ASUS Ascent QN10: 18 cores in a 0.7-litre mini PC for €2,195

ASUS’s 4 November press release covers the launch in Germany, Austria and Switzerland. It announces a recommended retail price including tax of €2,195 for the euro markets and CHF 2,195 for Switzerland. It confirms neither pricing nor availability in France: the presence of a product page on the French website is not enough to establish either.

The detailed configuration pairs the Snapdragon X2 Elite X2E-88-100 with 32 GB of LPDDR5x-8533 and a 512 GB NVMe SSD. It runs Windows 11 Pro and belongs to the Copilot+ PC category. The processor model is specified, but the press release does not provide a full retail SKU that could identify a potential French variant.

These details match the configuration in the regional press release and the current specification sheet. ASUS also publishes range-wide specifications that vary by version: it is better to check the seller’s exact model reference than to assemble a configuration from the best figures across several pages.

Enlarge

The three USB4 ports and HDMI output are useful for multi-display setups in particular. ASUS claims support for up to four 4K displays simultaneously. Two USB-C ports are accessible at the front, alongside two USB-A ports and the audio jack; the rear houses the third USB-C port, two USB-A ports, HDMI, Ethernet and power.

The compact design does impose a lasting limitation, however: ASUS states that the LPDDR5x memory is onboard, non-removable and non-upgradable, with a maximum capacity of 32 GB. The choice therefore needs to take future memory requirements into account, especially if you plan to load AI models locally.

The Hexagon NPU accelerates compatible AI workloads. Its theoretical throughput alone does not tell us how quickly a language model will respond, what size model will fit in memory or how many applications can work simultaneously. Those are questions of software, model and workload.

The QN10 should not be confused with the Ascent GX10 despite their similar names. To put NVIDIA-based machines for AI development in context alongside this Qualcomm-based Windows PC, our guide to the ASUS GX10 and ET900N G3 details their use cases and memory constraints.

Microsoft confirms that Windows 11 on Arm can run x86 and x64 applications through emulation, but recommends native Arm versions for performance. Drivers are a separate issue: kernel-mode drivers in particular must be available in Arm64 versions.

For a workstation, it is therefore important to check professional software, its extensions and specialist peripherals. An application that launches does not necessarily deliver the same performance as a native version, and a peripheral that relies on an incompatible driver may remain unusable.