Snapdragon Makes Its Way to Desktop PCs — ASUS Unveils a One-of-a-Kind New Device
Arkadiy Andrienko
At Computex 2026, ASUS fired back at everyone waiting for a compact desktop that’s not x86 but ARM-based. Together with Qualcomm, they unveiled the Ascent QN10 mini-PC. This marks the second time the Snapdragon X2 Elite platform has jumped from laptops into the desktop space, carving out a niche for ultra-compact systems — the first time was also an ASUS product, but an all-in-one.
Under the hood, you’ve got an 18-core Qualcomm Oryon chip with Adreno graphics, but the real focus isn’t on core count — it’s on the NPU, which delivers 80 TOPS. That’s what lets ASUS position this thing as a tool for working with AI agents without constantly hitting the cloud. On the demo floor, they highlighted local use cases like OpenClaw, Cursor, Claude Desktop, Hermes, OpenAI Codex, and OpenCode. The idea is to run heavyweight language models and agent orchestrators right on the device itself.
Qualcomm claims the Snapdragon X2 Elite’s power efficiency keeps the system cool and silent even during long sessions. As far as ports go, the Ascent QN10 is pretty well kitted out: three USB4 Type-C, three USB 3.2 Gen 2, and one USB 2.0. That means you can hook up multiple monitors (they claim support for up to four 4K displays at once) and peripherals without needing extra hubs. Snapdragon Guardian handles hardware-level security, which is a nice touch when you’re dealing with sensitive data on-device.
That said, ASUS didn’t box this thing in as just an “AI agent box.” They mention it’s also a solid fit for developers who care about compile times in VS Code and working with large datasets. Even gamers are promised solid multitasking performance — though let’s be real, it’s no replacement for a proper gaming rig with a discrete GPU.
The one real fly in the ointment? No release date or final pricing yet. ASUS kept the commercial details under wraps, which suggests this is still an early announcement — especially since the platform is just starting to spread its wings beyond laptops.
What do you think — will the ability to run AI agents locally become the deciding factor when choosing a mini-PC for home or office use, or will price and compatibility with good old x86 software still rule the day? Drop your thoughts in the comments.
