ASUS just started shipping the Ascent QN10, the first mini PC built around Qualcomm’s Snapdragon X2 Elite platform and its 80 TOPS NPU.
The 80 TOPS NPU is designed for local AI tasks such as language models, image processing, and automation workflows.
Up to this point, mini PCs running Windows on Arm have been dev kits, such as the Windows Dev Kit 2023. Those types of devices allowed developers to optimize apps for Arm, but they were not aimed at general customers.
Latest Videos FromWindows Central
The ASUS Ascent QN10 is a very different device. It’s a consumer‑ready PC with an anodized aluminum chassis rather than the plastic developer boxes we’ve seen in the past. Our friends at Tom’s Guide said the Ascent QN10 “could be the Mac mini M4 alternative for Windows users.”
The Ascent QN10 has a 0.7-liter body, making it smaller than many of the best mini PCs. Despite the small body, it packs plenty of ports.
|
Processor |
Qualcomm Snapdragon X2 Elite (18-Core Oryon CPU) |
|
Graphics |
Qualcomm Adreno X2-90 |
|
NPU |
Qualcomm Hexagon (Up to 80 TOPS) |
|
Memory |
Up to 32GB LPDDR5x |
|
Storage |
512GB, 1TB or 2TB |
|
Ports |
2 x USB4, 2 x USB-A, 1 x 3.5mm audio jack (front), 1 x USB4, 1 x HDMI 2.1, 2 x USB-A, 1 x 2.5 Gigabit Ethernet (rear) |
|
Connectivity |
Wi-Fi 7, Bluetooth 6.0 |
|
Dimensions |
5.1 x 5.1 x 1.5 inches |
|
Weight |
1.6 pounds |
The specs of the Ascent QN10 received a small bump since our colleagues at Tom’s Guide went hands-on with the mini PC. The Ascent QN10 features Bluetooth 6.0 rather than the older Bluetooth 5.4.
The Ascent QN10 has a clever cooling solution that keeps the heat from its CPU and SSD separate. Fresh air is brought in on one side of the device but then split and sent out separately on the other side of the device.
Overall, the Ascent QN10 looks like a strong first showing for Snapdragon X2 Elite in a mini PC. The big question is if there’s a demand for an AI-focused mini PC running Windows on Arm.
The Ascent QN10 can access Qualcomm’s AI Hub, which provides over 175 pre‑optimized models and tools for testing and deploying AI workloads.