AMD Radeon R9 M380 vs NVIDIA GeForce GTX 950 Comparison

Where Each One Wins

The recorded data splits these two mobile and desktop graphics cards almost perfectly down the middle, with each claiming one decisive benchmark victory. The AMD Radeon R9 M380 wins the Geekbench Metal test by a massive margin, while the NVIDIA GeForce GTX 950 takes the Geekbench OpenCL test. This is not a case of one card being uniformly superior; instead, the winner depends entirely on the API and workload in question.

For users operating within Apple's Metal ecosystem, the AMD Radeon R9 M380 is the clear choice. Its Metal score of 18,476 dwarfs the NVIDIA card's 7,172, a 157.6% advantage. This is a dominant, almost generational gap in a single test. The data suggests that any Metal-centric application, such as creative suites or games optimized for Apple's framework, will see a substantial performance lead on the AMD part.

Conversely, for OpenCL workloads, which are common in cross-platform compute tasks, the NVIDIA GeForce GTX 950 is the superior option. It scores 15,662 against AMD's 12,565, a 19.8% lead. While not as lopsided as the Metal result, this is still a clear and meaningful advantage. Users running OpenCL-based rendering, physics simulation, or data processing will favor the NVIDIA card.

The average benchmark scores reflect this split. The AMD card's average of 15,521 places it in the 58th percentile of all GPUs, while the NVIDIA card's average of 13,189 places it in the 53rd percentile. The AMD card holds a higher overall standing, but the NVIDIA card is more consistent across the two recorded tests. The AMD card's average is inflated by its extreme Metal score, whereas the NVIDIA card shows less variance between its Metal and OpenCL results.

Looking at the nearest rivals from the database provides additional context. The AMD Radeon R9 M380's average score of 15,521 puts it within 0.2% of the NVIDIA GeForce GTX 1080 Ti (15,548) and 1% behind the AMD Radeon Pro W5500 (15,679). It also leads the NVIDIA GeForce RTX 2060 (15,290) by 1.5% and the NVIDIA GeForce GTX 580 (15,283) by 1.6%. This is remarkable company, indicating that the R9 M380, despite its age and mobile origins, delivers average compute performance in the same tier as much more powerful desktop cards.

The NVIDIA GeForce GTX 950's nearest rivals tell a different story. Its average of 13,189 is 0.8% behind the NVIDIA GeForce GTX 480 (13,300) and 1% behind the AMD Radeon Pro 555X (13,321). It is 0.9% ahead of the NVIDIA Tesla M2090 (13,075) and 1.2% ahead of the AMD FirePro M6100 (13,354). These rivals are all older or lower-tier parts, which suggests the GTX 950's average performance is more modest in absolute terms, even if its OpenCL showing is strong.

In summary, the R9 M380 is the winner for Metal-centric workflows, while the GTX 950 is the winner for OpenCL-centric workflows. Neither card is a general-purpose champion, but each has a clear domain of superiority.

FAQ

Q: Which card wins in Geekbench Metal performance?

A: The AMD Radeon R9 M380 wins decisively with a score of 18,476, which is 157.6% higher than the NVIDIA GeForce GTX 950's score of 7,172 in the same test.

Q: Which card wins in Geekbench OpenCL performance?

A: The NVIDIA GeForce GTX 950 wins the OpenCL test with a score of 15,662, beating the AMD Radeon R9 M380's score of 12,565 by 19.8%.

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

A: The AMD Radeon R9 M380 has an average benchmark score of 15,521, while the NVIDIA GeForce GTX 950 has an average of 13,189. The AMD card sits in the 58th percentile of all GPUs, and the NVIDIA card sits in the 53rd percentile.

Q: How does the AMD Radeon R9 M380 compare to the NVIDIA GeForce GTX 1080 Ti in average score?

A: The R9 M380's average score of 15,521 is effectively tied with the GTX 1080 Ti's average of 15,548, showing a negligible 0.2% difference in favor of the NVIDIA card.

Q: How does the NVIDIA GeForce GTX 950 compare to the NVIDIA GeForce GTX 480 in average score?

A: The GTX 950's average score of 13,189 is 0.8% lower than the GTX 480's average of 13,300, indicating the older card holds a slight edge in overall average compute performance.

Q: Does either card support DirectX 12?

A: Both cards support DirectX 12. The AMD Radeon R9 M380 supports version 12 (12_0), while the NVIDIA GeForce GTX 950 supports a slightly higher feature level of 12 (12_1).

Head-to-Head Benchmarks

The two recorded head-to-head benchmarks show a clear division of labor between these GPUs. The first test, Geekbench Metal, is a complete rout in favor of the AMD Radeon R9 M380. It scores 18,476 points, while the NVIDIA GeForce GTX 950 manages only 7,172. The delta percentage of 157.6% means the AMD card more than doubles the NVIDIA card's output in this specific API. This is not a marginal win; it is a categorical difference in Metal compute capability. Any application that relies heavily on Metal will see roughly two and a half times the performance on the AMD card.

The second test, Geekbench OpenCL, flips the script. Here, the NVIDIA GeForce GTX 950 scores 15,662, while the AMD Radeon R9 M380 scores 12,565. The delta percentage of 19.8% in favor of NVIDIA indicates a solid, though not overwhelming, lead. In OpenCL, the GTX 950 is about a fifth faster than the R9 M380. This is a meaningful margin for compute workloads, but it is nowhere near the scale of the AMD card's Metal advantage.

The implications of these results are stark. The AMD card's average benchmark score of 15,521 is dragged upward by its exceptional Metal performance. Without the Metal test, its OpenCL score alone would place it far lower. Conversely, the NVIDIA card's average of 13,189 is more balanced, with its OpenCL score being its strongest showing and its Metal score being its weakest. The data shows that the R9 M380 is a specialist, excelling in one API while being mediocre in another, whereas the GTX 950 is more versatile, performing respectably across both APIs.

When placed against the broader database, the AMD card's Metal score of 18,476 is its single most important data point. It allows the R9 M380 to compete with far more powerful cards in average terms, as evidenced by its 0.2% gap to the GTX 1080 Ti. The NVIDIA card's OpenCL score of 15,662 is its key asset, but it is not enough to lift its average above the AMD card's. The head-to-head results, therefore, do not declare a single winner; they declare a winner per API.

Specification Differences

The two cards differ on nearly every core specification, which explains their divergent benchmark behavior. The AMD Radeon R9 M380 is built on a 28 nm process at TSMC, with 2,080 million transistors on a 160 mm² die. The NVIDIA GeForce GTX 950 uses the same 28 nm process and foundry, but packs 2,940 million transistors on a larger 228 mm² die. The transistor density is nearly identical (13.0M per mm² for AMD versus 12.9M per mm² for NVIDIA), but NVIDIA uses the extra silicon area for a different configuration.

Clock speeds favor NVIDIA. The GTX 950 has a base clock of 1024 MHz and a boost clock of 1188 MHz, while the R9 M380 has a base of 900 MHz and a boost of 1000 MHz. Memory clocks also favor NVIDIA: 1653 MHz (6.6 Gbps effective) versus AMD's 1500 MHz (6 Gbps effective). This contributes to a bandwidth advantage for NVIDIA: 105.8 GB/s versus 96.00 GB/s, despite both cards using a 128-bit bus.

Memory capacity is a differentiator, with the AMD card offering 4 GB of GDDR5, double the NVIDIA card's 2 GB. For modern workloads that exceed 2 GB of frame buffer, this could be critical, though the benchmark data does not isolate this effect. Both cards have 768 shading units and 48 texture mapping units, but the ROP count differs: AMD has 16, while NVIDIA has 32. This directly impacts pixel fill rate, where NVIDIA leads with 38.02 GPixel/s versus AMD's 16.00 GPixel/s. Texture fill rate also favors NVIDIA at 57.02 GTexel/s versus 48.00 GTexel/s. Floating-point performance, or FP32, is higher on the NVIDIA card at 1.825 TFLOPS versus 1.536 TFLOPS.

Power and physical specifications are only recorded for the NVIDIA card. The GTX 950 has a TDP of 90 W, is a dual-slot card, requires a single 6-pin power connector, and has a suggested PSU of 250 W. Its length is 202 mm (8 inches). No such data exists for the AMD mobile card. Display outputs are also only recorded for NVIDIA: 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2. The AMD card's output configuration is not in the database.

The release dates differ by roughly three months, with the AMD card launching on May 4, 2015, and the NVIDIA card on August 19, 2015. Both are marked as end-of-life. The NVIDIA card has a recorded launch MSRP of 159 USD. The AMD card has no launch MSRP recorded. The API support differs slightly: both support DirectX 12 and OpenGL 4.6, but NVIDIA supports Vulkan 1.4, while AMD supports Vulkan 1.2.170.

Architecture Differences

The architectural split is fundamental. The AMD Radeon R9 M380 is based on the Strato chip using Graphics Core Next 2.0 (GCN 2.0) architecture, part of the Gem System generation within the R9 M300 series. The NVIDIA GeForce GTX 950 is based on the GM206 chip using Maxwell 2.0 architecture, part of the GeForce 900 generation. These are completely different design philosophies from the same era.

GCN 2.0 was designed with compute uniformity in mind, emphasizing a large number of simple cores that can be scheduled efficiently for parallel workloads. The R9 M380's 768 shading units and 48 TMUs reflect this approach. Its lower clock speeds and lower ROP count suggest a design optimized for sustained compute throughput rather than peak fill rates. The 4 GB memory capacity is also a boon for larger datasets, which may explain its strong Metal showing.

Maxwell 2.0, by contrast, focused on power efficiency and higher clock speeds per core. The GTX 950's higher base and boost clocks, along with its doubled ROP count, indicate a design aimed at rasterization and geometry throughput. Its 32 ROPs and 38.02 GPixel/s pixel rate are nearly 2.4 times that of the AMD card, which is a massive advantage for traditional gaming workloads that rely on pixel fill. The smaller 2 GB memory capacity is a limitation, but the higher memory clock partially compensates with greater bandwidth.

The transistor counts tell a story of design density. NVIDIA packs 2,940 million transistors into 228 mm², while AMD fits 2,080 million into 160 mm². The densities are nearly equal, but NVIDIA chooses to spend its transistors on more ROPs and higher clocks, while AMD spends them on a larger memory buffer and compute-oriented GCN architecture. The Vulkan support difference (1.4 for NVIDIA versus 1.2.170 for AMD) further indicates that NVIDIA's driver stack and architecture are more forward-compatible with modern low-level APIs.

The predecessor and successor lines also differ. The AMD card's predecessor is the Solar System series, and its successor is Polaris Mobile, indicating a mobile-focused lineage. The NVIDIA card's predecessor is GeForce 700, and its successor is GeForce 10, showing a desktop-focused trajectory. This architectural divergence directly explains the benchmark results: GCN 2.0 excels in Metal, Maxwell 2.0 excels in OpenCL and rasterization-heavy tasks.

The Verdict

The data presents a clear, if bifurcated, verdict. For users whose primary workload is Metal-based, the AMD Radeon R9 M380 is the only rational choice. Its Metal score of 18,476 is more than double the NVIDIA card's 7,172, and this single metric dominates its average score. The R9 M380 also offers 4 GB of memory, which is a practical advantage for modern texture-heavy applications, and its average score of 15,521 places it in the 58th percentile, above the NVIDIA card's 53rd percentile. Any Mac or Metal-centric system will see dramatically better compute performance with the AMD card.

For users whose workloads are OpenCL-based, the NVIDIA GeForce GTX 950 is the superior part. Its OpenCL score of 15,662 beats the AMD card's 12,565 by 19.8%, and its higher clock speeds, doubled ROPs, and greater bandwidth (105.8 GB/s versus 96.00 GB/s) make it a more balanced performer across a wider range of tasks. The GTX 950 also has a recorded TDP of 90 W and a suggested PSU of 250 W, making it a predictable desktop component. Its Vulkan 1.4 support is more modern than AMD's 1.2.170, which could matter for future software compatibility.

The tie in head-to-head wins (one each) is the honest summary. The AMD card is a Metal specialist with a higher average score and larger memory buffer. The NVIDIA card is an OpenCL and rasterization specialist with higher fill rates and better API versioning. Neither card is a universal winner. The choice hinges entirely on the software environment. If the database only recorded one test, the verdict would be one-sided, but with two tests, the data demands a split decision. The R9 M380 for Metal, the GTX 950 for OpenCL, and no single card for everything.