2024 Taiwan Edge AI Day Promotes Business Opportunities for AI 2.0 Applications in Various Industries Around the World

Contributing Editor: Ashley Lu/ Yu-Ting Lu

The rapid development of AI technologies in generative AI (GenAI) and large language models (LLM) have enabled unlimited digital transformation opportunities in various industries. Taipei Computer Association (TCA) stated that according to a report by the International Data Corporation (IDC), global Edge Computing spending will reach USD 378 billion in 2028 with a CAGR in the double digits between 2023 and 2028. As the focus of AI shifts from training to inference, Edge AI has become one of the most important trends.

The 2024 Taiwan Edge AI Day Discusses the Trends of the AI 2.0 Era in Edge Tech+ 2024

TCA stated that the AI 1.0 era focused on high-speed cloud computing and AI model development. As users begin focusing on cybersecurity and AI computing efficiency, utilizing AI hardware on the premise with dedicated AI inference models has been an alternative. This is the start of the AI 2.0 era where data control is returned to the end users and integrates Cloud, Premise, and Device.

The 2024 Taiwan Edge AI Day aimed to promote cross-industry cooperation between Taiwan and Japan’s Edge AI companies in a localized method and to welcome the era of Edge AI together with Taiwan and Japan’s ICT and semiconductor industries. The event was organized by the Asia Silicon Valley Development Agency (ASVDA), Asia Silicon Valley Major League IoT, IC Taiwan Grand Challenge, TCA, Taiwan AI Alliance, and other organizations. The TCA Tokyo Office organized the seminar during the Edge Tech+ 2024 in Yokohama, Japan on November 21.

The 2024 Taiwan Edge AI Day invited speakers including Albert Liu –Founder & CEO of Kneron, KS Pua - Founder & CEO of Phison Electronics, and Kenji Tsuda - Editor-in-chief of News & Chips who shared their insights on the latest trends in semiconductor and Edge AI. The event attracted close to 200 professionals from Japan’s industry, government, academia, and research institutes; including senior executives and R&D personnel from many heavyweight Japanese tech companies.

Sakura Yang: Taiwan Actively Promotes the Development of AI and Semiconductor Industries and Welcomes Taiwanese and Japanese Companies to Cooperate in AI Digital and Net-Zero Dual Transformation Business Opportunities

The opening speech of the event was delivered by Sakura Yang, Deputy Secretary General of TCA. She stated that as GenAI moves from the cloud to the Edge, users can benefit from a faster and more personalized AI experience; which also brings new changes to the next generation of AI technology. As the Japanese government is actively revitalizing their domestic semiconductor industry and fully promoting AI applications and the deep mutual trust between Taiwan and Japan, the seminar focused on Edge AI and semiconductors. The event hoped to create more cooperation opportunities for Taiwanese and Japanese companies and expand greater business opportunities for AI applications in various industries together.

Ms. Yang stated that the Taiwan government has been actively promoting the development of AI and semiconductor industries and implemented them through cross-ministerial cooperation projects. For example, the ASVDA was jointly promoted by the National Development Council (NDC), Ministry of Economic Affairs (MOEA), the National Science and Technology Council (NSTC), and other ministries in 2016.

This year marks the beginning of ASVDA 3.0 which aims to accelerate the digital and net-zero transformation of Taiwan’s industries. This also coincides with the needs of Japanese industries, so there will be more room for cooperation in the future.

To leverage the advantages of Taiwan’s semiconductor industry, the Taiwan government began to promote the Taiwan Chip-based Industrial Innovation Program (Taiwan CbI) in March 2024. It not only provided investment, but also mobilized multiple government ministries to cooperate with the industry.

One part of the Taiwan CbI is the IC Taiwan Grand Challenge which is sponsored by the NSTC. The competition aims to leverage Taiwan’s strengths as the silicon island to gather global top startups and VCs in Taiwan. Registration for the second batch of IC Taiwan Grand Challenge is now open and startups from Japan are welcome to join.

In addition, the Taiwan AI Alliance which is managed by TCA actively connects companies in various fields including AI computing, software & hardware, and cybersecurity to jointly promote the development of Taiwan’s AI technology and applications. As of time of writing, the alliance has a total of 45 member companies and openly welcomes cross-border cooperation between Japanese companies and AI focused organizations.

Albert Liu: Edge AI SoC with Built-in NPU can Accelerate the Launch of Edge AI Hardware and Reduce Energy Consumption for AI Computing

The Founder and CEO of Kneron, Dr. Albert Liu stated that as AI computing has become a necessary function for major companies which allowed them to introduce various AI applications. However, they now face issues of slow response time, high energy consumption, and enterprise vital data security. As semiconductor process packaging and testing technologies and AI models continue to advance, the industry has begun to promote Edge AI to be the mainstream of the new generation of AI computing. This means AI computing and data processing are performed on the Edge without sending it to the cloud, reducing the chance of cybersecurity issues.

Dr. Liu stated that Kneron has key technologies and R&D teams for Neural Processing Unit (NPU) accelerators, AI chips, AI models, etc. Edge AI System-on-Chip (SoCs) with different specifications have been developed with AI Chip + Edge Computing + AI algorithms as the core. With the KNEO shared development platform, companies can rapidly develop various Edge AI applications in the fields of smart security, AIoT, automotive, Edge Server, etc. This also enables AI chips to be stacked to accelerate AI computing efficiency. Many internationally renowned companies have cooperated with Kneron to launch various Edge AI computing products. In addition, Kneron has launched the KNEO-300 and KNEO-330 solutions for enterprise GPT applications; enabling rapid developments of internal enterprise GPT applications on the NPU based Edge AI server. The users will not need to be worried about data breaches and the power consumption is also much lower than traditional AI servers.

KS Pua: Phison aiDAPTIV+ Solution Enables Ordinary Workstations to Train LLM and Mitigate the Issue of High Cost in Building AI Servers

KS Pua, founder & CEO of Phison Electronics stated that Phison is a technological leader in solid state drive (SSD) technology which has been applied in many fields, including the launch of the world’s first 4TB SSD and launching a NASA-certified SSD which will soon land on the moon. Phison’s SSD was already installed on NASA’s Mars rover – Perseverance in 2020 and have successfully operated normally in the Martian environment. Phison is a supplier of NAND storage solutions, supplying customers in the server, EV, smartphone, industrial/ commercial robot, aviation, drone, gaming console, surveillance system, PC, and various other fields.

KS Pua shared that by introducing AI technology, companies can not only improve the operational efficiency of various functions, but can also be widely used in production and manufacturing environments. However, since the current costs for building AI servers are still relatively high, Phison has developed their own aiDAPTIV+ solution and taken advantage of the cost advantage of flash memory to greatly improve the feasibility of using ordinary workstations to train large language model (LLM). The Phison aiDAPTIV+ Pro Suite is a LLM training platform that includes back-end model management functions, LLM private model internal training, and provides a code-free operation interface. This also allows enterprises to start using enterprise-wide GenAI functions, including Conversational AI, search engines, AI code generation, voice services, data analysis, and other AI applications. aiDAPTIVGuru can assist businesses to read uploaded files and generate & process data for LLM training, which can save up to 90% of the time required to pre-process model training data.

Leveraging the growing popularity of AI PC in recent years, KS Pua suggested that the aiDAPTIV+ solution can be introduced to upgrade AIPCs to be an AITPC (AI Training PC). This can accelerate the popularization of AI computing environments in educational environments, creating a one AITPC per person scenario which is ideal for AI education and training.

Kenji Tsuda: AI Applications Have Entered All Walks of Life. Japanese Companies can Learn from NVIDIA to Reinvent Industrial AI

Kenji Tsuda, Editor-in-chief of News & Chips stated that AI applications are growing rapidly and according to the report of Precedence Research, the global AI market will reach USD 1.6 trillion in 2030 with a CAGR as high as 38.1% from 2021 to 2030 and the entire market will grow by nearly 20 times. This year’s Nobel Prize in Physics this year was awarded to the American scientist John J. Hopfield and the Canadian scientist Geoffrey E. Hinton who is also known as the Godfather of AI. They won the award for their foundational discoveries and inventions in neural networks which enable monumental progress in machine learning and once again proved that AI is an irreversible trend.

AI is now also applied in various fields including in autonomous driving, 5G/6G telecommunications communication optimization, bank fraud prevention, genetic analysis, drug development, rail inspection, EDA tools (by TSMC), photolithography machine parameters (by TSMC), and more. The introduction of Edge AI continues to grow rapidly, including in the form of AI smartphones, AI PCs, computer vision, and more. Semiconductor companies such as Intel, AMD, Qualcomm, STMicroelectronics, and Tenstorrent have also launched SoCs and other products for Edge AI computing.

In his latest book “What is NVIDIA?”, Kenji Tsuda not only introduces NVIDIA from the perspective of corporate culture and management style, but also their strengths which include their advantages in GPU architecture, CUDA Toolkit for parallel computing, and their support for complete AI program libraries (PyTorch and TensorFlow), and AI modules developed for specific applications (NIM and NeMo) which fully explains how NVIDIA was able to win the AI market advantage.

Kenji Tsuda believes that Japanese companies can learn how to leverage AI to reengineer their business from NVIDIA, including how to utilize AI in manufacturing, tourism, agriculture, and other fields. There are many plausible directions for Japan to gain advantages under the AI trend.

TCA will Host the AI & Semiconductor Forum in Tokyo, Japan on December 13 to Discuss In-Depth Semiconductor Cooperation Opportunities Between Taiwan and Japan

TCA stated that to explore the cooperation opportunities between Taiwan and Japan to Japanese companies, TCA will host the AI & Semiconductor Forum in Tokyo, Japan on December 13, 2024. The forum will invite speakers from the semiconductor industry. This forum is also one of the official forums of SEMICON Japan and everyone is welcome to join the forum and visit the event.

■2024 Taiwan Edge AI Day Event Details

Event time: November 21, 14:00 – 16:00 (UTC+9)
Event Venue: PACIFICO YOKOHAMA ANNEX 201 @Edge Tech+ 2024
Organizers: Asia Silicon Valley Development Agency (ASVDA), Asia Silicon Valley Major League IoT, ICTGC (IC Taiwan Grand Challenge), Taipei Computer Association (TCA), Taiwan AI Alliance
Supported by: Taiwan GloRa Alliance (GloRa), Taiwan IOT Technology and Industry Association (TwIoTA), Taiwan Advanced Automotive Technology Development Association (TADA)

■《About COMPUTEX TAIPEI》

COMPUTEX TAIPEI was founded and named by the then Chairman of Taipei Computer Association (TCA), Stan Shih. In 1985, TCA invited TAITRA to be a co-organizer of COMPUTEX TAIPEI. In 2016, the startup focused event, InnoVEX was introduced.

● COMPUTEX 2025 Overview ●

Event Dates: May 20 – 23, 2025
Event Location: Taipei Nangang Exhibition Center (TaiNEX) Halls 1 & 2
Main Theme: AI NEXT
COMPUTEX Official Website: https://www.computex.biz/
Event information and pre-registration website: https://show.computex.biz/
Facebook: https://www.facebook.com/ComputexTaipei
YouTube Channel: https://www.youtube.com/user/COMPUTEXTAIPEIshow/
LinkedIn: https://www.linkedin.com/company/computex-taipei/