AMD Radeon 860M

AMD Radeon 860M
AMD Radeon 860M graphics card review

AMD Radeon 860M: New Architecture vs. 8 Graphics Units

The AMD Radeon 860M appears to be a significant update to AMD's integrated graphics: a fresh architecture, higher clock speeds, and the new Ryzen AI platform. However, in terms of configuration, it remains an 8-unit iGPU. Thus, it makes more sense to compare it with the Radeon 760M rather than the Radeon 780M or Radeon 890M.

For a thin laptop without a discrete graphics card, the Radeon 860M can handle basic tasks: user interface, video playback, light gaming, and some older AAA titles. However, the name should not create expectations at the level of the Radeon 890M or a discrete GPU.

What is Radeon 860M

The Radeon 860M is integrated graphics designed for the mobile Ryzen AI 7 processors. It is based on the RDNA 3.5 architecture and utilizes 8 graphics units, which means 512 shaders. Depending on the processor, the GPU clock speed reaches around 3000-3100 MHz.

The main difference from the Radeon 760M lies not in the number of units, but in the newer platform. The Radeon 860M receives a new architecture, higher clock speeds, and support for faster memory on laptops with Ryzen AI. However, the width of the GPU remains the same: it still features 8 integrated graphics units.

A GPU class cannot be determined by name alone. The Radeon 860M stands alongside the Radeon 880M and 890M only by index. In terms of unit count, it is closer to the Radeon 760M.

Position in AMD's Lineup

Integrated Graphics Architecture Graphics Units Role in Lineup
Radeon 760M RDNA 3 8 Previous 8-CU Class
Radeon 860M RDNA 3.5 8 New Mid-Range Version
Radeon 780M RDNA 3 12 Older but Wider iGPU
Radeon 880M RDNA 3.5 12 Higher Tier Ryzen AI
Radeon 890M RDNA 3.5 16 Top Integrated Graphics AMD

The Radeon 860M is not a junior version of the Radeon 890M, but rather an update to the previous 8-unit level. It features a newer core and higher clock speeds, but fewer execution units than the Radeon 780M, 880M, and 890M.

If a laptop is primarily needed for gaming, the GPU's name alone tells little about real-world performance. One must consider memory, power limits, cooling, and price in comparison to models with Radeon 780M or Radeon 890M.

The Radeon 780M is older in architecture, but it has 12 units instead of 8. In gaming and 3D benchmarks, this could be more significant than the transition from RDNA 3 to RDNA 3.5.

Gaming Performance

The Radeon 860M is designed for minimal gaming performance for iGPUs. It is suitable for esports games, indie titles, older projects, and less demanding AAA games. In modern heavy games, settings or resolution may need to be lowered, or upscaling turned on.

In the 3DMark Time Spy Graphics benchmark, the average scores for the Radeon 860M fall around 2400-2500 points. For integrated graphics, this is decent but does not reach the level of gaming laptops with discrete GPUs. This score offers room for occasional gaming rather than replacing a dedicated graphics card.

At Full HD, the Radeon 860M does not perform confidently in all games. Light projects can yield comfortable fps. In new heavy games, a more realistic approach is to look at 720p-900p, low settings, and FSR if supported.

The Radeon 860M is not meant to be purchased as gaming graphics. It is an integrated solution that allows gaming without a discrete graphics card, but only with clear limitations.

Why One Laptop with Radeon 860M May Be Faster Than Another

The Radeon 860M does not have its own video memory. It uses system RAM, so performance heavily depends on the laptop's configuration.

This graphics unit works best with fast LPDDR5X memory. Slow memory or single-channel modes can significantly reduce performance. For iGPUs, memory bandwidth is one of the main limiting factors.

The second factor is cooling and power limits. In a thin chassis, the CPU and GPU quickly hit thermal throttling. In a larger laptop, the Radeon 860M can maintain high clock speeds for longer durations. Therefore, sharing the same GPU name does not guarantee the same performance across different models.

How Radeon 860M Outperforms Older Integrations

For the Radeon 860M, gaming performance is not the only consideration. In a typical laptop, aspects like media handling, support for modern displays, energy efficiency, and video performance matter. Here, the new Ryzen AI platform is more advantageous than the old 8-unit iGPUs.

For browsing, video, office tasks, light photo editing, and simple video editing, the Radeon 860M is adequate. Limitations arise where stable high 3D performance is necessary: modern games, heavy video editing, 3D rendering, and GPU tasks.

Radeon 860M vs. Radeon 780M

The choice between Radeon 860M and Radeon 780M depends not just on the GPU. If a laptop with Radeon 860M is built on a newer Ryzen AI platform, has fast memory, and offers good battery life, it may be preferable as a modern solution.

However, if graphical performance is the priority, the Radeon 780M should not be discounted. It has 12 graphics units, and in well-optimized laptops, it may perform better than the Radeon 860M. The new name does not negate the physical differences in GPU configuration.

The situation with the Radeon 890M is clearer. This is AMD's higher-tier integrated graphics with 16 units. If a laptop is being purchased with gaming without a discrete graphics card in mind, one should indeed look towards the 890M rather than the 860M.

Conclusion

The AMD Radeon 860M is a modern mid-range integrated graphics solution. Its main advantages are the new Ryzen AI platform, RDNA 3.5 architecture, high clock speeds, and support for fast memory. The main disadvantage is the same 8 graphics units, which limit its performance in gaming and 3D benchmarks.

It is graphics for a versatile laptop with no discrete GPU: work, video, light gaming, older AAA projects, and basic content handling. The Radeon 860M is not a mini-890M, but an updated 760M for the new generation of Ryzen AI.

Basic

Label Name
Intel
Platform
Integrated
Launch Date
February 2025
Former Codename
Krackan Point / Gorgon Point
GPU Lithography
4 nm
Model Name
AMD Radeon 860M Graphics
Generation
Radeon 800M Series
Boost Clock
3000-3100 MHz
Bus Interface
Integrated
RT Cores
8
Compute Units
8
Tensor Cores
?
Tensor Cores are specialized processing units designed specifically for deep learning, providing higher training and inference performance compared to FP32 training. They enable rapid computations in areas such as computer vision, natural language processing, speech recognition, text-to-speech conversion, and personalized recommendations. The two most notable applications of Tensor Cores are DLSS (Deep Learning Super Sampling) and AI Denoiser for noise reduction.
No
TMUs
?
Texture Mapping Units (TMUs) serve as components of the GPU, which are capable of rotating, scaling, and distorting binary images, and then placing them as textures onto any plane of a given 3D model. This process is called texture mapping.
32
Foundry
TSMC
Process Size
4 nm
Architecture
RDNA 3.5

Memory Specifications

Memory Size
Shared system memory
Memory Type
System shared
Memory Bus
?
The memory bus width refers to the number of bits of data that the video memory can transfer within a single clock cycle. The larger the bus width, the greater the amount of data that can be transmitted instantaneously, making it one of the crucial parameters of video memory. The memory bandwidth is calculated as: Memory Bandwidth = Memory Frequency x Memory Bus Width / 8. Therefore, when the memory frequencies are similar, the memory bus width will determine the size of the memory bandwidth.
Dual-channel system memory, platform dependent
Memory Clock
System memory dependent

Display and Media

AMD FreeSync
Yes
AV1 Encode/Decode
Encode/Decode
DisplayPort Extensions
Adaptive-Sync, HDR Metadata, UHBR10
H.264 Hardware Encode/Decode
Encode/Decode
H.265 HEVC Hardware Encode/Decode
Encode/Decode
H.266 VVC Hardware Encode/Decode
No hardware support
HDCP Version
2.3
HDMI Version
2.1
Intel Quick Sync Video
No
Max Resolution DP
7680x4320 @ 60Hz
Max Resolution HDMI
7680x4320 @ 60Hz
Max Video Decode Bandwidth
1080p60 8bpc MPEG2, 1080p60 8bpc VC1, 1080p786 8/10bpc VP9, 2160p196 8/10bpc VP9, 4320p49 8/10bpc VP9, 1080p1200 8bpc H.264, 2160p300 8bpc H.264, 4320p75 8bpc H.264, 1080p786 8/10bpc H.265, 2160p196 8/10bpc H.265, 4320p49 8/10bpc H.265, 1080p960 8/10bpc AV1, 2160p240 8/10bpc AV1, 4320p60 8/10bpc AV1
Max Video Encode Bandwidth
1080p630 8bpc H.264, 1440p373 8bpc H.264, 2160p175 8bpc H.264, 1080p630 8bpc H.265, 1440p373 8bpc H.265, 2160p175 8bpc H.265, 4320p43 8bpc H.265, 1080p864 8/10bpc AV1, 1440p513 8/10bpc AV1, 2160p240 8/10bpc AV1, 4320p60 8/10bpc AV1
Number of Displays Supported
4
Outputs
HDMI 2.1, DisplayPort 2.1, USB-C DisplayPort Alt Mode; device dependent
USB Type-C DisplayPort Alternate Mode
Yes
Wireless Display
Miracast

Theoretical Performance

Pixel Rate
?
Pixel fill rate refers to the number of pixels a graphics processing unit (GPU) can render per second, measured in MPixels/s (million pixels per second) or GPixels/s (billion pixels per second). It is the most commonly used metric to evaluate the pixel processing performance of a graphics card.
48-49.6 GPixel/s
Texture Rate
?
Texture fill rate refers to the number of texture map elements (texels) that a GPU can map to pixels in a single second.
96-99.2 GTexel/s
FP16 (half)
?
An important metric for measuring GPU performance is floating-point computing capability. Half-precision floating-point numbers (16-bit) are used for applications like machine learning, where lower precision is acceptable. Single-precision floating-point numbers (32-bit) are used for common multimedia and graphics processing tasks, while double-precision floating-point numbers (64-bit) are required for scientific computing that demands a wide numeric range and high accuracy.
6.14-6.35 TFLOPS
FP64 (double)
?
An important metric for measuring GPU performance is floating-point computing capability. Double-precision floating-point numbers (64-bit) are required for scientific computing that demands a wide numeric range and high accuracy, while single-precision floating-point numbers (32-bit) are used for common multimedia and graphics processing tasks. Half-precision floating-point numbers (16-bit) are used for applications like machine learning, where lower precision is acceptable.
192-198 GFLOPS
FP32 (float)
?
An important metric for measuring GPU performance is floating-point computing capability. Single-precision floating-point numbers (32-bit) are used for common multimedia and graphics processing tasks, while double-precision floating-point numbers (64-bit) are required for scientific computing that demands a wide numeric range and high accuracy. Half-precision floating-point numbers (16-bit) are used for applications like machine learning, where lower precision is acceptable.
3.07 TFLOPS

AI Features

Intel Deep Learning Boost on GPU
No
NPU TOPS
Up to 50 TOPS
Processor Overall TOPS
Up to 66 TOPS

Miscellaneous

AMD SmartAccess Memory
Available
Native PCIe Lanes
14 total / 14 usable
PCI Express Version
PCIe 4.0
Shading Units
?
The most fundamental processing unit is the Streaming Processor (SP), where specific instructions and tasks are executed. GPUs perform parallel computing, which means multiple SPs work simultaneously to process tasks.
512
TDP
Shared with processor; 15-54 W cTDP
Vulkan Version
?
Vulkan is a cross-platform graphics and compute API by Khronos Group, offering high performance and low CPU overhead. It lets developers control the GPU directly, reduces rendering overhead, and supports multi-threading and multi-core processors.
1.4
OpenCL Version
2.1
OpenGL
4.6
CUDA
No
DirectX
12 Ultimate (12_2)
Power Connectors
None
ROPs
?
The Raster Operations Pipeline (ROPs) is primarily responsible for handling lighting and reflection calculations in games, as well as managing effects like anti-aliasing (AA), high resolution, smoke, and fire. The more demanding the anti-aliasing and lighting effects in a game, the higher the performance requirements for the ROPs; otherwise, it may result in a sharp drop in frame rate.
16
Shader Model
6.7

Benchmarks

FP32 (float)
Score
3.07 TFLOPS
3DMark Steel Nomad
Score
454.3
3DMark Time Spy
Score
2410
Vulkan
Score
29771
OpenCL
Score
23816

Compared to Other GPU

FP32 (float) / TFLOPS
3.3 +7.5%
3.196 +4.1%
2.935 -4.4%
2.86 -6.8%
3DMark Time Spy
4861 +101.7%
3754 +55.8%
2742 +13.8%
3DMark Steel Nomad
454.3
Vulkan
87196 +192.9%
60360 +102.7%
38421 +29.1%
29771
5522 -81.5%
OpenCL
61514 +158.3%
39179 +64.5%
23816
12186 -48.8%
6073 -74.5%