AMD Radeon R7 M380 vs NVIDIA GeForce GTX 675MX Comparison

Head-to-Head Benchmarks

The database records only one direct benchmark comparison between these two mobile GPUs: the Geekbench OpenCL test. In this run, the NVIDIA GeForce GTX 675MX scores 10,723 points, while the AMD Radeon R7 M380 scores 9,313 points. That gives the NVIDIA part a 13.1% advantage, a clear and decisive margin. The delta is not a razor-thin edge; it represents a substantial performance gap in compute workloads.

Looking at the broader benchmark landscape, the GTX 675MX also holds a higher raw average benchmark score of 8,427 across all recorded tests, compared to the R7 M380's 9,313, which is based on a single OpenCL entry. However, the average is skewed by the NVIDIA card's additional Metal benchmark score of 6,131, which drags its mean downward. The OpenCL result is the only apples-to-apples comparison, and there the NVIDIA card wins outright.

The percentile rankings tell a slightly different story. The R7 M380 sits at the 46th percentile among all GPUs, while the GTX 675MX sits at the 43rd percentile. This means the AMD part is positioned marginally higher in the overall distribution, even though it loses the direct head-to-head. The explanation lies in the rival sets. The R7 M380's nearest competitors are the GTX 850M (9,302, delta 0.1%), GTX 465 (9,294, delta 0.2%), GTX 960 (9,273, delta 0.4%), and Vega 8 (9,221, delta 1%). All of those scores are within 1% of the R7 M380, indicating a very tight cluster. The GTX 675MX, by contrast, faces rivals like the Radeon 880M (8,436, delta -0.1%), MX330 (8,458, delta -0.4%), HD 8870M (8,462, delta -0.4%), and R9 M375X (8,325, delta 1.2%). Its cluster is similarly tight but sits at a lower absolute level, which explains the percentile discrepancy.

The key takeaway from the head-to-head data is that the GTX 675MX is the faster card in raw OpenCL compute, but the R7 M380 is not far behind in relative terms, and its percentile placement suggests it competes with a slightly higher tier of GPUs overall. The 13.1% delta is meaningful, but not overwhelming.

FAQ

Q: Which GPU wins the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 675MX wins with a score of 10,723, compared to the AMD Radeon R7 M380's 9,313. The delta is 13.1% in NVIDIA's favor.

Q: How do the two cards rank among all GPUs?

A: The R7 M380 is at the 46th percentile, while the GTX 675MX is at the 43rd percentile. Despite losing the head-to-head, the AMD card ranks slightly higher in the overall distribution.

Q: What is the average benchmark score for each card?

A: The R7 M380 has an average benchmark score of 9,313 (based on one OpenCL result). The GTX 675MX has an average of 8,427, which includes both a 10,723 OpenCL score and a 6,131 Metal score.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce GTX 675MX has 960 shading units, while the AMD Radeon R7 M380 has 640.

Q: Which GPU has higher memory bandwidth?

A: The GTX 675MX has 115.2 GB/s of bandwidth, compared to the R7 M380's 32.00 GB/s. The NVIDIA card uses a 256-bit bus with GDDR5 memory, while the AMD card uses a 128-bit bus with DDR3.

Q: What is the pixel rate for each card?

A: The R7 M380 has a pixel rate of 14.64 GPixel/s, which is higher than the GTX 675MX's 13.08 GPixel/s. However, the NVIDIA card has a higher texture rate at 52.32 GTexel/s versus 36.60 GTexel/s.

Where Each One Wins

The data reveals a clear split: the NVIDIA GeForce GTX 675MX is the compute champion, while the AMD Radeon R7 M380 holds advantages in certain rasterization metrics. In the OpenCL benchmark, the GTX 675MX wins by 13.1%, which suggests it is better suited for general-purpose compute tasks, GPU-accelerated applications, and workloads that leverage raw shader throughput. Its 960 shading units versus 640, and its 80 texture mapping units versus 40, give it a substantial edge in parallel processing.

On the other hand, the R7 M380 wins in pixel rate, 14.64 GPixel/s versus 13.08 GPixel/s, which indicates it may fill pixels faster in certain rendering scenarios. This could translate to better performance in fill-rate-bound situations, such as high-resolution depth or stencil operations. The AMD card also has a higher transistor density (12.2M per mm² versus 12.0M per mm²) and a smaller die (123 mm² versus 294 mm²), which may imply better power efficiency per unit of silicon, though the database does not record a TDP for the R7 M380 to confirm this.

For gaming, the GTX 675MX's higher texture rate (52.32 GTexel/s) suggests it can handle texture-heavy scenes better. For compute, the NVIDIA card's FP32 output (1,255.7 GFLOPS) is slightly ahead of the AMD card's (1,171.2 GFLOPS), reinforcing its lead in raw number crunching. The R7 M380's win in pixel rate is the only direct benchmark metric where it beats the NVIDIA part, and it is a narrow one (14.64 versus 13.08, an 11.9% margin in AMD's favor). That is a real advantage, but it is isolated to a single rasterization stage.

The verdict on use cases: pick the GTX 675MX for OpenCL workloads, texture-heavy rendering, or any task that benefits from more shading units and higher bandwidth. Pick the R7 M380 if pixel throughput is the bottleneck, or if you prefer the higher percentile ranking and the closer alignment with GTX 850M and GTX 960-class performance.

Specification Differences

The two GPUs diverge sharply in their memory subsystems. The R7 M380 has 4 GB of DDR3 on a 128-bit bus, yielding 32.00 GB/s of bandwidth. The GTX 675MX has 2 GB of GDDR5 on a 256-bit bus, delivering 115.2 GB/s, a 3.6x bandwidth advantage. This is a fundamental difference: the NVIDIA card trades capacity for speed, while the AMD card offers more memory but at a fraction of the throughput.

Compute resources also differ significantly. The GTX 675MX has 960 shading units, 80 TMUs, and 32 ROPs. The R7 M380 has 640 shading units, 40 TMUs, and 16 ROPs. The NVIDIA card has 50% more shading units, double the TMUs, and double the ROPs. However, the R7 M380 achieves a higher pixel rate (14.64 GPixel/s) despite having half the ROPs, which suggests its clock speed compensates. The R7 M380 runs at a base clock of 900 MHz and boost of 915 MHz, while the GTX 675MX has no recorded base or boost clock in the database.

The memory clock also differs: the R7 M380 runs at 1000 MHz (2 Gbps effective), while the GTX 675MX runs at 900 MHz (3.6 Gbps effective). The effective data rate is higher on the NVIDIA card due to the GDDR5 interface.

The R7 M380 uses a PCIe 3.0 x16 bus interface, while the GTX 675MX uses an MXM-B (3.0) interface. The NVIDIA card is an MXM module and its display outputs are listed as "Portable Device Dependent," while the AMD card's display outputs are not recorded. The GTX 675MX has a TDP of 100 W, while the R7 M380 has no TDP listed. The NVIDIA card also has no power connectors, while the AMD card's power connectors are not specified.

Architecture Differences

The two GPUs represent different architectural generations from their respective vendors. The AMD Radeon R7 M380 uses the Tropo chip based on GCN 1.0 architecture, part of the "Gem System (R7 M300)" generation. The NVIDIA GeForce GTX 675MX uses the GK104 chip based on Kepler architecture, part of the "GeForce 600M" generation.

Both are built on a 28 nm process at TSMC, but the silicon differs dramatically. The GTX 675MX has 3,540 million transistors on a 294 mm² die, while the R7 M380 has 1,500 million transistors on a 123 mm² die. The NVIDIA chip is more than twice as large and packs more than twice the transistors, yet the transistor density is nearly identical: 12.0M per mm² for the NVIDIA chip versus 12.2M per mm² for the AMD chip.

The API support shows subtle differences. Both support DirectX 12, but the R7 M380 supports version 12 (11_1) while the GTX 675MX supports 12 (11_0). Both support OpenGL 4.6. For Vulkan, the R7 M380 supports 1.2.170, while the GTX 675MX supports 1.2.175, a slightly newer revision.

Neither GPU has ray tracing cores or tensor cores, as expected for these older mobile parts. The R7 M380 has no FP16 support recorded, and the GTX 675MX also has no FP16 support listed. The release dates differ: the R7 M380 was released on 2015-05-04, while the GTX 675MX was released on 2012-09-30, making the NVIDIA card nearly three years older.

The R7 M380's predecessor is "Solar System" and its successor is "Polaris Mobile." The GTX 675MX's predecessor is "GeForce 500M" and its successor is "GeForce 700M." Both are end-of-life products.

The Verdict

The benchmark data points to the NVIDIA GeForce GTX 675MX as the stronger GPU for compute-heavy tasks. It wins the only direct head-to-head test by 13.1%, has higher FP32 throughput (1,255.7 GFLOPS versus 1,171.2 GFLOPS), and offers 3.6x the memory bandwidth of the R7 M380. Its 960 shading units and 80 TMUs provide a clear advantage in parallel workloads that scale with shader count and texture throughput.

The AMD Radeon R7 M380, however, is not without merits. It holds a higher percentile ranking (46th versus 43rd), a higher pixel rate (14.64 GPixel/s versus 13.08 GPixel/s), and double the memory capacity (4 GB versus 2 GB). It also has a smaller die and higher transistor density, which could indicate better manufacturing efficiency.

The choice depends on the workload. For OpenCL compute, GPU-accelerated applications, or texture-heavy rendering, the GTX 675MX is the superior option based on the recorded scores. For fill-rate-bound scenarios or tasks that need more video memory, the R7 M380 is the better fit. The GTX 675MX's older release date (2012) versus the R7 M380's newer one (2015) does not appear to hurt the NVIDIA card's performance, as its architecture still delivers higher peak throughput.

In the end, the data suggests a nuanced verdict: the GTX 675MX is the faster card, but the R7 M380 is better positioned in the overall GPU hierarchy. If raw performance in the recorded benchmark is the only criterion, the NVIDIA part wins. If percentile standing and pixel throughput matter more, the AMD part has a case. The 13.1% delta in OpenCL is the single most decisive number, and it favors NVIDIA.