Mini PCs have been carving out a unique niche in the computing world over the past decade. What started as an experimental category dominated by Intel’s NUC line has grown into a competitive market with serious contenders from Geekom, Beelink, Minisforum, and even Apple’s Mac Mini. By 2025, these small form factor (SFF) systems are no longer novelties — they are viable desktop replacements for professionals, gamers, students, and AI enthusiasts alike.
The Geekom A9 Max AI is one of the most talked-about mini PCs of the year. With a price tag of $999 and powered by AMD’s latest Ryzen AI 9 HX 370 processor, this compact machine promises to combine raw computing muscle with next-generation AI acceleration. But does it live up to the hype, and should you actually buy one in 2025? After an extensive week of testing, here’s everything you need to know.
The Mini PC Market in 2025: Why Size Matters
In 2025, mini PCs are no longer just backup systems or media servers. Thanks to modern CPUs, GPUs, and NPUs (Neural Processing Units), they now compete head-to-head with bulky towers in areas like:
- Productivity (office suites, virtual machines, coding environments)
- Creative work (photo editing, video rendering, 3D design)
- AI workloads (local LLMs, Copilot+ integration, data processing)
- Gaming (esports titles, AAA games with mid-range graphics settings)
The global mini PC market has grown at a steady pace, driven by remote work, compact urban living spaces, and rising demand for energy-efficient computers. The Geekom A9 Max AI enters this space with an aggressive pitch: small size, big power, and AI-focused hardware.
Geekom A9 Max AI: Specs at a Glance
Here’s a breakdown of the A9 Max AI’s hardware configuration:
| Component | Specification |
|---|---|
| Processor | AMD Ryzen AI 9 HX 370 (up to 54W TDP) with XDNA 2 AI engine |
| AI Acceleration | Up to 80 TOPS total, 50 TOPS via dedicated NPU |
| Graphics | AMD Radeon 890M (RDNA 3.5, 16 compute units) |
| Memory | Up to 128GB DDR5-5600MHz dual-channel RAM |
| Storage | 2x M.2 PCIe 4.0 NVMe SSD slots (2280 + 2230) |
| Connectivity | Wi-Fi 7, Bluetooth 5.4, dual 2.5GbE RJ45 ports |
| Ports | Multiple USB 4.0 Type-C, HDMI 2.1, full-size SD reader |
| OS | Windows 11 24H2 |
| Dimensions | 135 × 132 × 46.9 mm |
On paper, this makes the A9 Max AI one of the most powerful mini PCs under $1,000, especially for users interested in AI-driven workflows.
Setup and First Impressions
Unboxing the A9 Max AI, you’re struck by how little desk space it occupies. Roughly the size of a stacked paperback novel, it feels solid with a premium build. Connectivity is plentiful — from USB4 ports to dual HDMI outputs for multi-monitor setups.
However, setup on Windows 11 is predictably slow. Between updates, Microsoft account sign-ins, and system configurations, it took nearly an hour before the machine was fully usable. This highlights one of the ongoing frustrations with Windows-powered mini PCs: the OS itself is often the bottleneck.
Windows vs. Linux: The Ongoing Battle
One of the most pressing questions for power users is whether mini PCs like the A9 Max AI run better on Linux. The short answer: yes.
During testing, Linux distributions such as Ubuntu and Pop!_OS consistently outperformed Windows in terms of efficiency and responsiveness. For example:
- Running the same large language model (DeepSeek-r1:8b) locally produced faster, cleaner, and more accurate results on Linux compared to Windows.
- Resource allocation was more efficient, with Linux using significantly less RAM and CPU overhead.
- Virtual machines ran smoother, with Ubuntu VMs outperforming Windows guests even when given fewer cores and memory.
If you’re comfortable with Linux or willing to learn, the A9 Max AI feels like a completely different machine under open-source software.
AI Performance: Local Models on a Mini PC
The A9 Max AI markets itself as an AI powerhouse, thanks to AMD’s Ryzen AI 9 HX 370 chip with its XDNA 2 NPU. The specs claim 80 TOPS of total AI performance, making it suitable for:
- Running Microsoft Copilot+ features locally
- Deploying large language models (LLMs) such as LLaMA and DeepSeek
- Handling content generation, transcription, and real-time translation
- Enabling AI-enhanced creative tools like Adobe Firefly
In practice, here’s how it performed:
- DeepSeek-r1:8b model (9GB): Completed queries faster on the A9 Max AI than on some larger desktops, though grammar and accuracy were weaker under Windows compared to Linux.
- GPT-OSS:120b model: Managed to run, albeit very slowly (~5 minutes per query). Impressively, the system didn’t crash despite lacking a discrete NVIDIA GPU.
The takeaway? For local AI experimentation, the A9 Max AI punches above its weight. But don’t expect it to replace a full AI workstation with high-end GPUs.
Everyday Performance: Apps, Virtualization, and Multitasking
Outside of AI testing, the A9 Max AI handled everyday workloads with ease:
- Productivity apps like LibreOffice and GIMP launched quickly.
- VirtualBox virtual machines ran smoothly, even with heavy allocations.
- Web browsing, media playback, and multitasking were seamless.
One caveat: the fan noise. Under heavy loads, the cooling system ramped up aggressively, making the machine audible even from another room. This could be a dealbreaker for users who need silence.
Benchmark Comparisons
Here’s how the A9 Max AI stacks up against popular alternatives:
| Mini PC | CPU/GPU | Max RAM | AI Performance | Starting Price |
|---|---|---|---|---|
| Geekom A9 Max AI | AMD Ryzen AI 9 HX 370 / Radeon 890M | 128GB | 80 TOPS | $999 |
| Intel NUC 13 Pro | Intel Core i7-1360P / Iris Xe | 64GB | ~30 TOPS | $899 |
| Beelink GTR7 | AMD Ryzen 9 7940HS / Radeon 780M | 64GB | ~40 TOPS | $849 |
| Apple Mac Mini (M4) | Apple M4 SoC / Integrated GPU | 64GB (unified) | ~45 TOPS | $1,099 |
| Minisforum UM890 Pro | AMD Ryzen 9 8945HS / Radeon 780M | 96GB | ~45 TOPS | $899 |
The Geekom A9 Max AI leads the pack in AI-specific acceleration, making it attractive to developers and researchers who want a compact system for experimenting with local models.
Pros and Cons
Pros
- Powerful AMD Ryzen AI 9 HX 370 processor
- Industry-leading AI acceleration in a mini PC
- Support for up to 128GB RAM
- Excellent connectivity (Wi-Fi 7, dual 2.5GbE, USB4, HDMI 2.1)
- Handles virtualization and multitasking with ease
Cons
- Windows 11 hampers efficiency compared to Linux
- Loud fans under load
- Lacks discrete GPU for high-end gaming or pro-level AI training
- $999 price point may feel high for casual users
Who Should Buy the Geekom A9 Max AI?
This mini PC isn’t for everyone. Here’s who it best serves:
- AI enthusiasts and developers testing local LLMs or Copilot+ workflows
- Tech professionals who need a portable yet powerful workstation
- Students and digital nomads requiring performance in a compact form factor
- Small businesses looking for energy-efficient desktop replacements
Who should skip it? Gamers who demand top-tier graphics, silence-focused users who need whisper-quiet PCs, and casual users who don’t need AI acceleration will likely find better value elsewhere.
The Verdict: A Serious Contender in 2025
At $999, the Geekom A9 Max AI isn’t cheap, but it justifies the price for the right audience. With cutting-edge AI acceleration, solid everyday performance, and generous expandability, it stands out as one of the best mini PCs of 2025.
Yes, it struggles with Windows inefficiencies and noisy fans, but under Linux, it transforms into a highly capable machine that rivals larger desktops in performance.
If you want a compact system ready for AI experimentation, virtualization, and productivity, the A9 Max AI deserves a place on your shortlist. It’s not perfect, but it’s a glimpse into the future of small form factor computing.
