Taipei, April 2, 2019 – QNAP® (QNAP® Systems, Inc.) will showcase the TS-2888X AI-Ready NAS for machine learning at the 2019 Artificial Intelligence Exhibition & Conference (AI EXPO 2019). The TS-2888X uses the exclusive QuAI Developer Suite and the Intel® OpenVINO™ Workflow Consolidation Tool (OWCT) to simplify AI application development processes, and helps organizations develop Computer Vision and deep learning solutions with cost effectiveness and high efficiency. QNAP will also unveil its smart home video companion robot, AfoBot, which integrates health, education and surveillance applications.

TS-2888X - All-in-one platform for AI solutions, storage, computing and development

The TS-2888X AI-Ready NAS features a high-performance Intel® Xeon® W processor with up to 18 cores/36 threads and up to 512 GB of DDR4 ECC memory. With a huge potential storage capacity that alleviates data storage problems that AI applications often face, the TS-2888X also supports up to 4 high-end graphics cards to significantly increase AI machine learning efficiency. Its hybrid storage architecture provides more flash storage space, allowing users to transfer pre-processed data to four U.2 SSDs for more efficient storage utilization. Through the training models generated by a TS-2888X, OWCT can be used for optimization and verification to facilitate deep learning inference in various industries.

QNAP NAS simplifies and verifies AI deep learning inference applications through the use of accelerator cards and OWCT acceleration

QNAP’s OWCT is a collaborative tool designed for deep learning applications and for integrating OpenVINO workflows. It allows trained AI models to be deployed to target environments for inference operations.
It comprises an OpenVINO's open source library that can help shorten development time and provides a model optimizer that can make inference up to 100 times faster. The most salient feature of the OWCT is its GUI that makes using the OpenVINO features a breeze. Users can also view rendered images and download/archive inference results.

A TVS-672XT NAS with a Mustang-F100-A10 PCIe FPGA accelerator card will be showcased at the Expo to demonstrate how the QNAP NAS may be used as an inference server. Inferential verification features that visitors can look forward to include Smart Coin Recognition, Traffic Volume Management and Facial Recognition.

"QNAP is set to position itself as an innovator in computing, networking and storage solutions from 2019 and is actively developing solutions for AI applications,” said Zhang Mingzhi, general manager of QNAP. “We have gained valuable experience from the practical applications of our AI solutions through our collaboration with renowned medical research institutions in Taiwan."

"AI applications integrate Big Data with cloud computing and powerful algorithms, and QNAP can integrate these key elements in both hardware platforms and open-source software. We look forward to the future and our appearance at the AI Expo in Japan", he added.

AfoBot video companion robot – A new solution that integrates health, education and surveillance

QNAP's first video companion robot AfoBot is powered by an Android™ 6-based operating system, and comes with 32 GB built-in storage to support voice and face recognition, as well as video calling. A wide range of features will be demonstrated at the Expo, including health care applications such as blood sugar measuring and recording, instructional videos on language learning, as well as safety surveillance. Not only is the AfoBot an intelligent assistant for daily life, it can also help general users or businesses to improve work efficiency.
 

AI EXPO Tokyo 2019 - Artificial Intelligence Exhibition & Conference
Exhibition date: April 3 to April 5, 2019
Venue: Tokyo Big Sight (Tokyo International Exhibition Center), 3-11-1 Ariake, Koto-ku, Tokyo, 135-0063, Japan
QNAP booth: No. 9-6