ASUS and NVIDIA Assist National Chung Hsing University in Building Advanced AI Infrastructure


High Performance Computing ASUS and NVIDIA Assist National Chung Hsing University in Building Advanced AI Infrastructure Print Page Caption: From left to right: Assistant Professor Chen Guan-Chen, Vice President for Research and Development Song Zhen-Ming, Department Chair Jian Rui-Yu, Associate Chair Wang Shi-Ming, Assistant Professor Lan Guo-Rui, Assistant Professor Wu Zhong-Han National Chung Hsing University (NCHU), with its strong research capacity and the latest AI infrastructure, has successfully connected cross-domain collaborations in smart manufacturing, smart healthcare, and smart agriculture. This not only implements high-performance computing in academic research and industrial applications, but also promotes NCHU as an important AI ecosystem hub in central Taiwan. With strong financial support from the Ministry of Education, NCHU commissioned the startup company Glows.ai to deploy the latest AI computing power and build a cloud management platform, deepening the application of AI in research fields such as smart manufacturing, smart healthcare, and smart agriculture. As a leading AI infrastructure provider in the industry, ASUS assisted NCHU and Glows.ai by supplying NVIDIA DGX™ and NVIDIA RTX™ hardware equipment and completing the construction of the server data center. NVIDIA DGX™ H200 is a relatively high-end model that provides powerful artificial intelligence foundations, while NVIDIA RTX™ A6000 is used for model inference after machine training, becoming a powerful computing resource required for the integration of teaching, research, and AI. Through ASUS’s diverse hardware product portfolio and software solutions, and leveraging its experience in addressing diverse requirements encountered in AI server construction, ASUS not only helps build and deploy AI ecosystem applications, but also provides one-stop platforms including ACC (ASUS Control Center), AIDC (ASUS Infrastructure Deployment Center), and BMC to meet construction, deployment, and operational needs, while also addressing data center lifecycle management requirements. This enables seamless integration and rapid scalability of AI systems across various application scenarios, strongly supporting NCHU and customers in different fields in achieving success in AI deployment. Further examining NVIDIA DGX™ H200, it belongs to the DGX platform and is equipped with eight H200 Tensor Core GPUs with a total of 1,128 GB of GPU memory. Each GPU features 18 NVIDIA NVLink™ connections, delivering 900 GB/s of bidirectional GPU-to-GPU bandwidth. It also integrates four NVIDIA NVSwitches, ten NVIDIA ConnectX-7 400 GB/s network interfaces, dual Intel Xeon Platinum 8480C processors, and 30 TB of NVMe SSDs, delivering outstanding AI computing performance. Caption: NVIDIA DGX™ H200 expands new frontiers of business innovation and optimization Song Zhen-Ming, Vice President for Research and Development at NCHU, pointed out that after the deployment of AI computing power was completed, NCHU became the first university in Taiwan to possess such powerful AI computing capabilities. It also supports interdisciplinary research and teaching across the Taiwan National University System (NUST), which consists of 15 universities. From daily learning and student projects to industry-academia collaborative research, AI can be extensively utilized to help industry solve problems. AI Enhancing Smart Manufacturing The NCHU Smart Manufacturing Integrated Talent Training Base, inaugurated in November 2023, is a demonstration factory incorporating the latest technologies. With funding support from the Ministry of Education and industry donations, the smart production line integrates nearly 70 advanced hardware and software systems, including gantry-type five-axis machines, turn-mill compound five-axis machines, three-axis machines, AR and VR virtual factory applications, coordinate measuring machines, X-ray residual stress analyzers, ultrasonic flaw detectors, collaborative robots, and robotic arms. In addition, the powerful AI computing capabilities driven by NVIDIA DGX™ H200 are integrated to connect the overall operation of the smart production line. In the past, mechanical manufacturing typically optimized production based on individual machines. Entering a new stage of smart manufacturing, planning is conducted from the perspective of the entire production line, where different machines and sensors communicate with each other, generating massive amounts of production data. Relying solely on a few production control personnel is insufficient to allocate resources in real time; instead, AI with powerful computing capabilities must be involved to perform real-time calculations and optimal planning, thereby improving production efficiency and yield rates. Chen Guan-Chen, Assistant Professor in the Department of Mechanical Engineering at NCHU, pointed out that when orders enter the smart manufacturing line, multiple products may be in production simultaneously. At this time, AI intervenes in the production management system to perform appropriate scheduling and confirm whether raw materials are sufficient. Collaborative robots in the production environment transport materials to processing machines, with AI assisting in planning transport routes. Combined with visual equipment on collaborative robots, the environment and objects are identified to assist operators. After processing is completed, collaborative robots deliver finished products for ultrasonic cleaning and laser engraving of production information, followed by packaging and shipment, completing the entire process. The demonstration factory uses three types of processing materials commonly found in central Taiwan as examples: shoe molds, turbine blades required by the aerospace industry, and fixtures used in semiconductor manufacturing. These meet the R&D needs of industries such as metal processing, aerospace and defense, and semiconductor packaging processes. AI Deepening Smart Healthcare and Smart Agriculture NCHU has invested in smart healthcare research and established collaborations with four major institutions: Taichung Veterans General Hospital, Changhua Christian Hospital, Show Chwan Healthcare System, and Tong’s General Hospital. Many collaborative projects incorporate deep learning and AI technologies. In cellular image recognition, traditional methods rely on medical laboratory technicians observing cellular morphology changes in dark rooms, which is time-consuming and harmful to eyesight. Lan Guo-Rui, Assistant Professor in the Department of Mechanical Engineering, shared that NCHU collaborated with Taichung Veterans General Hospital to use AI to quickly identify cellular features and provide reference information for physicians’ diagnosis. Wu Zhong-Han, Assistant Professor in the Department of Mechanical Engineering, also cited a collaboration case with the Department of Veterinary Medicine. When determining which type of cancer is present, complex procedures such as biopsies are typically required, and biomarkers are used to estimate the probability of cancer. While this is common in human medicine, it differs for pets. Veterinarians have far fewer animal cancer samples compared to humans, and pets often seek treatment only after disease onset, making it difficult to collect data from healthy conditions and identify differences. Therefore, when blood samples are taken from pets and biomarkers are analyzed, NCHU uses AI to assist veterinarians in analysis and detect cancer probability. In smart agriculture, NCHU developed an agriculture-specific generative AI, “Shennong TAIDE,” based on the National Science and Technology Council’s TAIDE model. It integrates over 18,000 documents from the Ministry of Agriculture and more than 8,000 master’s and doctoral theses archived in the NCHU library, enabling AI responses to be more specific and closer to real agricultural conditions. It provides professional agricultural knowledge to meet diverse needs in cultivation and management for farmers, academic research units, and agricultural enterprises. Microgrids Supporting NCHU’s AI Infrastructure At its main campus and Nantou campus, NCHU has implemented microgrids built by Taipower, utilizing sources such as solar energy and biomass energy to provide sufficient electricity for teaching, research, and campus operations. To deploy power-intensive equipment such as NVIDIA DGX™ H200 at NCHU’s Information and Computing Center, the server data center’s power system was re-planned and connected to NCHU’s microgrid. This enables islanded operation for a period of time. Even during emergency power outages, the equipment has sufficient buffer time for shutdown and power restoration, protecting data and users. Jian Rui-Yu, Chair of the Department of Mechanical Engineering, whose research spans catalyst reaction engineering, noted that catalyst development and reactor design require large volumes of experiments and data. Many experts have already summarized data into models to predict catalyst reaction capabilities and reactor scale. Integrating AI into such research also helps NCHU achieve energy saving and carbon reduction goals. Wang Shi-Ming, Associate Chair of the Department of Mechanical Engineering, added that enabling all industries to benefit is NCHU’s objective in actively applying AI. For a single company or institution, merely investing heavily in hardware infrastructure does not yield high marginal benefits. In contrast, NCHU possesses abundant research capacity, accumulating research data across manufacturing, healthcare, and agriculture. By collaborating with partner universities, industry associations, medical institutions, and government agencies, application results can be fed back to various industries. Through technical support from NVIDIA and ASUS, NCHU has built a forward-looking computing environment, providing replicable and effective AI solutions for various industries. Vice President for Research and Development Song Zhen-Ming emphasized that with the implementation of AI infrastructure, National Chung Hsing University has become an “AI Enabled University.” With deepened cross-university and cross-institution collaborations, NCHU will continue to drive more innovative outcomes with AI and expand its impact on industry and society. Caption: NCHU research team posing in front of the new AI infrastructure, declaring the university’s move toward an AI Enabled University. (Source: TechNews) Repost from TechNews

High Performance Computing

ASUS and NVIDIA Assist National Chung Hsing University in Building Advanced AI Infrastructure

Print Page

Caption: From left to right: Assistant Professor Chen Guan-Chen, Vice President for Research and Development Song Zhen-Ming, Department Chair Jian Rui-Yu, Associate Chair Wang Shi-Ming, Assistant Professor Lan Guo-Rui, Assistant Professor Wu Zhong-Han

National Chung Hsing University (NCHU), with its strong research capacity and the latest AI infrastructure, has successfully connected cross-domain collaborations in smart manufacturing, smart healthcare, and smart agriculture. This not only implements high-performance computing in academic research and industrial applications, but also promotes NCHU as an important AI ecosystem hub in central Taiwan.

With strong financial support from the Ministry of Education, NCHU commissioned the startup company Glows.ai to deploy the latest AI computing power and build a cloud management platform, deepening the application of AI in research fields such as smart manufacturing, smart healthcare, and smart agriculture.

As a leading AI infrastructure provider in the industry, ASUS assisted NCHU and Glows.ai by supplying NVIDIA DGX™ and NVIDIA RTX™ hardware equipment and completing the construction of the server data center. NVIDIA DGX™ H200 is a relatively high-end model that provides powerful artificial intelligence foundations, while NVIDIA RTX™ A6000 is used for model inference after machine training, becoming a powerful computing resource required for the integration of teaching, research, and AI. Through ASUS’s diverse hardware product portfolio and software solutions, and leveraging its experience in addressing diverse requirements encountered in AI server construction, ASUS not only helps build and deploy AI ecosystem applications, but also provides one-stop platforms including ACC (ASUS Control Center), AIDC (ASUS Infrastructure Deployment Center), and BMC to meet construction, deployment, and operational needs, while also addressing data center lifecycle management requirements. This enables seamless integration and rapid scalability of AI systems across various application scenarios, strongly supporting NCHU and customers in different fields in achieving success in AI deployment.

Further examining NVIDIA DGX™ H200, it belongs to the DGX platform and is equipped with eight H200 Tensor Core GPUs with a total of 1,128 GB of GPU memory. Each GPU features 18 NVIDIA NVLink™ connections, delivering 900 GB/s of bidirectional GPU-to-GPU bandwidth. It also integrates four NVIDIA NVSwitches, ten NVIDIA ConnectX-7 400 GB/s network interfaces, dual Intel Xeon Platinum 8480C processors, and 30 TB of NVMe SSDs, delivering outstanding AI computing performance.

Caption: NVIDIA DGX™ H200 expands new frontiers of business innovation and optimization

Song Zhen-Ming, Vice President for Research and Development at NCHU, pointed out that after the deployment of AI computing power was completed, NCHU became the first university in Taiwan to possess such powerful AI computing capabilities. It also supports interdisciplinary research and teaching across the Taiwan National University System (NUST), which consists of 15 universities. From daily learning and student projects to industry-academia collaborative research, AI can be extensively utilized to help industry solve problems.

AI Enhancing Smart Manufacturing

The NCHU Smart Manufacturing Integrated Talent Training Base, inaugurated in November 2023, is a demonstration factory incorporating the latest technologies. With funding support from the Ministry of Education and industry donations, the smart production line integrates nearly 70 advanced hardware and software systems, including gantry-type five-axis machines, turn-mill compound five-axis machines, three-axis machines, AR and VR virtual factory applications, coordinate measuring machines, X-ray residual stress analyzers, ultrasonic flaw detectors, collaborative robots, and robotic arms. In addition, the powerful AI computing capabilities driven by NVIDIA DGX™ H200 are integrated to connect the overall operation of the smart production line.

In the past, mechanical manufacturing typically optimized production based on individual machines. Entering a new stage of smart manufacturing, planning is conducted from the perspective of the entire production line, where different machines and sensors communicate with each other, generating massive amounts of production data. Relying solely on a few production control personnel is insufficient to allocate resources in real time; instead, AI with powerful computing capabilities must be involved to perform real-time calculations and optimal planning, thereby improving production efficiency and yield rates.

Chen Guan-Chen, Assistant Professor in the Department of Mechanical Engineering at NCHU, pointed out that when orders enter the smart manufacturing line, multiple products may be in production simultaneously. At this time, AI intervenes in the production management system to perform appropriate scheduling and confirm whether raw materials are sufficient. Collaborative robots in the production environment transport materials to processing machines, with AI assisting in planning transport routes. Combined with visual equipment on collaborative robots, the environment and objects are identified to assist operators. After processing is completed, collaborative robots deliver finished products for ultrasonic cleaning and laser engraving of production information, followed by packaging and shipment, completing the entire process.

The demonstration factory uses three types of processing materials commonly found in central Taiwan as examples: shoe molds, turbine blades required by the aerospace industry, and fixtures used in semiconductor manufacturing. These meet the R&D needs of industries such as metal processing, aerospace and defense, and semiconductor packaging processes.

AI Deepening Smart Healthcare and Smart Agriculture

NCHU has invested in smart healthcare research and established collaborations with four major institutions: Taichung Veterans General Hospital, Changhua Christian Hospital, Show Chwan Healthcare System, and Tong’s General Hospital. Many collaborative projects incorporate deep learning and AI technologies.

In cellular image recognition, traditional methods rely on medical laboratory technicians observing cellular morphology changes in dark rooms, which is time-consuming and harmful to eyesight. Lan Guo-Rui, Assistant Professor in the Department of Mechanical Engineering, shared that NCHU collaborated with Taichung Veterans General Hospital to use AI to quickly identify cellular features and provide reference information for physicians’ diagnosis.

Wu Zhong-Han, Assistant Professor in the Department of Mechanical Engineering, also cited a collaboration case with the Department of Veterinary Medicine. When determining which type of cancer is present, complex procedures such as biopsies are typically required, and biomarkers are used to estimate the probability of cancer. While this is common in human medicine, it differs for pets. Veterinarians have far fewer animal cancer samples compared to humans, and pets often seek treatment only after disease onset, making it difficult to collect data from healthy conditions and identify differences. Therefore, when blood samples are taken from pets and biomarkers are analyzed, NCHU uses AI to assist veterinarians in analysis and detect cancer probability.

In smart agriculture, NCHU developed an agriculture-specific generative AI, “Shennong TAIDE,” based on the National Science and Technology Council’s TAIDE model. It integrates over 18,000 documents from the Ministry of Agriculture and more than 8,000 master’s and doctoral theses archived in the NCHU library, enabling AI responses to be more specific and closer to real agricultural conditions. It provides professional agricultural knowledge to meet diverse needs in cultivation and management for farmers, academic research units, and agricultural enterprises.

Microgrids Supporting NCHU’s AI Infrastructure

At its main campus and Nantou campus, NCHU has implemented microgrids built by Taipower, utilizing sources such as solar energy and biomass energy to provide sufficient electricity for teaching, research, and campus operations.

To deploy power-intensive equipment such as NVIDIA DGX™ H200 at NCHU’s Information and Computing Center, the server data center’s power system was re-planned and connected to NCHU’s microgrid. This enables islanded operation for a period of time. Even during emergency power outages, the equipment has sufficient buffer time for shutdown and power restoration, protecting data and users.

Jian Rui-Yu, Chair of the Department of Mechanical Engineering, whose research spans catalyst reaction engineering, noted that catalyst development and reactor design require large volumes of experiments and data. Many experts have already summarized data into models to predict catalyst reaction capabilities and reactor scale. Integrating AI into such research also helps NCHU achieve energy saving and carbon reduction goals.

Wang Shi-Ming, Associate Chair of the Department of Mechanical Engineering, added that enabling all industries to benefit is NCHU’s objective in actively applying AI. For a single company or institution, merely investing heavily in hardware infrastructure does not yield high marginal benefits. In contrast, NCHU possesses abundant research capacity, accumulating research data across manufacturing, healthcare, and agriculture. By collaborating with partner universities, industry associations, medical institutions, and government agencies, application results can be fed back to various industries.

Through technical support from NVIDIA and ASUS, NCHU has built a forward-looking computing environment, providing replicable and effective AI solutions for various industries. Vice President for Research and Development Song Zhen-Ming emphasized that with the implementation of AI infrastructure, National Chung Hsing University has become an “AI Enabled University.” With deepened cross-university and cross-institution collaborations, NCHU will continue to drive more innovative outcomes with AI and expand its impact on industry and society.

Caption: NCHU research team posing in front of the new AI infrastructure, declaring the university’s move toward an AI Enabled University. (Source: TechNews)

Repost from TechNews