AMD Radeon 860M vs AMD Radeon 890M Comparison

FAQ

Q: How do the two GPUs compare in raw compute performance?

A: The AMD Radeon 890M delivers 5.939 TFLOPS FP32, which is 93.3% higher than the 3.072 TFLOPS of the AMD Radeon 860M. The 890M also has double the shading units (1024 vs 512), double the texture mapping units (64 vs 32), and double the render output units (32 vs 16).

Q: Which GPU has the higher boost clock speed?

A: The AMD Radeon 860M has a higher boost clock at 3000 MHz, while the AMD Radeon 890M boosts to 2900 MHz. However, the 890M's base clock is lower at 400 MHz compared to 600 MHz on the 860M.

Q: How do the benchmark scores compare between the two?

A: In Geekbench OpenCL, the 890M scores 37254 versus 22759 for the 860M, a 38.9% advantage. In Geekbench Vulkan, the 890M scores 40808 versus 30043, a 26.4% advantage. The 890M wins both head-to-head tests.

Q: What is the transistor count and die size difference?

A: The 890M uses a chip with 34,000 million transistors on a 233 mm² die, with a transistor density of 145.9M per mm². The 860M's transistor count and die size are listed as unknown in the database.

Q: Do both GPUs support the same graphics APIs?

A: Yes, both support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. They also share the same PCIe 4.0 x8 bus interface and the same 15 W TDP.

Q: Which GPU has better ray tracing capabilities?

A: The AMD Radeon 890M has 16 ray tracing cores, double the 8 ray tracing cores found on the AMD Radeon 860M. Both use the RDNA 3.5 architecture.

Architecture Differences

The AMD Radeon 860M and AMD Radeon 890M both use the RDNA 3.5 architecture and belong to the Navi III IGP (Strix Point Mobile) generation, but they are built on different chips. The 860M uses the Krackan Point chip, while the 890M uses the Strix Point chip. Both are fabricated on TSMC's 4 nm process node.

The 890M carries 34,000 million transistors on a 233 mm² die, giving it a transistor density of 145.9M per mm². The 860M's transistor count and die size are recorded as unknown, which means the database cannot confirm whether the physical chip dimensions differ beyond the architectural specifications.

The most significant architectural difference is the execution resource count. The 890M doubles the shading units to 1024, doubles the texture mapping units to 64, and doubles the render output units to 32. The ray tracing core count also doubles from 8 on the 860M to 16 on the 890M. This doubling of resources across all major pipeline stages directly explains the 890M's higher pixel rate of 92.80 GPixel/s versus 48.00 GPixel/s on the 860M, and its texture rate of 185.6 GTexel/s versus 96.00 GTexel/s.

Clock behavior differs between the two. The 860M runs a 600 MHz base clock and a 3000 MHz boost clock. The 890M has a lower 400 MHz base clock and a 2900 MHz boost clock. The 890M compensates for its lower clock speeds with its doubled resource count, achieving substantially higher throughput in every measured metric.

Both GPUs share the same memory architecture: system shared memory, system dependent bandwidth, and no dedicated VRAM. Both use a PCIe 4.0 x8 interface and have no power connectors, consistent with their 15 W TDP and IGP form factor. The display outputs on both are portable device dependent.

The Verdict

The benchmark data clearly favors the AMD Radeon 890M in every recorded head-to-head test. The 890M wins both Geekbench OpenCL and Geekbench Vulkan, with a 38.9% advantage in OpenCL and a 26.4% advantage in Vulkan. The 890M also delivers 5.939 TFLOPS FP32 versus 3.072 TFLOPS on the 860M, nearly double the compute throughput.

The 860M does hold a small clock speed advantage with a 3000 MHz boost versus 2900 MHz on the 890M, but this does not translate into a benchmark win in any recorded test. The 890M's doubled shading units, texture mapping units, render output units, and ray tracing cores overwhelm the 860M's clock advantage.

The 890M's average benchmark score of 9210 places it at the 45th percentile of all GPUs, with nearest rivals including the AMD Radeon Vega 8 (0.1% lower), NVIDIA GeForce GTX 960 (0.7% lower), NVIDIA GeForce GTX 465 (0.9% lower), and NVIDIA GeForce GTX 850M (1% lower). The 860M sits at the 72nd percentile with an average score of 26401, placing it near the NVIDIA GeForce MX550 (0.1% lower), NVIDIA GeForce RTX 5060 (0.3% higher), AMD Radeon RX 5700 XT 50th Anniversary (0.6% lower), and NVIDIA RTX A4000 (1.1% lower).

The 860M's higher percentile ranking reflects its stronger absolute benchmark scores in the database's overall ranking system. However, within this direct comparison, the 890M is the faster part. The 860M is the better choice for users prioritizing the higher boost clock, while the 890M is the clear pick for maximum compute and graphics throughput.

Specification Differences

| Specification | AMD Radeon 860M | AMD Radeon 890M |

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

| Chip | Krackan Point | Strix Point |

| Transistors | unknown | 34,000 million |

| Die Size | unknown | 233 mm² |

| Transistor Density | null | 145.9M / mm² |

| Base Clock | 600 MHz | 400 MHz |

| Boost Clock | 3000 MHz | 2900 MHz |

| Shading Units | 512 | 1024 |

| TMUs | 32 | 64 |

| ROPs | 16 | 32 |

| RT Cores | 8 | 16 |

| Pixel Rate | 48.00 GPixel/s | 92.80 GPixel/s |

| Texture Rate | 96.00 GTexel/s | 185.6 GTexel/s |

| FP32 | 3.072 TFLOPS | 5.939 TFLOPS |

| FP16 | 3.072 TFLOPS (1:1) | 5.939 TFLOPS (1:1) |

| Release Date | 2025-02-28 | 2024-07-14 |

The two GPUs share identical specifications for memory size, type, bus width, and bandwidth (all system shared or system dependent), TDP (15 W), slot width (IGP), power connectors (none), bus interface (PCIe 4.0 x8), display outputs (portable device dependent), and all three recorded APIs (DirectX 12 Ultimate, OpenGL 4.6, Vulkan 1.4). Both are active production parts with no successor listed, and both use the RDNA 3.5 architecture on TSMC's 4 nm process.

Head-to-Head Benchmarks

The head-to-head benchmark results show a decisive victory for the AMD Radeon 890M, which wins both recorded tests.

In Geekbench OpenCL, the 890M scores 37254 against 22759 for the 860M. This is a 38.9% delta, meaning the 890M delivers roughly two-thirds more compute throughput in this workload. The OpenCL test exercises general-purpose GPU compute, and the 890M's doubled shading units and texture units directly contribute to this substantial margin. The 860M's higher boost clock of 3000 MHz cannot offset the resource disadvantage.

In Geekbench Vulkan, the 890M scores 40808 against 30043 for the 860M, a 26.4% advantage. The Vulkan test measures graphics and compute performance through the Vulkan API, and while the margin is smaller than in OpenCL, the 890M still holds a commanding lead. The 890M's doubled render output units and ray tracing cores help drive this result, as Vulkan workloads often utilize these fixed-function units more heavily than OpenCL.

The 890M's pixel rate of 92.80 GPixel/s versus 48.00 GPixel/s on the 860M indicates a 93.3% advantage in fill rate. Its texture rate of 185.6 GTexel/s versus 96.00 GTexel/s shows the same proportional advantage. These theoretical throughput figures align with the observed benchmark deltas, confirming that the 890M's resource doubling translates into real performance gains.

The 860M does have a 100 MHz higher boost clock, but this represents only a 3.4% clock advantage over the 890M's 2900 MHz boost. The 890M's 100% increase in shading units, texture mapping units, and render output units far outweighs this small clock edge.

The database records zero wins for the 860M and two wins for the 890M in head-to-head testing. The 890M's average benchmark score of 9210 places it among older discrete GPUs like the GeForce GTX 960 and GTX 465, while the 860M's average of 26401 places it near the GeForce RTX 5060 and RTX A4000. This discrepancy between the head-to-head results and the overall percentile rankings suggests the two GPUs are measured in different benchmark pools, with the 860M benefiting from a more modern set of comparison points.