MSI has introduced a new line of AI-focused mini PCs at CES 2026, targeting users who require compact systems for local large language model (LLM) processing and similar AI workloads. The diverse range includes systems based on AMD's RDNA 3.5 architecture, NVIDIA's RTX 5070 GPU, and the most notable EdgeXpert model which utilizes NVIDIA's DGX Spark platform. The EdgeXpert features 20 Arm CPU cores paired with an AI Blackwell GPU and NVIDIA's NVLink C2C technology, providing five times the bandwidth of PCIe 5.0 for enhanced CPU-GPU data transfer.

This configuration delivers one petaFLOP of AI performance, supported by 128 GB of LPDDR5x unified system memory and a vapor chamber cooling system. These new offerings represent MSI's strategic move into the edge AI computing market, providing developers and professionals with powerful, compact solutions for on-device AI processing. The introduction of these systems demonstrates the growing demand for specialized hardware that can efficiently handle AI workloads locally, reducing dependency on cloud-based solutions while maintaining high performance in a small form factor.