ASUS has unveiled the Ascent QN10, a new ultra-compact mini PC built around Qualcomm’s Snapdragon X2 Elite platform and designed with local AI workloads in mind. The company is positioning it as the world’s first AI mini PC with an 80-TOPS NPU, bringing the same push toward on-device AI that has defined the latest Copilot+ laptops into a desktop that takes up less than 0.7 liters of space.

The Ascent QN10 combines an 18-core Qualcomm Oryon CPU with an integrated Qualcomm Adreno GPU and a dedicated Hexagon NPU capable of up to 80 TOPS of AI processing. That gives the tiny system enough dedicated AI hardware to handle compatible workloads locally rather than sending every task to the cloud. ASUS is targeting developers, creators, prosumers, and businesses that want to experiment with local and agentic AI while keeping the hardware footprint small.

A Snapdragon-powered desktop built around AI

The Snapdragon X2 Elite is the star of the show. ASUS says the 18-core Qualcomm Oryon CPU is designed to deliver high performance while maintaining the platform’s power efficiency, while the integrated Adreno GPU handles graphics and display workloads.

The Hexagon NPU is arguably the more important component for the QN10, however. With up to 80 TOPS of INT8 AI performance, it can accelerate compatible AI workloads directly on the machine. ASUS says the system supports multiple intelligent workloads simultaneously, including agentic AI applications.

The QN10 is also designed to work with Qualcomm’s AI Hub, giving developers access to optimized machine-learning models and tools for deploying applications locally. ASUS has specifically highlighted use cases involving local LLMs, AI-assisted development, inference, and agent orchestration.

ASUS also says the QN10 supports agentic AI frameworks including OpenClaw and Hermes Agent. The company envisions the machine handling automated workflows and other AI tasks locally, while hybrid setups can still combine on-device processing with cloud intelligence when necessary.

The system is a Microsoft Copilot+ PC as well, enabling supported Windows 11 AI features that rely on the NPU.

A surprisingly small desktop

The QN10 measures just 130 × 130 × 40 mm and weighs around 620 grams according to ASUS’s current product information. ASUS puts the chassis at less than 0.7 liters, making it 86% smaller than a conventional 5-liter desktop system.

Despite the tiny enclosure, ASUS has fitted the machine with a dedicated cooling system designed to keep the Snapdragon platform running smoothly. The design uses a high-efficiency CPU fan and vapor chamber, along with dedicated cooling for the SSD. ASUS says the QN10 is designed for quiet operation under lighter workloads while maintaining sustained performance when required.

Memory is available in configurations of up to 32GB of LPDDR5X, with ASUS listing speeds of up to 8533 MHz on its product page. The memory is onboard rather than user-upgradable, which is a trade-off for the QN10’s compact design.

Storage is more flexible. The QN10 uses a dual-M.2 SSD configuration, giving users two slots for storage expansion. ASUS’s specifications support PCIe-based NVMe storage, with configurations starting at 512GB and going higher depending on the configuration.

Plenty of ports for a tiny PC

ASUS has also avoided turning the QN10 into a dongle-dependent desktop.

The front of the machine includes two USB4 Type-C ports, alongside USB-A connectivity and a 3.5mm audio jack. The rear adds another USB4 Type-C port, two additional USB-A ports, HDMI 2.1, 2.5Gb Ethernet, and the DC power input. In total, the machine offers seven USB ports.

The USB4 ports support DisplayPort 1.4 and USB Power Delivery, with bandwidth of up to 40Gbps. The integrated graphics can also drive up to four 4K displays simultaneously through the HDMI output and three USB4 connections.

Wireless connectivity includes Wi-Fi 7, while ASUS lists Bluetooth 6.0 on its current product page. A 2.5GbE Ethernet connection is included for users who need faster wired networking.

That combination makes the QN10 suitable for more than a conventional desktop setup. ASUS is also pitching it for edge computing, digital signage, smart kiosks, enterprise analytics, and other applications where a small system can sit behind a display or inside a larger deployment.

Built with enterprise security in mind

ASUS isn’t treating the QN10 as simply a tiny consumer PC. The company has incorporated hardware-backed security features intended for business and enterprise deployments.

The system uses Qualcomm’s Secure Processing Unit and supports Microsoft Pluton, alongside fTPM 2.0 and Qualcomm’s Snapdragon Guardian technology. ASUS describes the approach as chip-to-cloud security designed to protect the device, data, and AI workloads.

That could become particularly relevant as more organizations begin running AI agents and private models locally. Keeping sensitive prompts, documents, and other data on the device can reduce the need to send every AI operation to a remote service, although the actual privacy benefits will ultimately depend on how individual applications are configured.

The Ascent QN10 was first shown by ASUS and Qualcomm at Microsoft Build 2026 in June, but ASUS is now formally announcing the system as part of its broader AI PC portfolio.

ASUS has not announced pricing for the Ascent QN10 in the announcement material, so the final cost and availability will determine where it sits among the growing crop of Snapdragon-powered mini PCs.

For now, though, the QN10 is an interesting demonstration of where Qualcomm’s PC platform is heading. Instead of putting Snapdragon X2 Elite exclusively into thin laptops, ASUS is using its efficiency and dedicated AI hardware to build a tiny desktop that can function as a local AI workstation, developer machine, or always-on edge computer without taking over the desk.