AMD Radeon R7 M380 vs NVIDIA GeForce GTX 960 Comparison

The Verdict

The benchmark data presents a clear split between these two end-of-life graphics products. The AMD Radeon R7 M380 and NVIDIA GeForce GTX 960 occupy nearly identical positions in the overall performance distribution, with the R7 M380 at the 46th percentile and the GTX 960 at the 45th percentile. Their average benchmark scores are remarkably close: 9313 for the AMD part versus 9273 for the NVIDIA part, a gap of only 0.4% in the GeForce's nearest-rival listing. However, the single head-to-head benchmark available — Geekbench OpenCL — tells a very different story, with the GTX 960 scoring 18925 against the R7 M380's 9313, a 50.8% advantage. This contradiction between aggregate scores and direct comparison suggests the two cards are built for different workloads and software environments.

From the data, the GTX 960 is the pick for users running OpenCL-accelerated applications that can fully utilize its larger shader array and GDDR5 memory bandwidth. The R7 M380 is the pick for users who need 4 GB of video memory, a lower power footprint (the GTX 960 lists a 120 W TDP while the R7 M380 has none listed), or who are constrained by a mobile form factor. The R7 M380's closest rival is the GTX 850M at a 0.1% delta, while the GTX 960's closest rival is the GTX 465 at -0.2%, indicating both cards sit in a dense cluster of similar-performing GPUs. For pure compute throughput, the GTX 960 wins decisively; for memory capacity and efficiency, the R7 M380 holds the edge.

Architecture Differences

The two GPUs come from fundamentally different architectural generations, though both are built on the same 28 nm TSMC process node. The AMD Radeon R7 M380 uses the Tropo chip based on GCN 1.0 architecture, part of the Gem System (R7 M300) generation. It packs 1,500 million transistors on a 123 mm² die, yielding a transistor density of 12.2M per mm². The NVIDIA GeForce GTX 960 uses the GM206 chip based on Maxwell 2.0 architecture, part of the GeForce 900 generation. It contains 2,940 million transistors on a 228 mm² die, with a slightly higher transistor density of 12.9M per mm². The GTX 960's die is nearly double the size and carries nearly double the transistor count.

The compute resources differ substantially. The R7 M380 has 640 shading units, 40 texture mapping units, and 16 raster operation pipelines. The GTX 960 has 1024 shading units, 64 TMUs, and 32 ROPs — exactly 60% more shaders and TMUs, and double the ROPs. Clock speeds also favor NVIDIA: the GTX 960 runs at 1127 MHz base and 1178 MHz boost, while the R7 M380 runs at 900 MHz base and 915 MHz boost. This combination of more units and higher clocks produces a massive throughput advantage for the GTX 960: 2.413 TFLOPS FP32 versus 1,171.2 GFLOPS, 75.39 GTexel/s versus 36.60 GTexel/s, and 37.70 GPixel/s versus 14.64 GPixel/s.

Memory architecture diverges sharply. The R7 M380 uses 4 GB of DDR3 on a 128-bit bus, running at 1000 MHz (2 Gbps effective) for 32.00 GB/s bandwidth. The GTX 960 uses 2 GB of GDDR5 on the same 128-bit bus, but at 1753 MHz (7 Gbps effective) for 112.2 GB/s bandwidth — 3.5 times higher memory bandwidth. The GTX 960 also lists a 120 W TDP, dual-slot cooling, a 1x 6-pin power connector, and a 300 W suggested PSU, while the R7 M380 lists none of these, consistent with a mobile-oriented part. API support differs as well: the GTX 960 supports DirectX 12 (12_1) and Vulkan 1.4, while the R7 M380 supports DirectX 12 (11_1) and Vulkan 1.2.170. Both support OpenGL 4.6. The GTX 960 also has a launch MSRP of 199 USD (stated once here) and measures 241 mm (9.5 inches) in length.

Where Each One Wins

The data indicates a clear division of labor. The NVIDIA GeForce GTX 960 wins on raw compute performance across every measured throughput metric. Its FP32 performance of 2.413 TFLOPS is more than double the R7 M380's 1,171.2 GFLOPS. Its texture rate of 75.39 GTexel/s is more than double the R7 M380's 36.60 GTexel/s. Its pixel rate of 37.70 GPixel/s is more than 2.5 times the R7 M380's 14.64 GPixel/s. In the only direct head-to-head benchmark, Geekbench OpenCL, the GTX 960 scores 18925 versus 9313, a 50.8% lead. This makes the GTX 960 the clear choice for OpenCL compute workloads, high-resolution texture rendering, and any task that benefits from high memory bandwidth (112.2 GB/s versus 32.00 GB/s).

The AMD Radeon R7 M380 wins on memory capacity and power characteristics. Its 4 GB of DDR3 is double the GTX 960's 2 GB of GDDR5, which matters for workloads that require large working sets but not necessarily high bandwidth. The R7 M380 lists no TDP, power connector, or PSU requirement, whereas the GTX 960 lists 120 W TDP, a 1x 6-pin connector, and a 300 W suggested PSU — indicating the AMD part is designed for lower-power mobile or compact systems. The R7 M380 also has a higher average benchmark score (9313 versus 9273) and a higher percentile ranking (46th versus 45th), though these differences are within noise at 0.4%. In the nearest-rival context, the R7 M380 edges out the GTX 960 by 0.4% in the NVIDIA card's own rival list, while the GTX 960 trails the R7 M380 by the same margin in the AMD card's list.

The GTX 960 also offers more modern API support: DirectX 12 (12_1) versus DirectX 12 (11_1), and Vulkan 1.4 versus Vulkan 1.2.170. This suggests better longevity for newer games and applications that leverage these features. However, the R7 M380's 4 GB frame buffer could prove more useful in texture-heavy scenarios at lower resolutions, even if the bus bandwidth limits overall throughput.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The AMD Radeon R7 M380 has a slightly higher average benchmark score of 9313 compared to the NVIDIA GeForce GTX 960's 9273, a difference of 0.4% in favor of the AMD part.

Q: How do the two cards compare in the Geekbench OpenCL test?

A: The GTX 960 scores 18925 in Geekbench OpenCL, while the R7 M380 scores 9313. The GTX 960 wins by 50.8%, which is the only direct head-to-head benchmark listed.

Q: What are the memory specifications for each card?

A: The R7 M380 has 4 GB of DDR3 memory on a 128-bit bus with 32.00 GB/s bandwidth. The GTX 960 has 2 GB of GDDR5 memory on a 128-bit bus with 112.2 GB/s bandwidth.

Q: Which GPU has more shading units and what is the clock speed difference?

A: The GTX 960 has 1024 shading units running at 1127 MHz base and 1178 MHz boost. The R7 M380 has 640 shading units running at 900 MHz base and 915 MHz boost.

Q: What are the API compatibility differences?

A: The GTX 960 supports DirectX 12 (12_1) and Vulkan 1.4, while the R7 M380 supports DirectX 12 (11_1) and Vulkan 1.2.170. Both support OpenGL 4.6.

Q: Which card is in a higher performance percentile overall?

A: The R7 M380 is at the 46th percentile of all GPUs, while the GTX 960 is at the 45th percentile. The R7 M380 also has a closer rival relationship with the GTX 850M (0.1% delta) than the GTX 960 has with any listed rival.

Head-to-Head Benchmarks

The only direct comparison available is the Geekbench OpenCL test, and it is lopsided. The NVIDIA GeForce GTX 960 scores 18925, while the AMD Radeon R7 M380 scores 9313. The delta is -50.8% from the AMD card's perspective, meaning the GTX 960 is more than twice as fast in this workload. This result is consistent with the architectural differences: the GTX 960 has 1024 shading units versus 640, a boost clock of 1178 MHz versus 915 MHz, and 112.2 GB/s memory bandwidth versus 32.00 GB/s. The OpenCL test appears to be heavily sensitive to both compute unit count and memory bandwidth, both of which favor the NVIDIA part dramatically.

However, the aggregate benchmark picture complicates this narrative. The R7 M380's average benchmark score of 9313 is higher than the GTX 960's 9273, despite the GTX 960's massive OpenCL advantage. This is likely because the R7 M380's score is composed solely of its Geekbench OpenCL result (9313), while the GTX 960's average includes multiple tests: Geekbench OpenCL at 18925, Geekbench Metal at 8773, Geekbench Vulkan at 9231, and 3DMark Steel Nomad DX12 at 162. The GTX 960's average is dragged down by its poor Metal and Vulkan scores relative to its OpenCL result, while the R7 M380 has no comparable low-scoring tests to dilute its average.

In the nearest-rival context, the R7 M380's closest competitor is the GTX 850M at 9302 (0.1% delta), followed by the GTX 465 at 9294 (0.2% delta), the GTX 960 at 9273 (0.4% delta), and the Radeon Vega 8 at 9221 (1% delta). The GTX 960's closest competitor is the GTX 465 at 9294 (-0.2% delta), then the GTX 850M at 9302 (-0.3% delta), the R7 M380 at 9313 (-0.4% delta), and the Radeon Vega 8 at 9221 (0.6% delta). Both cards sit within a 0.6% performance band of each other and their immediate rivals, suggesting that in mixed workloads, the two GPUs are effectively interchangeable, despite the GTX 960's dominance in pure OpenCL compute.

The win count reflects this split: the GTX 960 has 1 head-to-head win, while the R7 M380 has 0. Yet the R7 M380 holds a higher percentile rank (46 versus 45) and a higher average score (9313 versus 9273). The data suggests that the GTX 960 is the superior compute part for OpenCL-heavy applications, while the R7 M380's 4 GB memory capacity and lower power profile make it competitive in broader system contexts where the GTX 960's additional tests (Metal at 8773, Vulkan at 9231) underperform relative to its OpenCL peak. Users should weigh the specific benchmark workloads they care about: if OpenCL throughput is paramount, the GTX 960 is the clear winner; if balanced system performance and memory capacity matter more, the R7 M380 holds its own.