AMD 라이젠 AI 할로, 4천 달러대 AI 개발 미니 PC 공개

The AMD Ryzen AI Halo is a truly mini-PC built around the Zen 5 AMD Ryzen AI Max+ 395 processor (16 core, 32 thread) that streamlines learning AI development with ROCm or AMD hardware. The Max+ 395 processor is equipped with AMD Radeon 8060S integrated graphics which will be doing most of the heavy lifting, and an NPU which historically doesn't do much, but we were finally able to use. AMD Ryzen AI Halo It comes in a single hardware configuration with a removable 2 TB M.2 SSD and 128 GB of unified LPDDR5x-8000 memory capable of 256 GB/s bandwidth. 2 TB is a good amount of storage to hoard local models, and 128 GB is certainly enough memory to load a couple reasonably sized models into memory while reserving some space for system operation. Spec AMD Ryzen AI Halo DGX Spark Mac Studio Framework Desktop CPU 16-core Zen 5 AMD Ryzen AI Max+ 395 20-core Arm (10P+10E) GB10 Up to 32-core (24P+8E) M3 Ultra Up to 16-core Zen 5 AMD Ryzen AI Max+ 395 GPU Integrated 40 RDNA 3.5 Compute Unit AMD Radeon 8060S 6144-CUDA-core GB20B Up to 80-core M3 GPU cores Up to Integrated 40 RDNA 3.5 Compute Unit AMD Radeon 8060S NPU/Tensor AMD XDNA 2 NPU 5th Generation Tensor Cores Up to 32-core Neural Engine AMD XDNA 2 NPU Memory 128 GB LPDDR5x-8000 unified memory 128 GB LPDDR5x unified memory Up to 512 GB unified memory Up to 128 GB LPDDR5x-8000 unified memory Memory Bandwidth 256 GB/s 273 GB/s Up to 819 GB/s 256 GB/s Storage 2 TB 4 TB Up to 16 TB Up to 16 TB 1 of 2 AMD Ryzen AI Halo Product Page The AI Halo can be purchased for $3,999.99 USD in a single hardware configuration, preloaded with either Windows 11 Pro or Linux. You are able to load your own OS on the system once you have it, but as far as we know AMD won’t be making the ‘factory’ Linux and Windows installs(packaged drivers, programs, and models) available. AMD has sent us the Linux version of the Halo which is running a custom AMD Linux distribution based on Debian 13.4. Check out the interactive CT scan of the Ryzen AI Halo below! View Interactive AMD Ryzen AI Halo CT Scan The Hardware Despite the marketing images presenting it as the size of a datacentre, the Halo is an incredibly small box with only a square 15 cm(6 in) footprint and at less than 5 cm(2 in) tall. It weighs 1.2 kg, but if you’re planning on putting this in your backpack then also consider the required 240 W power brick. 1 of 3 AMD Ryzen AI Halo on the Box The power button and all of the ports are on the back face of the chassis: four USB 3.2 Type-C ports, an HDMI 2.1 port, and a 10 GbE ethernet port . Besides the connectivity on the rear it features Wi-Fi 7 and Bluetooth 5.4. The USB Type-C port closest to the power button is dedicated to USB-C Power Delivery(PD) power input. 1 of 4 Rear of the Ryzen AI Halo There aren’t any clear affordances for stacking them, but the corner feet and air intakes on all sides should make it viable if you need Windows and Linux, or if you want to cluster them. The Halo contains two blower fans to draw air in through the top and sides of the case which is then blown through the heat sink and out the back. This box usually sits quietly, but can ramp up the fans to dissipate the 120 W TDP of the processor inside. View Interactive AMD Ryzen AI Halo CT Scan - Cooling The best feature is the white ring of light around the bottom of the case.(pulsing blue when asleep) It doesn’t cast much light, and it can be turned off, but it gives it a nice look without being gaudy. Ryzen AI Halo Light Bar Teardown Being a tightly integrated mini PC there isn’t too much to see inside but you only have to remove four bolts beneath the removable magnetic feet to lift off the bottom cover. 1 of 2 Inside the Bottom Cover of the Ryzen AI Halo The removable M.2 2280 SSD is easily accessible with no further dissection. Removing the top shell to expose the compute core only requires addressing a few more connections. 1 of 2 Inside the Bottom Cover of the Ryzen AI Halo - Annotated The core can be pulled out but there isn’t much else to be done. The bottom metal plate visible when first removing the case bottom is removable with four bolts, but we didn’t remove it so as to not mess with the thermal compound underneath. 1 of 3 Ryzen AI Halo Core The Performance The AMD Ryzen AI Max+ 395(Strix Halo) processor has been available since Spring 2025 and the Halo doesn’t offer anything new on that front . The 2 TB SSD and 128 GB of memory are as expected, but it’s all been seen before in other hardware like the Framework Desktop , Beelink GTR9 Pro , X+ Rival , and ACEMAGIC M1A PRO . We’ve run some benchmarks to show that it is capable of what is expected, but the main focus of this product is the ‘batteries included’ software which is covered in the next section. Test Method For previous “AI” specific hardware we’ve used MLPerf and Procyon to measure representative values of hardware performance, but as we get further into locally hosted LLMs and agentic workflows, we’re looking a little deeper. For this testing we’ve focused on using llama-bench, the benchmarking tool packaged with llama.cpp . We’re still exploring tests we can apply with llama-bench as well as other programs that can provide insight, so constructive feedback on test selection is very welcome! We’ve found that nearly every benchmark and benchmarking decision can be met with a “that only applies to these specific cases”. We can’t cover every angle, but for now we’ll at least contribute the tests we conducted on this hardware. LLM token per second performance can be extremely sensitive to compatibility differences, and have high variances; so consider results below with that in mind and consider results from multiple sources. llama-bench For those unfamiliar with llama-bench or llama.cpp itself, llama.cpp is an open source inference engine that can be used to load and run large language models(in the GGUF format). It has seen wide adoption for running LLMs because of its simplicity of operation and hardware compatibility. It takes the list of numbers constituting the LLM and orchestrates execution of the model, interfacing with hardware-specific drivers. llama-bench is one of the many command-line-interface(CLI) utilities that are packaged with–and support–llama.cpp. Along with many options and configurations, it conducts two main tests: prompt-processing(pp), and token generation(tg), also known as the pre-fill and decoding phases of inference. Prompt processing is the part of LLM inference where the LLM ‘reads through’ what the user has said to it, and token generation is when it begins outputting tokens back to the user. When used from the command line it will output results as shown below(markdown format), but we output to csv for more detailed results. llama-bench Markdown Results Format Default llama-bench Test The first test is the default llama-bench pp512/tg128 configuration, simulating a user providing 512 tokens, and the LLM generating 128 in response. We tested with Qwen 3.6 35B A3B(Q4_K_M) , Gemma 4 31B IT(IQ4_XS) , and GLM 4.7 Flash(Q8_0) models. These are 17-32 GB models that have been receiving a lot of attention recently. As with test parameters, preferred models are always changing, but these should give an idea of performance with the currently favoured LLMs. We tested the AMD Halo alongside a Framework Desktop (AI Max+ 395, 128 GB), M2 Ultra(76-core GPU) Mac Studio with 128 GB of unified memory, and M3 Ultra(80-core GPU) Mac Studio with 512 GB of unified memory. For the AI Halo and Framework Desktop we also used both the ROCm/HIP and Vulkan runtimes (backends). None of these devices are direct competitors, but they provide some context. Note that the owner of our company has invested in Framework. 1 of 2 llama-bench Default Test(Flash Attention enabled) The Apple Silicon Mac Studios outperform the AMD Ryzen AI Max+ 395 machines. This is likely primarily due to their much higher 800 GB/s memory bandwidth compared to only 256 GB/s for the Max+ 395. The prompt p