overall, but the MX330 is not far behind, and in Vulkan workloads it is actually ahead. The use case dictates the winner.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 960M has an average benchmark score of 9645, while the NVIDIA GeForce MX330 averages 8458. The GTX 960M also sits higher in the overall GPU percentile ranking at 46 versus 43 for the MX330.

Q: How large is the gap in OpenCL performance?

A: In Geekbench OpenCL, the GTX 960M scores 11045 and the MX330 scores 7896. That is a 39.9% difference in favor of the GTX 960M.

Q: Does the MX330 win any benchmark?

A: Yes, the MX330 wins the Geekbench Vulkan test with a score of 9019 against 8245 for the GTX 960M, an 8.6% advantage.

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

A: The GTX 960M sits within 0.7% of the NVIDIA Quadro P4000 and Tesla C2070, and 0.1% of the Quadro K5000, with the AMD Radeon Pro WX 2100 0.1% ahead. The MX330 is essentially tied with the AMD Radeon HD 8870M, sits 0.3% behind the AMD Radeon 880M, 0.4% ahead of the GeForce GTX 675MX, and 1.2% behind the Intel Arc A380.

Q: Which card supports the newer DirectX version?

A: The MX330 supports DirectX 12 (12_1), while the GTX 960M supports DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4.

Q: What is the transistor density difference?

A: The GTX 960M has a transistor density of 12.6 million transistors per square millimeter, while the MX330 has 24.3 million per square millimeter. The MX330 is built on a more modern 14 nm process from Samsung, versus 28 nm from TSMC for the GTX 960M.

Head-to-Head Benchmarks

The two recorded head-to-head benchmarks paint a clear picture of divergent strengths. In Geekbench OpenCL, the GTX 960M posts 11045 points, a 39.9% lead over the MX330's 7896. This is the largest margin in either direction between the two cards. The GTX 960M's 640 shading units, 40 texture mapping units, and 128-bit memory bus with 80.19 GB/s of bandwidth are simply too much for the MX330's 384 shaders, 24 TMUs, and 64-bit bus with 56.06 GB/s. In compute-heavy tasks, the MX330 is outmatched.

The Vulkan test flips the result. The MX330 scores 9019, beating the GTX 960M's 8245 by 8.6%. This is a smaller margin, but it is meaningful. The MX330's Pascal architecture, with its higher base clock of 1531 MHz and boost of 1594 MHz, outperforms the GTX 960M's Maxwell design, which runs at 1097 MHz base and 1176 MHz boost. The MX330 also has a higher pixel rate at 25.50 GPixel/s versus 18.82 GPixel/s for the GTX 960M, despite having the same 16 ROPs. The newer architecture extracts more efficiency from fewer resources.

The texture rate tells a different story. The GTX 960M achieves 47.04 GTexel/s, while the MX330 manages 38.26 GTexel/s, a 22.9% advantage for the older card. FP32 compute is also higher on the GTX 960M at 1.505 TFLOPS versus 1,224.2 GFLOPS for the MX330. The GTX 960M's raw compute muscle is apparent in every metric except pixel rate and Vulkan performance.

Memory capacity and bandwidth also favor the GTX 960M. It has 4 GB of GDDR5 versus 2 GB for the MX330, and bandwidth of 80.19 GB/s versus 56.06 GB/s. The MX330 compensates with faster effective memory speed at 7 Gbps versus 5 Gbps, but the narrower 64-bit bus limits total throughput. For large textures or compute buffers, the GTX 960M's memory advantage is significant.

Specification Differences

The two cards differ in nearly every core specification. The GTX 960M uses the GM107 chip on a 28 nm process from TSMC, while the MX330 uses the GP108B chip on a 14 nm process from Samsung. The GTX 960M has 640 shading units, 40 TMUs, and 16 ROPs. The MX330 has 384 shading units, 24 TMUs, and 16 ROPs. The GTX 960M has more of everything except ROPs, where they are equal.

Clock speeds favor the MX330. It runs at 1531 MHz base and 1594 MHz boost, while the GTX 960M runs at 1097 MHz base and 1176 MHz boost. The MX330's memory clock is also higher at 1752 MHz with 7 Gbps effective speed, versus 1253 MHz with 5 Gbps effective for the GTX 960M.

Memory configuration strongly favors the GTX 960M. It has 4 GB of GDDR5 on a 128-bit bus with 80.19 GB/s bandwidth. The MX330 has 2 GB of GDDR5 on a 64-bit bus with 56.06 GB/s bandwidth. The GTX 960M has double the capacity and 43% more bandwidth.

Pixel and texture rates split between the two. The MX330 has a higher pixel rate at 25.50 GPixel/s versus 18.82 GPixel/s. The GTX 960M has a higher texture rate at 47.04 GTexel/s versus 38.26 GTexel/s. FP32 compute goes to the GTX 960M at 1.505 TFLOPS versus 1,224.2 GFLOPS. The MX330 has a listed FP16 rate of 19.13 GFLOPS with a 1:64 ratio, while the GTX 960M has no listed FP16 rate.

Power and physical design are completely different. The GTX 960M has a TDP of 75 W and uses an MXM Module slot width with an MXM-B (3.0) bus interface. The MX330 has a TDP of 10 W, uses an IGP slot width, and connects via PCIe 3.0 x4. Neither card requires power connectors. The GTX 960M is far more power-hungry, while the MX330 is designed for low-power integration in thin laptops.

Transistor counts are nearly identical at 1,870 million for the GTX 960M and 1,800 million for the MX330, but die sizes differ dramatically. The GTX 960M measures 148 mm², while the MX330 is just 74 mm². This gives the MX330 a transistor density of 24.3 million per mm² versus 12.6 million per mm² for the GTX 960M.

Architecture Differences

The GTX 960M is built on the Maxwell architecture, released in 2015 as part of the GeForce 900M generation. The MX330 uses the Pascal architecture, released in 2020 in the GeForce MX (3xx) generation. This five-year gap is the defining difference between the two GPUs.

The process node difference is substantial. The GTX 960M uses 28 nm from TSMC, while the MX330 uses 14 nm from Samsung. This smaller node allows the MX330 to pack 1,800 million transistors into a 74 mm² die, while the GTX 960M spreads 1,870 million transistors across 148 mm². The MX330's transistor density is nearly double that of the GTX 960M.

API support differs in DirectX version. The GTX 960M supports DirectX 12 (11_0), while the MX330 supports DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4. The newer Pascal architecture provides better low-level API performance, which is reflected in the MX330's Vulkan win.

The MX330's Pascal architecture also delivers higher pixel throughput despite fewer resources. It achieves 25.50 GPixel/s with 16 ROPs, while the GTX 960M achieves 18.82 GPixel/s with the same number of ROPs. This indicates a more efficient ROP design in the newer architecture.

Neither GPU has ray tracing cores or tensor cores. Both are traditional rasterization designs. The GTX 960M is listed as end-of-life with a predecessor in the GeForce 800M and a successor in the GeForce 10 Mobile. The MX330 is also end-of-life but has no listed predecessor or successor.

The physical form factors reflect their intended use. The GTX 960M is an MXM module, a removable graphics card for laptops. The MX330 is an IGP, an integrated graphics processor. The MX330's 10 W TDP makes it suitable for ultraportable designs, while the GTX 960M's 75 W TDP requires a larger chassis with more cooling.

The Verdict

The data supports a clear split decision. For compute-heavy workloads, OpenCL applications, and tasks that benefit from more memory and bandwidth, the NVIDIA GeForce GTX 960M is the stronger GPU. Its 39.9% OpenCL advantage, 4 GB of VRAM, 128-bit bus, and 1.505 TFLOPS of FP32 compute make it the choice for GPU-accelerated work.

For Vulkan-based workloads and modern gaming, the NVIDIA GeForce MX330 is the better option. Its 8.6% Vulkan lead, higher clocks, and newer Pascal architecture deliver better performance in that specific API. The MX330 also consumes far less power at 10 W versus 75 W, making it practical for thin laptops where the GTX 960M's MXM module and power draw would be impractical.

The average benchmark scores place the GTX 960M ahead overall at 9645 versus 8458, and it ranks higher in the 46th percentile versus 43rd. The GTX 960M's nearest rivals are within 0.7% of its average score, while the MX330's closest rival, the AMD Radeon HD 8870M, is exactly at 0% delta. Both cards are mid-pack performers in the database.

The choice comes down to workload and platform. If the requirement is maximum compute performance and memory capacity in a laptop that can accommodate an MXM module, the GTX 960M is the answer. If the requirement is a low-power integrated GPU with better Vulkan performance and modern API support, the MX330 is the correct pick. Neither card dominates the other across the full range of recorded benchmarks.