AMD Radeon R9 M380

AMD Radeon R9 M380 is a mobile GPU built on the 28 nm process at TSMC, using the GCN 2.0 architecture with the Strato chip. It packs 2,080 million transistors on a 160 mm² die, with a transistor density of 13.0M / mm². The card operates at a base clock of 900 MHz and a boost clock of 1000 MHz, while its 4 GB GDDR5 memory runs at 1500 MHz (6 Gbps effective) across a 128-bit bus, yielding 96.00 GB/s of bandwidth. In aggregate benchmarks, it scores 15,504 on average, placing it at the 57th percentile of all GPUs in the database. Its Geekbench Metal and OpenCL scores are 18,476 and 12,532, respectively.

Who Should Consider It

The R9 M380’s benchmark position suggests it is a mainstream performer, not a high-end part. Its 4 GB frame buffer and 96.00 GB/s bandwidth are modest by modern standards, and the 128-bit memory bus will likely constrain performance at very high resolutions. The pixel rate of 16.00 GPixel/s and texture rate of 48.00 GTexel/s further indicate that the card is best suited to typical gaming resolutions where fill-rate demands are moderate. Given that it sits within 2% of the GeForce GTX 1080 and GTX 690 in average score, the data shows it can hold its own against those desktop cards in synthetic workloads, but those cards are not necessarily representative of modern gaming performance. For users targeting mainstream 1080p-class gaming with medium to high settings, the R9 M380’s memory and compute resources are adequate. Those aiming for ultra-high resolutions or heavy texture filtering may find the 96.00 GB/s bandwidth and 16 ROPs limiting. The 768 shading units and 48 TMUs provide a balanced compute-to-texture ratio, but the card is not designed for extreme workloads. In short, the R9 M380 is a reasonable choice for a laptop or compact system where power efficiency and moderate performance are priorities, and where the user is not pushing beyond standard high-definition gaming.

Ray Tracing and Feature Set

The R9 M380 does not include dedicated ray tracing cores or tensor cores; both fields are null in the specification data. This is consistent with its GCN 2.0 architecture, which predates hardware-accelerated ray tracing. Instead, the card relies on traditional rasterization techniques. It supports DirectX 12 (feature level 12_0), OpenGL 4.6, and Vulkan 1.2.170. The DirectX 12 support allows access to modern rendering features such as bindless resources and asynchronous compute, but without hardware RT or DLSS-like tensor operations, the card cannot accelerate ray-traced effects or AI-based upscaling. Users who require those features will need to look at newer architectures. The API coverage is solid for its era, but the lack of tensor and RT hardware means the feature set is firmly rooted in the pre-RT generation.

Power and Cooling

The FACT PACK does not list a TDP, PSU recommendation, or power connector requirements for the R9 M380. As such, power consumption cannot be quantified from the available data. The only power-related specifications are the 28 nm process node and the 2,080 million transistor count, which hint at a mid-range power envelope typical of mobile GPUs from that generation. The card uses a PCIe 3.0 x16 interface, which is standard and draws power from the slot or a proprietary mobile connector, though no details are provided. Without a TDP figure, system integrators must rely on the laptop’s overall thermal design. The absence of a suggested PSU is unsurprising for a mobile part, as power delivery is handled by the host system. The card is end-of-life, so cooling solutions are not actively marketed, but a capable air cooler or the laptop’s existing thermal solution should suffice given the modest specifications.

How It Compares

Against the NVIDIA GeForce GTX 1080, the R9 M380 scores 15,504 versus 15,531, a 0.2% deficit. This is a negligible difference in synthetic benchmarks, though the GTX 1080 is a desktop part with far higher memory bandwidth and compute capacity in real-world scenarios. The delta is within the margin of error, but the R9 M380’s mobile nature likely means it will throttle under sustained load.

Compared to the NVIDIA GeForce GTX 660 Ti, the R9 M380 is 1.3% faster, with scores of 15,504 and 15,303 respectively. This places the R9 M380 slightly ahead of a card that was a mid-range desktop option in its day, suggesting the R9 M380 offers comparable or better performance in compute tasks.

The NVIDIA GeForce GTX 690 posts an average score of 15,730, which is 1.4% higher than the R9 M380. The GTX 690 was a dual-GPU flagship, so the R9 M380’s near-parity in this metric is notable, though the GTX 690’s actual gaming performance would vary widely depending on driver and game support.

Finally, the AMD Radeon Pro W5500 scores 15,786, 1.8% higher than the R9 M380. The W5500 is a professional workstation card, and the R9 M380 trails it by a small margin in this benchmark, indicating that the R9 M380 is not far behind even in a professional compute context.

Memory Subsystem

The R9 M380 is equipped with 4 GB of GDDR5 memory on a 128-bit bus, running at 1500 MHz (6 Gbps effective). This configuration yields a bandwidth of 96.00 GB/s. The 128-bit bus is narrow compared to many desktop cards, which often use 256-bit or wider interfaces. As a result, the memory bandwidth is a limiting factor for high-resolution textures and large frame buffers. At 4 GB, the capacity is sufficient for many games at standard settings, but it may be insufficient for titles that demand more than 4 GB of VRAM at high resolutions. The effective 6 Gbps speed is standard for GDDR5, but the narrow bus halves the potential throughput compared to a 256-bit design. For users who push beyond 1080p, the bandwidth bottleneck will become apparent in texture streaming and high-detail scenes. The 16.00 GPixel/s pixel rate and 48.00 GTexel/s texture rate are consistent with the memory bandwidth, further indicating that the card is not designed for extreme resolutions. Overall, the memory subsystem is adequate for its target segment but will not scale gracefully to 4K or high-refresh-rate gaming.

FAQ

Q: What is the FP32 compute performance of the R9 M380?

A: The FP32 performance is 1.536 TFLOPS, derived from 768 shading units at a boost clock of 1000 MHz.

Q: Does the R9 M380 support hardware ray tracing?

A: No. The card has no ray tracing cores or tensor cores, and its GCN 2.0 architecture predates hardware RT.

Q: Which graphics APIs are supported?

A: The card supports DirectX 12 (feature level 12_0), OpenGL 4.6, and Vulkan 1.2.170.

Q: What is the memory configuration?

A: It has 4 GB of GDDR5 memory on a 128-bit bus, with a bandwidth of 96.00 GB/s and an effective memory clock of 6 Gbps.

Q: When was the R9 M380 released?

A: The release date is 2015-05-04, and the production status is end-of-life.

Q: How does the R9 M380 compare to the GeForce GTX 1080 in average benchmark score?

A: The R9 M380 scores 15,504, which is 0.2% lower than the GTX 1080’s 15,531.