AMD Radeon R5 Graphics vs NVIDIA GeForce GTX 860M Comparison

Head-to-Head Benchmarks

The benchmark data shows a decisive performance gap between these two mobile graphics solutions. Across the two shared compute tests, the NVIDIA GeForce GTX 860M wins both, and the margins are substantial.

In Geekbench OpenCL, the GTX 860M scores 10,461 points against 5,183 for the AMD Radeon R5 Graphics. That is a delta of -50.5%, meaning the NVIDIA part delivers roughly twice the raw compute throughput in this workload. The AMD IGP is not close.

The Vulkan result is even more lopsided. The GTX 860M posts 9,648 points while the Radeon R5 manages only 2,582, a delta of -73.2%. This suggests the NVIDIA architecture handles the lower-level API overhead far more efficiently, or simply has enough shader and texture hardware to run away with the test.

Looking at the broader database averages, the GTX 860M sits at 2,967 average benchmark score, with nearest rivals including the GeForce GTX 750 Ti (0.5% ahead) and the GeForce 820A (0.5% behind). The Radeon R5 averages 3,883, which places it near the Quadro 2000 (0.4% behind) and the Quadro K2000D (0.9% behind). Interestingly, the Radeon R5 has a higher average score across all recorded benchmarks than the GTX 860M, despite losing both head-to-head tests. That is because the Radeon's benchmark pool is limited to two OpenCL and Vulkan runs, while the GTX 860M includes several DirectX and Passmark tests that drag its average down.

Percentile rankings tell a similar story of overall placement: the Radeon R5 sits at the 23rd percentile among all GPUs, while the GTX 860M is at the 19th percentile. Neither is a high-flying part, but the GTX 860M's individual compute wins are undeniable.

Where Each One Wins

The AMD Radeon R5 Graphics wins nowhere in the direct head-to-head comparisons. It has zero wins against the GTX 860M's two. That is the short version.

However, the Radeon R5 does hold a higher aggregate average benchmark score (3,883 vs 2,967), which is worth contextualizing. That average is skewed by the fact that the Radeon only has two recorded tests, both of which are compute-oriented. The GTX 860M has ten recorded tests, including Passmark DirectX 9, 10, 11, and 12 runs, plus G2D and G3D tests. Those older DirectX tests score very low (15, 23, 13, and 55 for DX10, DX11, DX12, and DX9 respectively), pulling the average down. So if a workload is purely OpenCL or Vulkan compute, the GTX 860M is the clear pick. If a workload is something else, the database does not have direct head-to-head evidence for the Radeon R5, and the GTX 860M's broader test suite at least shows it can handle a variety of legacy APIs.

For gaming, the GTX 860M has the hardware fundamentals: 1,152 shading units, 96 texture mapping units, and 16 ROPs. The Radeon R5 has 256 shading units, 16 TMUs, and only 4 ROPs. Even without a direct gaming benchmark, the pixel rate difference (21.96 GPixel/s vs 3.032 GPixel/s) and texture rate difference (87.84 GTexel/s vs 12.13 GTexel/s) point to one obvious conclusion: the GTX 860M will rasterize and texture far faster.

Architecture Differences

The two chips come from different foundries and different design philosophies. The AMD Radeon R5 is built on GCN 2.0 architecture, specifically the Spectre SL die, fabricated by GlobalFoundries on a 28 nm process. It is an integrated graphics processor (IGP) designed for the Kaveri generation of AMD APUs. The chip integrates 2,410 million transistors on a 245 mm² die, giving a transistor density of 9.8 million per square millimeter.

The NVIDIA GeForce GTX 860M uses the Kepler architecture, specifically the GK104 die, built by TSMC also on a 28 nm process. It packs 3,540 million transistors on a larger 294 mm² die, yielding a higher density of 12.0 million per square millimeter. Despite being a mobile part, it is a discrete-style GPU with its own memory.

Memory is a fundamental split. The Radeon R5 uses system shared memory, meaning its frame buffer is carved from the main system RAM. Bandwidth is system dependent, so performance scales with whatever DDR3 or DDR4 memory the host laptop has. The GTX 860M has 2 GB of dedicated GDDR5 on a 128-bit bus, delivering 80.00 GB/s of fixed bandwidth. That dedicated memory is a major advantage for sustained workloads.

Clock speeds: the GTX 860M runs at a base of 797 MHz with a boost up to 915 MHz, and its memory runs at 1250 MHz (5 Gbps effective). The Radeon R5 has no listed base or boost clock in the database, and its memory clock is simply "System Shared."

Compute resources differ wildly. The Radeon R5 has 256 shading units, 16 TMUs, and 4 ROPs, producing 388.1 GFLOPS of FP32 compute. The GTX 860M has 1,152 shading units, 96 TMUs, and 16 ROPs, producing 2.108 TFLOPS. That is over five times the FP32 throughput.

Power draw is another differentiator: the Radeon R5 is rated at 15 W TDP, while the GTX 860M is rated at 75 W TDP. The NVIDIA part consumes five times the power, which is expected given the performance gap. The Radeon is an IGP with no power connectors and motherboard-dependent display outputs; the GTX 860M has no power connectors either (it draws from the laptop's power delivery) and its display outputs are portable-device dependent.

API support: both support DirectX 12, but the Radeon R5 supports feature level 12_0 while the GTX 860M only supports 11_0. Both have OpenGL 4.6. Vulkan versions are close: 1.2.170 for AMD, 1.2.175 for NVIDIA. The GTX 860M also has a recorded Metal score (4,900), which the Radeon R5 does not.

FAQ

Q: Which GPU is faster in OpenCL compute?

A: The NVIDIA GeForce GTX 860M scores 10,461 in Geekbench OpenCL, more than double the AMD Radeon R5's 5,183. That is a 50.5% gap in favor of NVIDIA.

Q: Can the AMD Radeon R5 play modern games?

A: The data indicates limited capability. With only 4 ROPs and 3.032 GPixel/s pixel rate, plus system shared memory, the Radeon R5 is far below the GTX 860M in raw rasterization. The GTX 860M's 16 ROPs and 21.96 GPixel/s suggest it is the only one of the two suitable for any kind of gaming beyond light titles.

Q: Why does the AMD Radeon R5 have a higher average benchmark score than the GTX 860M?

A: The Radeon R5's average is 3,883 based on only two tests (OpenCL and Vulkan), both of which are compute-heavy. The GTX 860M's average is 2,967 based on ten tests, including several low-scoring Passmark DirectX runs that drag down its mean.

Q: Which GPU has better memory bandwidth?

A: The GTX 860M has 80.00 GB/s of dedicated GDDR5 bandwidth on a 128-bit bus. The Radeon R5 has no fixed bandwidth; it uses system shared memory, so bandwidth is system dependent and typically much lower.

Q: What is the power consumption difference?

A: The Radeon R5 is rated at 15 W TDP, the GTX 860M at 75 W TDP. The NVIDIA part uses five times more power, consistent with its much larger shader count and dedicated memory.

Q: Do both support the same DirectX version?

A: Both support DirectX 12, but the Radeon R5 supports feature level 12_0 while the GTX 860M only supports 11_0. That means the AMD part has a higher feature level for DX12, though the GTX 860M is far more performant.

The Verdict

The data is unambiguous: the NVIDIA GeForce GTX 860M is the superior performer in every direct head-to-head benchmark. It wins both OpenCL and Vulkan tests by enormous margins, and its hardware specifications (1,152 shading units vs 256, 96 TMUs vs 16, 16 ROPs vs 4, 2.108 TFLOPS vs 388.1 GFLOPS) leave no room for doubt.

Who should pick the GTX 860M? Anyone who needs actual compute performance in a laptop GPU, whether for gaming, OpenCL workloads, or Vulkan-based applications. The 2 GB GDDR5 memory and 80 GB/s bandwidth make it a functional mobile GPU for its era. The 75 W TDP is high for a laptop part but justified by the performance.

Who should pick the AMD Radeon R5? Only someone who has no choice, as it is an integrated GPU on a Kaveri APU. At 15 W TDP, it is efficient and fine for basic desktop tasks, but the benchmark data shows it is not competitive with the GTX 860M in any compute scenario. The 23rd percentile ranking versus the GTX 860M's 19th percentile is misleading: the Radeon's percentile is based on only two tests, and both are losses in the head-to-head.

The GTX 860M's nearest rivals in the database are the GTX 750 Ti (0.5% ahead), the GeForce 820A (0.5% behind), and the Quadro P600 (1.5% behind). The Radeon R5's nearest rivals are the Quadro 2000 (0.4% behind) and the Quadro K2000D (0.9% behind). Neither part is a performance king, but the GTX 860M occupies a higher tier of actual capability.

For a builder or buyer choosing between these two, the decision is simple: the GTX 860M wins on every measurable front except power draw and DX12 feature level. If the system must stay under 15 W, take the Radeon R5. If performance matters at all, take the GTX 860M.

Specification Differences

| Specification | AMD Radeon R5 Graphics | NVIDIA GeForce GTX 860M |

|---|---|---|

| Architecture | GCN 2.0 | Kepler |

| Process Node | 28 nm | 28 nm |

| Foundry | GlobalFoundries | TSMC |

| Transistors | 2,410 million | 3,540 million |

| Die Size | 245 mm² | 294 mm² |

| Transistor Density | 9.8M / mm² | 12.0M / mm² |

| Base Clock | Not listed | 797 MHz |

| Boost Clock | Not listed | 915 MHz |

| Memory Size | System Shared | 2 GB |

| Memory Type | System Shared | GDDR5 |

| Memory Bus Width | System Shared | 128 bit |

| Memory Bandwidth | System Dependent | 80.00 GB/s |

| Shading Units | 256 | 1,152 |

| TMUs | 16 | 96 |

| ROPs | 4 | 16 |

| Pixel Rate | 3.032 GPixel/s | 21.96 GPixel/s |

| Texture Rate | 12.13 GTexel/s | 87.84 GTexel/s |

| FP32 | 388.1 GFLOPS | 2.108 TFLOPS |

| TDP | 15 W | 75 W |

| Bus Interface | IGP | PCIe 3.0 x16 |

| DirectX | 12 (12_0) | 12 (11_0) |

| Vulkan | 1.2.170 | 1.2.175 |

| Release Date | 2014-09-16 | 2014-03-09 |