The technology is helping plug holes created by unprecedented remote working and improving efficiency for manufacturers
The COMPUTEX Forum continued today with its second session, titled AI Empowerment. In keeping with the two-day forum’s overarching theme — The New Era of Intelligence — COMPUTEX’s organizer, the Taiwan External Trade Development Council (TAITRA), invited speakers from industry leaders to share their perspectives on the impact of artificial intelligence (AI) on the technology industry and its potential implications for security.
NXP Semiconductors president and CEO Kurt Sievers said that his future vision of technology is “about a world where technology is actually automated, is actually advancing and anticipating what we want.”
“It is a world where simple tasks become automatic, and complex tasks are being foreseen by the machines we are interacting with,” Sievers added.
The complex task cannot be accomplished by any company alone. Sievers showed the close partnership and collaboration with leading companies such as Advantech, Amazon Web Services, BMW, Garmin, Microsoft, Samsung, and TSMC.
Applications on the edge are changing how people consume, pay, move, interact and work, fueling massive growth in devices. The demand for silicon for edge computing has grown significantly, but wafer supply is constrained. However, the faster that the industry moves to smaller process nodes and newer technologies, the better the availability will be to meet the required capacity, Sievers predicted.
“Rene Haas, the President of Arm’s IP Products Group (IPG), expressed confidence that 100 percent of the world’s digitally-shared data will soon be securely-processed on Arm-based technology. He cited the Arm-partners’ more than 190 billion chip shipments to date and looked to the future with the Armv9 architecture, which will play a critical role in securing our digital future and increasing our trust in intelligent devices that will process our personal or professional data.
Haas also stated that when entering a new era of compute where AI and ML will cut across every computing environment, one-size-fits-all chips designed for general-purpose compute with big single-threaded processors are not enough. Those complex workloads will be enabled by specialized processing. Arm will continue to innovate new ways to enable its partners to differentiate and bring specialized processors to market faster, and no device will be left behind.”
Amid the unprecedented global COVID-19 pandemic, hundreds of millions of people began working from home — and continue to do so today. This has laid bare many cybersecurity shortfalls in corporations across the globe.
For example, “companies have to open up access” to their most sensitive corporate assets “from remote endpoints to ensure business continuity,” said Kev Hau, security evangelist at Check Point Software Technologies.
Hau added that this new remote workforce “has never had to develop” a high level of vigilance, as they previously depended on enterprise-grade security for their work. This has left them more susceptible to “phishing scams, malicious emails and data loss incidents.”
This increased risk is why firms like Check Point are adopting AI for predictive threat intelligence. Properly trained AI can perform the work of dozens of human analysts in a fraction of the time, all while maintaining a broader scope of analysis, predicting unknown threats and improving the accuracy and precision of responses.
Of course, the expansive capabilities of AI extend beyond security to other areas, such as manufacturing, where an intelligence-driven manufacturing will help factories adapt to “a paradigm shift for the manufacturing industry,” said Dr. Steven Chen of Delta Research Center, Delta Electronics, Inc.
“Today’s customers are more demanding and diverse. They keep manufacturers busy trying to meet the needs for personalization and immediacy.” The need for high-mix, low-volume production while maintaining utilization means that manufacturers must devise innovative approaches to stay ahead.
To that goal, Delta’s industrial solutions include machines that are designed to cope with changes by incorporating standardized, modular and intelligent units, lowering technical barriers for data collection & analysis, as well as AI adoption and upgrade.
Such uses of AI are growing more commonplace, with about 90 percent of enterprise applications forecast to have embedded AI by 2025. The “pervasive growth of AI” has been fueled by the tremendous amount of data available, which is growing on “a daily basis,” said Raj Hazra, senior vice president and general manager of Micron Technology’s compute and networking business unit.
Hazra cited some areas where AI will undoubtedly find new uses, such as conversational interfaces and the ability to deal with multiple sources of data to train and make predictions, thus enabling applications including smart manufacturing.
To deal with the massive increases in abilities needed to support such advanced applications of AI, the very nature of memory is changing. It is no longer just a resource directly connected to a CPU and storage. Looking forward, Hazra said that distributed computing would provide the right amount of resources needed for applications, right when they are needed.