QNAP’s plug-and-play edge AI accelerator modules deliver up to 3 TOPS of processing power, utilizing either an M.2 2280 PCIe slot or a USB 3.2 Gen 1 port.
QNAP has introduced two new edge AI accelerators, the QAI-M100 and QAI-U100, designed to enhance the performance of AI applications such as image and video recognition on its NAS systems. Aimed at small to medium-sized businesses and individual users, these devices are designed to make advanced AI applications more accessible without requiring a complete overhaul of existing setups.
The QAI-U100 connects through a USB 3.2 Gen 1 Type-A interface and offers up to 3 TOPS of processing power. It is equipped with 1GB of onboard memory and features a heat dissipation design that helps maintain performance during continuous operation. Additionally, it has an LED status light that provides users with at-a-glance feedback on the device’s operational status.
QAI-M100 M.2 2280 PCIe AI Accelerator Specifications
| Specification | Detail |
| Computing Accelerator Cards | Edge AI Accelerator |
| Standard Warranty | 1 year |
| Controller | Cortex-A35 2C 1.6GHz RK1808 |
| Connector | M.2 2280 PCIe Gen 2 x1 (M+B key) |
| Form factor | M.2 2280 |
The QAI-M100 offers 3 TOPS of computing performance and utilizes an M.2 2280 PCIe Gen 2 x1 (B+M key) interface. Similar to the USB version, it is equipped with 1GB of memory and features integrated cooling to maintain stable performance. This model is designed for NAS units that feature an M.2 2280 PCIe slot or support QNAP’s QM2 expansion card.
QAI-U100 USB AI Accelerator Specifications
| Specification | Detail |
| Computing Accelerator Cards | Edge AI Accelerator |
| Standard Warranty | 1 year |
| Controller | Cortex-A35 2C 1.6GHz RK1808 |
| Connector | USB 3.2 Gen1 |
| Dimensions (L x W x H) | 92.5 x 29 x 11 mm |
Both accelerators are designed for ease of use, featuring plug-and-play installation and flexible configuration options. According to results from internal testing by QNAP Labs, adding one of these accelerators to a NAS system can increase facial recognition speeds by up to 22% and object recognition speeds by up to 36% when using QuMagie’s Smart Album feature. The devices also allow for multiple-unit installation, meaning users can install two accelerators on a single NAS to further scale up AI performance.
The QAI-M100 is equipped with a Cortex-A35 dual-core 1.6GHz RK1801 architecture and comes with a one-year standard warranty. It follows the M.2 2280 form factor, ensuring compatibility with nearly all systems. The QAI-U100 features the same architecture and warranty, with dimensions of 92.5 x 29 x 11 millimeters.
QNAP also states that the new accelerators are designed to deliver improved energy efficiency. By offloading AI tasks to dedicated hardware, the system can process complex computations with lower power consumption, thereby reducing overall operational costs.
Engage with StorageReview
Newsletter | YouTube | Podcast iTunes/Spotify | Instagram | Twitter | TikTok | RSS Feed
