AMD Radeon R7 M380 vs NVIDIA GeForce GTX 870M Comparison

The NVIDIA GeForce GTX 870M and AMD Radeon R7 M380 are two end-of-life mobile graphics solutions from different architectural eras, and the benchmark data shows a clear, decisive winner. In the single head-to-head benchmark available (Geekbench OpenCL), the GTX 870M delivers a 35.6% higher score than the R7 M380, a substantial margin that underscores its superior compute capability. This performance gap is consistent with the broader specifications, where the NVIDIA part features a larger die, more shading units, and significantly higher memory bandwidth, making it the unequivocal choice for users prioritizing raw processing power over the AMD alternative.

Head-to-Head Benchmarks

The only direct benchmark comparison in the data is the Geekbench OpenCL test, which measures general-purpose compute performance across the GPU. Here, the NVIDIA GeForce GTX 870M scores 12,630 points, while the AMD Radeon R7 M380 trails significantly with 9,313 points. The delta is a decisive 35.6% in favor of the GTX 870M, a margin that indicates not just a minor edge but a class-level difference in compute throughput.

This result is further contextualized by each card's average benchmark score. The GTX 870M boasts an average score of 9,959, placing it in the 48th percentile of all GPUs. In contrast, the R7 M380's average score is 9,313, which lands it in the 46th percentile. While a two-percentile difference might seem modest in isolation, the head-to-head delta reveals the true nature of the gap, as the GTX 870M's average is pulled down by a weaker Geekbench Metal score of 7,288, a test the R7 M380 does not have a recorded result for.

Looking at the nearest rivals provides additional insight into where each card sits in the performance hierarchy. The GTX 870M's average score of 9,959 puts it just 0.5% behind the AMD Radeon Pro 5300M and 0.8% behind the NVIDIA Quadro K5100M. It is also 1.1% behind the AMD Radeon R9 M375, yet it manages to outperform the NVIDIA Quadro 6000 by 1.1%. This clustering within a tight band of roughly 1% suggests the GTX 870M is a very consistent performer relative to its peers.

The R7 M380, with its 9,313 average, finds itself in a similarly competitive neighborhood but at a lower tier. It is a mere 0.1% ahead of the NVIDIA GeForce GTX 850M and 0.2% ahead of the NVIDIA GeForce GTX 465. It also leads the NVIDIA GeForce GTX 960 by 0.4% and the AMD Radeon Vega 8 by a more comfortable 1%. These results show the R7 M380 competing effectively with a set of older or lower-tier NVIDIA parts, but it is clear from the head-to-head that it cannot match the compute output of the higher-end GTX 870M.

The disparity in raw compute is also reflected in theoretical throughput figures. The GTX 870M delivers 2.599 TFLOPS of FP32 performance, while the R7 M380 is limited to 1,171.2 GFLOPS. This is not a subtle difference; the NVIDIA part has more than double the floating-point capability. Similarly, the texture fill rate tells a stark story, with the GTX 870M achieving 108.3 GTexel/s compared to the R7 M380's 36.60 GTexel/s, a near three-to-one ratio. The pixel rate follows suit, with the GTX 870M managing 27.08 GPixel/s versus 14.64 GPixel/s for the AMD card. These numbers align directly with the 35.6% OpenCL victory, confirming that the NVIDIA architecture is simply more powerful in this matchup.

The Verdict

The data is unambiguous: the NVIDIA GeForce GTX 870M is the superior product for compute-heavy workloads. Its 35.6% lead in the OpenCL benchmark is a decisive verdict that outweighs any other consideration. The GTX 870M's higher average score, higher percentile ranking, and dramatically better theoretical throughput numbers all point to the same conclusion. For users running applications that leverage GPGPU acceleration, such as OpenCL-based rendering or scientific computing, the GTX 870M is the clear choice.

The AMD Radeon R7 M380, while a competent performer in its own right, is outclassed in this comparison. Its 9,313 OpenCL score is respectable and puts it in line with parts like the GTX 850M, but it is not in the same league as the GTX 870M. The R7 M380's 46th percentile ranking versus the GTX 870M's 48th is a more accurate representation of their relative standing than the raw scores alone, but even that modest gap understates the magnitude of the difference in the direct test.

There is no scenario in the data where the R7 M380 wins. It has zero wins in the head-to-head benchmarks, while the GTX 870M has one. This is not a case of trade-offs or differing strengths; it is a straightforward case of one GPU being significantly faster than the other in the available metrics. The GTX 870M should be the default pick for any application where compute performance is the primary criterion.

Where Each One Wins

The GTX 870M wins in every category defined by the benchmark data. Its sole head-to-head benchmark win is the Geekbench OpenCL test, and it also demonstrates superiority in various theoretical performance metrics. In terms of raw compute, the GTX 870M's 2.599 TFLOPS FP32 output is more than double the R7 M380's 1,171.2 GFLOPS, making it the clear winner for tasks like video encoding, physics simulations, or any workload that relies on parallel processing.

The R7 M380's only potential advantage lies in its memory capacity, with 4 GB of DDR3 compared to the GTX 870M's 3 GB of GDDR5. However, this is a narrow benefit. While having an extra gigabyte of VRAM can be useful for holding larger textures, the R7 M380's memory bandwidth of 32.00 GB/s is a severe bottleneck, especially when compared to the GTX 870M's 120.0 GB/s. The GDDR5 memory on the NVIDIA card is fundamentally faster, which is why it achieves nearly four times the bandwidth despite a narrower 192-bit bus versus the AMD's 128-bit bus. For gaming, the GTX 870M is likely to provide a smoother experience due to this bandwidth advantage, even if the R7 M380 can technically accommodate larger texture sets.

The R7 M380 does not win any benchmark in the provided data. Its performance profile is closely matched to the NVIDIA GeForce GTX 850M and GTX 465, which are older or lower-tier parts. Therefore, the use case for the R7 M380 is limited to scenarios where the 4 GB memory capacity is an absolute requirement and the lower bandwidth is acceptable, or where the user is constrained by a PCIe 3.0 x16 interface that the GTX 870M's MXM-B (3.0) bus does not use.

FAQ

Q: Which GPU has a higher Geekbench OpenCL score?

A: The NVIDIA GeForce GTX 870M scores 12,630, which is 35.6% higher than the AMD Radeon R7 M380's 9,313.

Q: How do their average benchmark scores compare?

A: The GTX 870M has an average score of 9,959, placing it in the 48th percentile, while the R7 M380 has an average of 9,313, placing it in the 46th percentile.

Q: What is the memory configuration difference?

A: The GTX 870M has 3 GB of GDDR5 memory on a 192-bit bus with 120.0 GB/s bandwidth, while the R7 M380 has 4 GB of DDR3 memory on a 128-bit bus with 32.00 GB/s bandwidth.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce GTX 870M has 1,344 shading units, whereas the AMD Radeon R7 M380 has 640.

Q: What are the theoretical fill rates for each card?

A: The GTX 870M has a pixel rate of 27.08 GPixel/s and a texture rate of 108.3 GTexel/s, while the R7 M380 has a pixel rate of 14.64 GPixel/s and a texture rate of 36.60 GTexel/s.

Q: How does the GTX 870M compare to the AMD Radeon Pro 5300M in average score?

A: The GTX 870M's average score of 9,959 is 0.5% lower than the AMD Radeon Pro 5300M's average of 10,013.

Architecture Differences

The two GPUs are built on different architectural designs, which explains their performance disparity. The NVIDIA GeForce GTX 870M is based on the Kepler architecture, implemented on a chip codenamed GK104. This is a large, complex die manufactured on a 28 nm process at TSMC, containing 3,540 million transistors on a 294 mm² die. The high transistor count and large die size contribute to its substantial compute capabilities, with a transistor density of 12.0M / mm².

The AMD Radeon R7 M380, conversely, uses the GCN 1.0 architecture with a chip codenamed Tropo. This is a much smaller design, also built on TSMC's 28 nm process, but it packs only 1,500 million transistors on a 123 mm² die, resulting in a slightly higher transistor density of 12.2M / mm². The smaller scale of the Tropo chip directly correlates with its lower performance figures, as it has fewer resources to draw upon.

The difference in core configuration is stark. The GTX 870M is equipped with 1,344 shading units, 112 texture mapping units (TMUs), and 24 raster operation pipelines (ROPs). In contrast, the R7 M380 has 640 shading units, 40 TMUs, and 16 ROPs. This configuration gives the NVIDIA chip a significant advantage in both pixel and texture processing, as reflected in the fill rates. The GTX 870M also supports DirectX 12 (11_0) and Vulkan 1.2.175, while the R7 M380 supports DirectX 12 (11_1) and Vulkan 1.2.170, though the practical difference in API support is minimal for most workloads.

Specification Differences

The specifications where the two cards differ are numerous and highlight the GTX 870M's overall advantage. The clock speeds are similar, with the GTX 870M running at a base of 941 MHz and a boost of 967 MHz, while the R7 M380 runs at 900 MHz base and 915 MHz boost. However, the memory clocks are vastly different, with the GTX 870M using 1250 MHz (5 Gbps effective) GDDR5 versus the R7 M380's 1000 MHz (2 Gbps effective) DDR3.

The memory subsystem is a key differentiator. The GTX 870M has a 192-bit memory bus providing 120.0 GB/s of bandwidth, while the R7 M380 has a 128-bit bus delivering only 32.00 GB/s. Although the R7 M380 has more memory capacity at 4 GB versus 3 GB, the GTX 870M's GDDR5 memory is fundamentally faster, and the bandwidth difference of 88 GB/s is a decisive factor in performance. The power consumption also differs, with the GTX 870M rated at 100 W TDP, while the R7 M380 has no TDP listed in the data.

The physical and interface specifications also differ. The GTX 870M uses an MXM-B (3.0) bus interface and is a slot-width MXM Module, whereas the R7 M380 uses a standard PCIe 3.0 x16 interface. The GTX 870M also has a release date of March 2014, while the R7 M380 was released in May 2015, making it a newer product despite its lower performance. The GTX 870M's FP32 performance is 2.599 TFLOPS, and its texture rate is 108.3 GTexel/s, compared to the R7 M380's 1,171.2 GFLOPS and 36.60 GTexel/s, respectively.