Those who own a Qnap NAS and want more AI computing power can purchase it in the form of a USB stick or M.2 PCIe module.
Qnap launches the QAI-M100 and QAI-U100. With these devices, the NAS specialist focuses the AI hype on a stick. The modules have a modest AI accelerator on board that can be connected via USB or an M.2 PCIe slot, respectively.
read also
Getting started with Qnap QTS 5.2: powerful yet accessible
The small devices don’t have a lot of AI horsepower. Qnap equips both models with an NPU featuring three TOPS, supported by 1 GB of memory. However, this should suffice to accelerate applications such as object and facial recognition on a Qnap NAS.
Plug Play
Users can directly insert the QAI-U100 or QAI-M100 into their NAS, either via a USB 3.2 Gen 1 Type-A port or through a built-in M.2 2280 PCI interface if available on the device. In both cases, the accelerator should work immediately if the correct software is running. This includes:
- NAS operating system: QTS 5.2.1.2930 build 20241025 or QuTS hero h5.2.1.2929 build 20241025 (or newer)
- Compatible software: QNAP AI Core v3.5.0 (or newer), Multimedia Console v2.7.0 (or newer), QuMagie v1.5.1 (or newer)
Qnap predicts an acceleration of AI object recognition by up to 36 percent and facial recognition by 22 percent. How important the speed of AI recognition is on a NAS performing background tasks will depend on the usage scenario. In any case, Qnap suspects there is demand for a modest AI upgrade of its storage servers.
The expansion cards should be immediately available, but they are not so easily found in local stores. Online, they are available within Europe for approximately 65 euros for the USB version and 55 euros for the PCIe version, each excluding VAT.