AMD Radeon R9 M380 vs NVIDIA GeForce RTX 3070 Comparison

FAQ

Q: What is the average benchmark score for each GPU in the database?

A: The NVIDIA GeForce RTX 3070 records an average benchmark score of 17,208, while the AMD Radeon R9 M380 averages 15,521.

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

A: The RTX 3070 scores 112,821, which is 797.9% higher than the R9 M380's 12,565. This is the only head-to-head benchmark recorded between them.

Q: What percentile rank does each GPU hold among all GPUs?

A: The RTX 3070 sits at the 61st percentile, while the R9 M380 sits at the 58th percentile.

Q: Which GPU has the higher pixel rate and texture rate?

A: The RTX 3070 has a pixel rate of 165.6 GPixel/s and a texture rate of 317.4 GTexel/s; the R9 M380 has 16.00 GPixel/s and 48.00 GTexel/s, respectively.

Q: What is the process node for each chip?

A: The RTX 3070 uses an 8 nm process at Samsung, while the R9 M380 uses a 28 nm process at TSMC.

Q: What memory configurations do the two cards carry?

A: The RTX 3070 has 8 GB of GDDR6 on a 256-bit bus with 448.0 GB/s bandwidth; the R9 M380 has 4 GB of GDDR5 on a 128-bit bus with 96.00 GB/s bandwidth.

Architecture Differences

The NVIDIA GeForce RTX 3070 is built on the GA104 chip using the Ampere architecture, fabricated on an 8 nm process at Samsung. It integrates 17,400 million transistors across a 392 mm² die, yielding a transistor density of 44.4 million per mm². In contrast, the AMD Radeon R9 M380 uses the Strato chip with GCN 2.0 architecture, fabricated on a 28 nm process at TSMC. That chip contains 2,080 million transistors on a 160 mm² die, giving a density of 13.0 million per mm².

The RTX 3070 belongs to the GeForce 30-series and the GeForce 30 generation, with the GeForce 20 series as its predecessor and GeForce 40 as its successor. The R9 M380 is part of the Gem System (R9 M300) generation, with the Solar System as its predecessor and Polaris Mobile as its successor. Both are listed as end-of-life products.

The RTX 3070 includes 5,888 shading units, 184 texture mapping units, and 96 ROPs. It also carries 46 ray tracing cores and 184 tensor cores, features entirely absent from the R9 M380. The R9 M380 has 768 shading units, 48 TMUs, and 16 ROPs, with no RT or tensor core support.

Clock speeds differ substantially. The RTX 3070 runs at a base clock of 1500 MHz and boosts to 1725 MHz, with memory clocked at 1750 MHz (14 Gbps effective). The R9 M380 runs at 900 MHz base and 1000 MHz boost, with memory at 1500 MHz (6 Gbps effective). The FP32 compute throughput is 20.31 TFLOPS for the RTX 3070 versus 1.536 TFLOPS for the R9 M380, a factor of more than 13. The RTX 3070 also supports FP16 at 20.31 TFLOPS in a 1:1 ratio, while the R9 M380 has no recorded FP16 capability.

API support also differs. The RTX 3070 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The R9 M380 supports DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170. The RTX 3070 uses a PCIe 4.0 x16 interface, while the R9 M380 uses PCIe 3.0 x16. The RTX 3070 has display outputs of 1x HDMI 2.1 and 3x DisplayPort 1.4a; the R9 M380 has no recorded display outputs.

Head-to-Head Benchmarks

The database contains a single head-to-head benchmark between these two cards: Geekbench OpenCL. In that test, the NVIDIA GeForce RTX 3070 scores 112,821, while the AMD Radeon R9 M380 scores 12,565. The delta is 797.9% in favor of the RTX 3070. That is a decisive margin, far beyond any single-generation improvement; it reflects a complete gap in compute architecture, memory bandwidth, and shading resources.

To contextualize the RTX 3070's OpenCL result, its nearest rivals in the database include the AMD Radeon RX 7600 XT with an average score of 17,083 (0.7% lower), the NVIDIA GeForce GTX 690 at 17,037 (1% lower), the AMD Radeon HD 7970M at 17,019 (1.1% lower), and the NVIDIA Tesla K40c at 17,468 (1.5% higher). The RTX 3070's average benchmark score of 17,208 places it just above the RX 7600 XT and GTX 690, and just below the Tesla K40c. In OpenCL specifically, the RTX 3070 is dramatically ahead of its own average, suggesting that test heavily favors its compute resources.

For the R9 M380, its nearest rivals in the database are the NVIDIA GeForce GTX 1080 Ti with an average score of 15,548 (0.2% higher), the AMD Radeon Pro W5500 at 15,679 (1% higher), the NVIDIA GeForce RTX 2060 at 15,290 (1.5% lower), and the NVIDIA GeForce GTX 580 at 15,283 (1.6% lower). The R9 M380's average of 15,521 sits within a narrow band of these cards, all within roughly 1.6% of each other. Its OpenCL score of 12,565, however, is well below its average, indicating that the R9 M380 performs better in other recorded tests, such as Geekbench Metal where it scores 18,476.

The overall benchmark wins tally is 1 for the RTX 3070 and 0 for the R9 M380, reflecting the single head-to-head test result. The magnitude of that win, however, is so large that it dominates any comparison. A 797.9% advantage in OpenCL is not a marginal edge; it is an order-of-magnitude difference in raw compute throughput. The RTX 3070's FP32 output of 20.31 TFLOPS versus 1.536 TFLOPS for the R9 M380 corroborates this gap. The RTX 3070 also has 448.0 GB/s of memory bandwidth versus 96.00 GB/s, which directly feeds its compute advantage.

Specification Differences

The two GPUs differ in nearly every measurable specification. Starting with the chip: the RTX 3070 uses GA104 (Ampere) on 8 nm, while the R9 M380 uses Strato (GCN 2.0) on 28 nm. Transistor count is 17,400 million versus 2,080 million, and die size is 392 mm² versus 160 mm². Transistor density is 44.4M per mm² versus 13.0M per mm².

Memory differs completely: 8 GB GDDR6 on a 256-bit bus with 448.0 GB/s bandwidth versus 4 GB GDDR5 on a 128-bit bus with 96.00 GB/s bandwidth. Memory clock is 1750 MHz (14 Gbps effective) versus 1500 MHz (6 Gbps effective). The shading unit count is 5,888 versus 768. TMUs are 184 versus 48. ROPs are 96 versus 16. The RTX 3070 has 46 RT cores and 184 tensor cores; the R9 M380 has none.

Pixel rate is 165.6 GPixel/s versus 16.00 GPixel/s. Texture rate is 317.4 GTexel/s versus 48.00 GTexel/s. FP32 compute is 20.31 TFLOPS versus 1.536 TFLOPS. The RTX 3070 supports FP16 at 20.31 TFLOPS (1:1); the R9 M380 has no recorded FP16 value. The RTX 3070 has a TDP of 220 W, while the R9 M380 has no TDP recorded. The RTX 3070 is dual-slot with a 1x 12-pin power connector and a suggested PSU of 550 W; the R9 M380 has no slot width, power connector, or PSU data. Bus interface is PCIe 4.0 x16 versus PCIe 3.0 x16. Display outputs are 1x HDMI 2.1 and 3x DisplayPort 1.4a for the RTX 3070; none are recorded for the R9 M380.

Dimensions also differ: the RTX 3070 is 242 mm long (9.5 inches) and 112 mm tall (4.4 inches); the R9 M380 has no recorded dimensions. The RTX 3070 was released on 2020-08-31 and has a launch MSRP of 499 USD. The R9 M380 was released on 2015-05-04 and has no recorded launch MSRP. The RTX 3070's APIs include DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4; the R9 M380 has DirectX 12 (12_0), OpenGL 4.6, and Vulkan 1.2.170.

The Verdict

The data is unambiguous. The NVIDIA GeForce RTX 3070 outperforms the AMD Radeon R9 M380 in the only recorded head-to-head benchmark by a margin of 797.9%. Its average benchmark score of 17,208 is higher than the R9 M380's 15,521. Its percentile rank among all GPUs is 61 versus 58. Every architectural metric measured supports the same conclusion: more shading units, more TMUs, more ROPs, more memory bandwidth, more compute throughput, and a more recent process node.

The R9 M380 is not without merit in its own context. Its average score of 15,521 places it within 1.6% of cards like the GTX 1080 Ti, the Radeon Pro W5500, the RTX 2060, and the GTX 580, all of which are far more recent or higher-tier parts. That suggests the R9 M380, despite its age, holds a reasonable position in the database's overall ranking. Its Geekbench Metal score of 18,476 shows it can perform well in specific API workloads.

However, when directly compared to the RTX 3070, the gap is too large to bridge. The RTX 3070 is ahead in every recorded category, and the single head-to-head test shows a near nine-fold advantage. The verdict, strictly from the data, is that the RTX 3070 is the superior GPU for essentially any compute or graphics workload measured here. The R9 M380's lower memory bandwidth, older architecture, and smaller shader count place it in a different performance class entirely.

Where Each One Wins

The RTX 3070 wins in the Geekbench OpenCL test, the only direct comparison available. It also wins on every aggregate metric: average benchmark score, percentile rank, and all architectural specifications recorded. Its strengths lie in raw compute throughput, memory bandwidth, and modern API support. The 20.31 TFLOPS FP32 figure and 448.0 GB/s bandwidth make it suited for demanding workloads like ray tracing, tensor operations, and high-resolution gaming. The presence of 46 RT cores and 184 tensor cores further extends its capability into ray-traced rendering and AI-accelerated tasks, areas where the R9 M380 has no hardware support whatsoever.

The R9 M380, by contrast, has no recorded wins in any head-to-head or aggregate benchmark category against the RTX 3070. Its only recorded benchmark scores are Geekbench OpenCL at 12,565 and Geekbench Metal at 18,476. The Metal score is not part of the head-to-head comparison, but it does indicate that the R9 M380 can perform respectably in Apple-centric workloads. For users constrained to Metal-based environments, the R9 M380 may still be a functional option, though the database does not provide a direct RTX 3070 Metal score for comparison.

Where the R9 M380 might be considered is in scenarios where its lower resource footprint is relevant. It has no recorded TDP, slot width, or power connector data, so no power efficiency claims can be made from the database. Its 4 GB memory and 128-bit bus are smaller, which could matter for compatibility with older systems, but the data does not support any performance advantage from those traits. In every measured category, the RTX 3070 is the clear choice for users prioritizing compute performance, modern feature support, and benchmark leadership.