← Back to LLM Mini PC
GMKtec
GMKtec EVO-X2
Local LLM Inference, AI Development, Long Context Workloads, Professional Workstation
Overview
The GMKtec EVO-X2 is a flagship LLM mini PC built around AMD's Ryzen AI MAX+ 395 'Strix Halo' APU. With up to 128GB of fast LPDDR5X-8000 unified memory shared with the Radeon 8060S iGPU, it runs 70B-class models locally at usable speeds, making it a serious compact AI workstation for developers and researchers who want to keep large workloads off the cloud.
Key Highlights
- Flagship local LLM mini PC built on the AMD Ryzen AI MAX+ 395 'Strix Halo' APU with 16 Zen 5 cores
- Up to 128GB of fast LPDDR5X-8000 unified memory shared with the Radeon 8060S iGPU
- Runs 70B-class models locally in 4-bit quantization with usable speed
- Dual USB4 40Gbps plus OCuLink for scaling with an external GPU
- 2.5GbE Ethernet and Wi-Fi 7 networking for fast model and data transfer
Technical Specifications
| Brand | GMKtec |
|---|---|
| Model | GMKtec EVO-X2 |
| Processor (CPU) | AMD Ryzen AI MAX+ 395 'Strix Halo' (16 Cores, 32 Threads, up to 5.1 GHz, Zen 5) |
| Graphics (GPU) | AMD Radeon 8060S integrated (RDNA 3.5, unified with system memory) |
| Memory (RAM) | 96GB or 128GB LPDDR5X-8000 unified memory |
| Storage | Multiple M.2 NVMe slots (config dependent) |
| NPU / AI TOPS | ~50 TOPS NPU (XDNA, platform-wide AI compute higher) |
| Ports & I/O | Dual USB4 40Gbps OCuLink HDMI 2.1 DisplayPort USB-A headphone jack |
| Network | 2.5GbE Ethernet, Wi-Fi 7, Bluetooth |
| Operating System | Windows 11 |
| Dimensions | Compact flagship-class mini PC chassis (official dimensions to confirm) |
| Price | Approx $2,349 for the 96GB configuration; approx $3,299 for the 128GB configuration |