One Petaflop Local AI Computing
MSI Unveils Compact EdgeMesa N AI+ Desktop
- MSI launched the 1.3-liter EdgeMesa N AI+ desktop powered by NVIDIA RTX Spark
- Delivers up to 1 petaflop FP4 AI performance with 128 GB unified memory
- Features vapor chamber cooling to sustain 140W TDP for 24/7 AI agents
Global AI computing and innovative technology leader MSI has unveiled the EdgeMesa N AI+ series, a compact AI desktop engineered to bring high-performance local AI computing into everyday workspaces. Powered by NVIDIA RTX Spark, this 1.3-liter mini-PC combines advanced AI acceleration with MSI's precision engineering, enabling developers, creators, and AI enthusiasts to build, run, and deploy AI applications locally without cloud dependency. Running on Windows 11 with a space-saving footprint, it serves as a practical platform to transition AI experimentation into real-world workflows while maintaining strict control over data and computing resources.
Designed for sustained AI workloads, the EdgeMesa N AI+ series houses a refined 1.3-liter aluminum chassis with an anti-fingerprint finish, weighing just 0.55 kg net (1.55 kg gross). Its MSI AI Pulse Cooling System integrates vapor chamber technology and dual turbo fans to effectively dissipate heat, sustaining 140W TDP performance and supporting reliable 24/7 operation of always-on AI agents without cluttering desk space. As AI evolves from simple command-driven tools into intelligent teammates capable of reasoning, planning, and taking action, the system empowers developers to build coding assistants, deploy Retrieval-Augmented Generation (RAG) applications using private knowledge bases, and automate repetitive tasks across document processing, vision AI, and voice interactions—bringing NVIDIA’s “From Tool to Teammate” vision closer to everyday users.
At its core, the NVIDIA RTX Spark N1X superchip integrates an up to 6,144-core NVIDIA Blackwell RTX GPU and an up to 20-core NVIDIA Grace CPU, delivering up to 1 petaflop of FP4 AI performance. Equipped with up to 128 GB of LPDDR5X unified memory, one NVMe M.2 storage slot, and native NVIDIA® CUDA® acceleration, the system allows developers to run large AI models and accelerate local inference seamlessly. Creators benefit from NVIDIA Studio acceleration for 3D rendering, video editing, and AI-assisted content generation, while gamers enjoy immersive experiences backed by ray tracing, DLSS, and Reflex technologies.
High-speed connectivity extends its capabilities beyond a single workstation. Featuring 10 Gb LAN, Wi-Fi 7, and Bluetooth 5.4, users can rapidly transfer massive datasets and sustain connected AI workflows. Its rear I/O includes four USB 20Gbps Type-C ports (three supporting DisplayPort Alt Mode 2.1 and one for power-delivery input), one HDMI 2.1 (FRL) port, and one RJ45 (10 Gb LAN) port, enabling flexible multitasking across up to four displays. Security and power are backed by TPM 2.0 and a 240W AC adapter. Furthermore, NVIDIA® Personal AI Router (PAIR) technology enables compatible devices to form a private AI inference cluster, intelligently routing independent inference requests across available local compute nodes to maximize resource efficiency.
//DBTech/NJP/SMP/MIST//





