As technology advances apace, the modern demand for high-quality A/V experience has also continued to rise. From mobile phones, computers, and televisions, the demand has even extended to vehicle cabins. Major auto makers, looking to generate revenue, have also been quite keen to introduce new technologies to raise the added value of products. The smart cabin market is gradually taking shape alongside the development of the supply and demand sides. Mindtronic AI's eye-tracking AI system for vehicles, which can instantly track precise movements of the driver's eyes, is finding use in a variety of applications, and has been seen favorably by the industry since its release. More recently, Mindtronic has cooperated with Taiwan panel factories to enter the smart car market.The auto industry has recently ushered in a new industrial revolution, one in which autonomous driving, electric vehicles, and smartification are the three key trends. According to analysis by Mike Huang, CTO of Mindtronic AI, the development of self-driving (autonomous) cars has not been as fast as expected, due to the need for highly integrated peripheral systems and innovative business models. With the simple structure of electric vehicle chassis, however, the technical threshold is relatively low compared with fuel vehicles, resulting in a large number of investors and major competitive pressures to existing car manufacturers. In such a scenario, introducing forward-looking technology to provide a high-quality ride, and at the same time creating a smart cabin with a range of value, has become an essential strategy for auto makers.Huang also pointed out that the panel plays an important role in the smart cabin system. Most high-end vehicles on the current market come with customized touch panels. These panels incorporate the appearance and shape of next-gen cabin design and extend from the driver's seat to the front passenger seat. They also provide various functions in response to user needs, from driving assistance to high-resolution audio-visual applications. The smart cabin system requires a new type of human-machine interface (HMI) that includes not just touch technology, but also eye tracking as one of its main features.Eye tracking, a technology that has been under development for many years, has found wide use in on-board vehicle systems. Its main purpose is to determine the mental state of the driver and issue alerts when the driver is tired or not paying attention. Mindtronic AI has further integrated with AI by making AI the "copilot". In situations not favorable to driving, such as when the driver is not mentally capable of driving, the ADAS of the in-vehicle system will temporarily "take the wheel." In addition to assisted driving, Mindtronic AI's systems also combine eye-tracking technology with intention inference algorithms. Driver can make use of the eye-sight manipulated control panel and the head-up display (HUD) screen function on the windshield, and the system also has the function to actively infer the user's intention and learn their preference or target of interest (TOI). Combined with the recommendation engine to actively provide information, entertainment, and driving service functions, this function can also become a commercial mechanism that can generate operating income for car dealers or platform operators.Huang says that the current business model for auto makers is basically new car sales with follow-up maintenance. But with electric vehicles becoming a key policy in governments of various countries, the depreciation rate for electric vehicles off the lot is much lower than that for gas-driven vehicles, and the maintenance rate is thus also greatly reduced, so car manufacturers will need to find another source of revenue. Smart cabins may just be one source for new business opportunities. Huang further pointed out that there are two types of the software business service mechanism of the smart car cabin: in-cabin and out-cabin. The in-cabin service mechanism focuses on driving safety functions provided by auto makers, while out-cabin services are similar to smartphones. Consumers can download software products from app makers and use them on the smart cabin platform. From a functional point of view, the out-cabin service mechanism offers more functions and more business opportunities. Nevertheless, this mechanism is built on a complete ecosystem, like a smart phone, so it requires all those involved to work together for promotion.Eye tracking will be an important control technology for both in-cabin and out-cabin systems. Huang mentioned that current control technologies for the HMI include touch, voice, and eye movement, each performing its respective function in the smart cabin. When the driver is parked or other passengers who there are those who are not in charge of driving, the operating system function mainly uses touch mode, while voice control is mainly used for inputting commands to the navigation system. Eye tracking technology also has its own use here: it is used for selecting functions and reading information during driving. However, Huang also pointed out that for eye tracking control to be practical in the smart cabin system, the two key objectives, i.e., accuracy and immediacy, must be met to create a high-quality experience for users. The technology developed by Mindtronic AI is not affected by ambient lights inside the cabin. With its fast and precise tracking of eye movement, the technology has won wide acceptance in the market, not only winning the US CES Innovation Award for 4 consecutive years, but also being used in cooperation with panel manufacturers and large auto makers.While unique and practical, Mindtronic AI's eye-tracking AI system for vehicles is favored by the industry due to the innovation-friendly environment created by the government in recent years, argues Huang. He pointed out that the National Science and Technology Council's value creation program and the Taiwan Tech Arena (TTA) have been providing massive supports for startups, especially when taking a startup team to the United States to participate in CES, an excellent opportunity to gain exposure. Huang pointed out that startup teams are certainly aware of the importance of big overseas exhibitions, but that most teams are unable to participate in exhibitions due to limited resources. Also, small booth area limits the final exposure. So, the TTA brought together a large number of Taiwanese start-up teams and set up the TTA Pavilion to magnify the exposure for exhibitors. This expanded exposure has benefited Mindtronic AI in the past. That is why Huang suggests that startups can make good use of the support offered by government policy to expand product exposure while enhancing their own technological capabilities.Mindtronic AI product developmentPhoto: Company

Owing to consistent advancements in chip computing and AI algorithms, the scope of AI applications has grown exponentially across multiple sectors. The early years saw a primary focus on image recognition, followed by gradual yet impactful inroads into speech and text applications. A notable breakthrough was the introduction of ChatGPT, a major achievement in AI speech technology that gained widespread acclaim since 2022.At present, AI applications in the speech sector fall into three distinct categories. Foremost, chatbots that engage with consumers via text or speech, have become integral components in customer service frameworks. Second, voice-to-text applications translate recorded files or user speech into corresponding text, thereby reducing manual processing time considerably - an optimal solution for tasks like recording meeting minutes or for media professionals.DeepWave's Chief Operating Officer, Sung-tsan Yeh, cites noise reduction and the separation of human voices from background noise as the third noteworthy application of AI in the speech sector. Traditional methods to enhance speech quality required meticulous professional input, making it a time-intensive and expensive process only affordable for high-budget video projects. In contrast, individual content creators such as YouTubers or online instructors with limited budgets could only rely on pre-production noise-canceling equipment as they could not afford professional post-production services. Given the rising popularity of YouTube and online teaching, a significant number of companies are investing in AI technology to address the need for noise reduction and voice-background noise separation.DeepWave Revolutionizes Noise Reduction with AI Acoustic Recognition TechnologyFounded in September 2019 by Professor Roger Jang of National Taiwan University's Computer Science & Information Engineering Department, CTO Ken Yeh, and CEO Henry Lin, DeepWave is a startup created under the aegis of Taiwan's National Science and Technology Council's Germination Program. This pioneering company is committed to offering cost-effective AI acoustic recognition technology, aiming to streamline manpower requirements, elevate productivity, and augment industrial value across numerous sectors requiring speech recognition. The startup offers over 20 varieties of API services for acoustic AI recognition and processing, closely collaborating with sectors such as music post-production, education, corporate interviews, and anomaly detection in mechanical noises.A key product from DeepWave, the Noise Eraser, exemplifies intelligent AI noise reduction. Trained on approximately 10,000 human voice and noise files, it mimics professional sound effects engineers' noise-reduction techniques. This enables users to effortlessly achieve professional-grade noise reduction and enhance voice clarity in their videos. The user-friendly tool requires only three steps and an average minute to process five-minute footage, significantly alleviating disruptions from environmental noises for content creators and educators, thereby enhancing video quality.COO Sung-tsan Yeh notes that conventional noise-removal methods often render audio signals unauthentic, a common issue with most free tools. DeepWave's approach, built on foundational AI technology and backed by extensive experimentation and user testing, allows optimal application ratios for a broad spectrum of use-cases. Users can fine-tune the voice-to-noise ratio based on their preferences. With over two decades of presence and numerous victories in international speech and music analysis contests, DeepWave seeks to translate promising technologies into steppingstones to expedite businesses' digital transformation, thereby contributing to Taiwan's digital revolution.DeepWave's Noise Eraser: A Global Success Story and a Glimpse into the FutureDeepWave operates its flagship product, Noise Eraser, under a freemium model where users can process up to five-minute videos for free, while unlimited access is provided to subscribers of the paid version. The product's superior noise reduction capabilities, coupled with affordable subscription pricing, have attracted a significant international user base, contributing to 70% of the company's revenue. Testament to its robust technological prowess, DeepWave has received numerous accolades, including the "2021 GO NEXT DIGITAL" award under the Bureau of Industry's "Demo: Pitch Talent Cultivation Program", "Purpose Demo Show" by Startup Taipei, Chang Gung University and United Innovation, and the AIGO AI Talent Training Program award by the Administration for Digital Industries, Ministry of Digital Affairs.CEO Henry Lin acknowledges the support DeepWave has received from various legal entities and government agencies during its growth phase. Apart from the National Science and Technology Council's Taiwan Germination Program, the company has participated in multiple events facilitated by the TTA. Lin notes that these engagements have not only amplified brand and product exposure but also provided opportunities to connect with overseas investors and clients, bolstering the company's long-term growth prospects.In light of Noise Eraser's market success, DeepWave has ambitious plans for 2023, including the launch of new services like meeting minutes management. Their meeting minutes service, capable of discerning different human voices and organizing all spoken content during meetings, has emerged as a promising tool. Several government agencies currently trialing this service have lauded its efficacy and potential.Henry Lin, CEO of DeepWavePhoto: CompanySung-Tsan Yeh, Chief Operating Officer of DeepWavePhoto: CompanyNoise Eraser AI Noise-reduction Service Rolled Out by DeepWavePhoto: Company

STMicroelectronics has introduced the ST-ONEHP integrated digital controller, the world's first IC certified by USB-IF according to the USB Power Delivery Extended Power Range (USB PD 3.1 EPR) specification.As the third controller in the ST-ONE family, the ST-ONEHP has 28V output capability to simplify building chargers and power adapters rated up to 140W. Sharing the ST-ONE architecture, the converter is a non-complementary active-clamp-flyback zero-voltage-switching (ZVS) topology that ensures superior efficiency at high power output and switching frequency. Essential features including secondary regulation and the USB PD communication interface are integrated, saving bill-of-materials costs as well as PCB size and complexity. Built-in synchronous rectification helps maximize efficiency and reinforced galvanic isolation permits a solution that is both compact and safe.Inside the ST-ONEHP, an Arm Cortex-M0+ microcontroller maintains overall control. The MCU is supplied pre-loaded with certified USB PD firmware, which eases end-product approval to qualify for the USB logo. With 64KByte of flash storage, the MCU provides room for customizing the power-conversion firmware.The ST-ONEHP is well suited to use with ST's MasterGaN single chip, integrating the Company's third-generation gallium nitride (GaN) power transistors and optimized gate drivers. The SiP ease the adoption of GaN technology in energy-conscious applications to realize benefits including superior thermal performance and switching efficiency compared to conventional silicon transistors.Complementing the properties of the ST-ONE converter, MasterGaN helps maximize performance and allows high switching frequencies that permit smaller passive components. By combining the ST-ONEHP controller and a 9mm x 9mm MasterGaN1 half-bridge device, ST has produced the EVLONE140W reference design smaller than a standard 65W laptop charger.Its volume of 90cm^3 equates to industry-leading power density of 25W/in^3, while the peak efficiency reaches beyond 94%. From an environmental perspective, the charger can use just a quarter of the plastics and is 2% more efficient than average chargers. Using these products in all chargers produced worldwide would save 3.5 million tons of CO2 emissions. Samples of the ST-ONEHP are already available for selected customers in a SSOP36 leaded package.For further information please go to www.st.com/st-onehp.STMicroelectronics grows ST-ONE controller family for USB Power Delivery applications up to 140W

STMicroelectronics has released a toolchain and accompanying software package for programming the intelligent sensor processing unit (ISPU) embedded in the latest-generation intelligent MEMS IMUs, ISM330IS and LSM6DSO16IS.The toolchain and software help employ the ISPU to handle motion-related workloads such as activity recognition and anomaly detection directly in the sensor. This permits reducing system power and latency, offloading the local microcontroller, and uniquely specializing the behavior of the sensor to the application.Using the ISPU toolchain, developers can program the sensor's intelligent processing unit using the familiar and widespread C programming language. They can choose to work from a command line interface (CLI) or an Eclipse-based environment like STM32CubeIDE, and to use a graphical user interface (GUI) such as AlgoBuilder and Unicleo.The X-CUBE-ISPU software package contains templates and example projects as well as ready-to-use libraries that help developers quickly understand how to use and program the sensors' ISPU and can be used as a starting point to implement custom algorithms. Pre-built files are also available, which lets users load the X-CUBE-ISPU examples directly into the sensor using one of the GUIs, with no coding required. In addition, a GitHub repository is available providing more examples, tutorials, and other development resources.Using these resources helps make short work of developing applications such as personal electronics including wearable devices for activity recognition and health monitoring, as well as industrial devices such as asset trackers, equipment-condition monitors, robots, and machine controllers.ST's ISM330IS and LSM6DSO16IS inertial modules contain an always-on 3D accelerometer and 3D gyroscope with the embedded ISPU. They feature low power consumption, drawing as little as 0.46mA in low-power mode, and low noise at 70μg/√Hz in high-performance mode. Sensor hub functionality allows them to collect data from up to four additional external sensors. An embedded temperature sensor is also included, and each device is housed in a compact 2.5mm x 3mm x 0.83mm plastic land grid array (LGA) package.SPM Instrument (https://www.spminstrument.com/), a Strängnäs, Sweden, based innovator in condition monitoring and process optimization, relied on ST's ISPU toolchain and the X-CUBE-ISPU to tailor, in a short time, the ISPU behavior in their new products for vibration-severity analysis that contain the ISM330IS sensor.SPM's sensing solution is ideal for remote condition monitoring of standard production equipment such as pumps and fans, as well as inaccessible machines and equipment placed in hostile or risky environments. The ISM330IS enabled designers to meet a tight power budget, also overcoming the limited processing capabilities available in the local microcontroller.The ISPU-toolchain, X-CUBE-ISPU and GitHub model zoo are ready now to download, free of charge, at st.com. The ISM330IS and LSM6DSO16IS intelligent MEMS sensors are in production now and available.Please visit www.st.com/ispu for more information.New toolchain and software package from STMicroelectronics ease development of edge processing with intelligent inertial sensors.

During the peak of the pandemic, businesses embraced remote work, and schools adopted distance learning to ensure uninterrupted productivity and education. As we transition into the post-pandemic era, work and education are undergoing a transformation, leading businesses and schools to explore new operational models. "Hybrid work" and "hybrid education" are emerging as the new normal. In this context, enterprises and educational institutions require a flexible and highly customizable collaborative sharing tool that enables users to achieve their work and educational goals through real-time meetings and audiovisual communication, regardless of their location.Addressing this trend, Cypress Technology (CYP) has launched the CPLUS-V32SDM, a 4K ultra-slim modular matrix switcher. This innovative device allows for seamless integration of various audiovisual and environmental devices, including laptops, personal computers, tablets, smartphones, displays, projectors, video cameras, audio systems, keyboards, and mice. With the CPLUS-V32SDM, users can easily manage devices, share information, and engage in real-time collaboration, enabling both local and remote participants to focus on their conferences and classes.Based on the design concept of "One for all, all in one," the CPLUS-V32SDM is our first ultra-compact and highly versatile matrix switch on the market. It features a modular card slot that expands the unit's functionality through the use of our Signal Diversified Modules (SDMs). Users can customize the unit's configuration by easily swapping SDMs, catering to the unique requirements of conferences, education, and entertainment. Furthermore, the CPLUS-V32SDM supports USB 3.0 transmission and includes one USB-B and two USB-A ports for seamless device extension.Equipped with HDMI and the now mainstream USB-C input interfaces, the CPLUS-V32SDM, when paired with CYP's Signal Diversified Modules (SDMs), also offers support for wireless sharing, delivering a seamless presentation and collaboration experience. Various additional audiovisual signal types are also supported such as HDBaseT, SDI, and SDVoE, accommodating up to three different signal sources. The output interface provides two 4K@60Hz HDMI 2.0 outputs. Users can leverage this modular matrix switcher to centrally manage devices and swiftly switch between input and output signal sources, enhancing conferences and education, unlocking infinite possibilities.CYP is actively developing vertical market sectors, including OR medical imaging integration, hybrid conference rooms, smart classrooms, distance learning, esports, and command centers. The integration of audiovisual and IT has emerged as a prominent trend, enabling intelligent environments through the fusion of professional audiovisual and IT technologies. By seamlessly integrating various audiovisual and environmental devices, CYP empowers these fields with intelligence, driving the advancement of smart solutions.CYP Introduces a 4K Ultra-Slim Modular Matrix Switcher, Unlocking Infinite Possibilities for Conferences and Education

iCatch Technology and Prophesee have jointly announced the launch of the iEVCam, powered by iCatch Technology's V57 AI Imaging SoC and Prophesee's Event-Based Metavision® Technology. The iEVCam is a smart camera reference system designed for developers to deploy innovative event-based edge AI solutions across various industries. With the accompanying edge AI SDK, it significantly simplifies the integration of event-based algorithms onto the camera device, accelerating time-to-market of event-based products. This makes the iEVCam an ideal platform for scaling event-based technology solutions in the rapidly growing edge AI market. Prophesee will be showing the live demonstration of the iEVCam at Vision China 2023, from July 11 to July 13.iEVCam Developer Kit. Photo: iCatch Technology The iEVCam utilizes the IMX646 Event-based Vision Sensor (EVS) realized in collaboration between Sony Semiconductor Solutions and Prophesee. It offers high privacy data, high dynamic range, extreme low light performance, and low data rate. The AI Imaging processor used inside is iCatch Technology's V57 SoC, the world's first ASIC that can directly support EVS with edge AI. V57 AI Imaging SoC features event-based processing accelerators such as event decoder and event optical flow engine, along with computer vision Digital Signal Processor (DSP), Neural Processing Unit (NPU), and AI SDK for edge AI processing. These features enable developers to migrate their event-based algorithms from PC to camera device, allowing real-time processing and decision-making at the edge. The iEVCam is a perfect platform for product prototyping to reduce time-to-market for applications such as Driver Monitoring System (DMS), elderly care, home security camera, traffic monitoring, and much more.Elderly Care Scenario: RGB (Left) and Event (Right) Figures. Photo: iCatch Technology"We are thrilled to be working with Prophesee, the pioneer and market leader in event-based vision, to launch the iEVCam. Our partnership will expedite the adoption of this highly efficient vision sensing approach in a range of industries, especially in those where low power consumption is critical," said Weber Hsu, CEO of iCatch Technology. Luca Verre, CEO of Prophesee, noted, "As an established leader in key sectors such as automotive, security, and consumer/IoT, iCatch Technology has a proven and innovative vision solution that is well suited to the needs of many Edge AI IoT applications. With iEVCam, product developers now have a very useful way to evaluate how the sensor performs for multiple use cases and market-specific machine vision systems, which will advance the penetration of event sensor technology into new markets."The iEVCam is now available for purchase from iCatch Technology. For more information about the iEVCam, please visit this link. If you have any inquiries, please contact https://www.icatchtek.com/Requests .

In 2020, the escalation of the pandemic severely impacted the global retail and food service industries. In order to survive, most businesses have adopted digital technology to enhance consumer experience and attempted new business models to address labor shortage issues. To date, although the pandemic has officially ended and in-store shopping has gradually resumed, the wave of smartization and self-service in retail equipment, along with the rising AI boom, remains a prominent trend in the industry's future development. DataVan (stock code: 3521), which has worked to take root in the point-of-sale (POS) industry for many years, markets its products worldwide under proprietary brands. At COMPUTEX this year, which was held for the first time after the pandemic, DataVan showcased a variety of smart retail, catering, and service equipment integrated with AI technology under the theme of "Smart Airport." The exhibition managed to attract many domestic and overseas buyers.Big Business Potential for GPT-integrated KiosksThe first thing one would see upon entering the "Smart Airport" exhibition area was DataVan's newly-launched slim kiosk "Genius," which incorporates the DataVanGPT AI intelligent recommendation system. The kiosk will assume the role of an information hotspot for domestic and international travelers, offering smart recommendations of travel information through its text and image interface and built-in generative AI voice. Travelers can input their needs by text, or allow DataVanGPT to identify their gender and age with the camera on the kiosk to provide the best travel recommendations based on the results. For the management, DataVanGPT can also provide statistics and big data on passenger inquiries for subsequent analysis and application. In terms of commercial application value, according to DataVan, the kiosks can also be used by venue operators in collaboration with travel agencies, museums, handicraft stores, restaurants, as well as in other tourist attraction spots. Apart from providing information, they can serve as a one-stop service operator to enhance the travel experience, because coupons can be printed and scanned using the printer and 2D barcode scanner on the kiosks.AI Helps Retailers in Precision MarketingThe on-site airport scenario also included a duty-free shop, which used a duty-free store rack display to demonstrate the "AI Vision Intelligent Visual Service System" as the core solution of smart retail. AI Vision mainly consists of the eye-tracking analysis technology and AI Vision Snapshot, the real-time rack management system, which can help retailers determine the age and gender of customers standing in front of the racks, and record the position and duration of their visual focus. Analysis of the data will help determine hot zones in a store for more effective product placement, or make personalized recommendations during special holidays and promotional events, thereby realizing precision marketing and optimizing sales per unit area. In addition, the AI Vision Snapshot system will update the quantity of products on the racks in real time and notify store personnel to restock immediately after customers remove items from the racks. It can also be integrated with the inventory management system and the AI Robot Arm (an intelligent robot arm developed by DataVan that can be installed on racks to attract customers). Moreover, it can be applied in unmanned stores to enhance the convenience of self-checkout, thereby saving labor costs.New Multi-Functional POS, Highly Efficient and Light WeightAnother DataVan's featured product is a new POS device with the latest platform. It offers not only high performance to duty-free shops, but also has extremely flexible expansion features, providing consumers with a thoughtful design, such as a second display to confirm purchase items and amount. Meanwhile, DataVan has also developed a conceptual model with a multi-functional slim tablet design. It can be placed vertically as a mini-kiosk or horizontally as a regular POS. If officially commercialized in the future, the model could be used in the food service industry, serving as a digital signage placed in a commercial space. Alternatively, it could be combined with a Box PC for medical use or automated production lines. The operation status of POS and kiosk terminal devices can be remotely monitored and their fault alerted through the DataVan Central management platform to optimize efficiency. In addition to comprehensive POS product lines comprising desktops and mobile devices, DataVan also has its own production and excellent research and development capabilities, which allow it to provide customers with more customized designs and OEM services options.Smart Ordering System Opens Up New Opportunities in Food Service IndustryIn the meantime, DataVan also simulated an airport lounge, primarily to demonstrate its AI Waiter, an intelligent ordering service system. The system has an intuitive interface, and offers voice ordering function to help the elderly or people with mobility issues to verbally place orders and check out. Those still unaccustomed to the voice ordering system can fill out a paper order form, which will then be scanned by DataVan's MenuSync menu recognition device. The front-end intelligent ordering system will also incorporate the KDS (Kitchen Display System), allowing kitchen staff to receive and prepare the orders as soon as possible. This can reduce customer waiting time and increase the restaurant's table turnover rate.In the lounge, DataVan's smart beer dispenser system was also on display. Through the touch screen, it can dispense beer automatically and adjust the amount of beer dispensed according to the volumes of different glasses and maintain the best liquid-foam ratio. The beer tap at the top of the machine and the refrigerator below can be customized so that the dispenser can be aligned with the decoration of the store. Meanwhile, its software can be integrated with the membership system to allow businesses to analyze member information and personalize marketing.DataVan Targets Smart Retail with Hardware and Software PlanningFrom DataVan's exhibition this year, it is evident that AI is playing an increasingly important role in the retail and food service systems. DataVan pointed out that all AI-driven solutions and software-hardware integration have been developed by its teams. In the future, DataVan will continue to explore more potential solutions for AI and smart retail applications, and make more diverse planning for potential industrial applications and customers. At the same time, DataVan will work to provide more comprehensive smart software and hardware integration solutions to help customers improve management efficiency and reduce operating costs, thereby creating a quality experience for consumers.DataVan flexed its muscle at this year's COMPUTEX, showcasing its software-hardware integration capabilities by demonstrating a variety of smart retail solutions under the theme of "Smart Airport."

Skymizer, a leading AI system software provider, today announced that it had achieved top-tier performance in the MLPerf Tiny v1.1 benchmarks. Skymizer's submissions used both hardware/software co-optimization and pure compiler optimization, and both approaches achieved impressive results.In the hardware/software co-optimization case, Skymizer worked with hardware vendor Nuvoton to develop a custom microcontroller that was optimized for AI inference. This microcontroller, the M467HJHAN, operates at a frequency of 200MHz and has only 512KB of SRAM. Despite these limitations, Skymizer was able to achieve top-tier performance on the MLPerf Tiny v1.1 benchmarks in the entire ARM Cortex-M4 segment.In the pure compiler optimization case, Skymizer used its ONNC compiler to optimize the performance of AI models on typical microcontrollers. ONNC achieved a 1.24x speed-up on average compared with the open-source compiler. ONNC's performance gains were particularly impressive on the IC (Image Classification) benchmark, where it achieved a 1.31x speed-up. This was due to ONNC's ability to split large tensors into smaller sizes that fit within the memory constraints of the microcontroller. ONNC also achieved a significant speed-up on the VWW (Visual Wake Word) benchmark, where it achieved a 1.52x speed-up. This was due to ONNC's ability to handle sparse tensors. ONNC's unique optimization algorithm can remove sparsity in the tensor and keep the precision of the neural network at the same time.Skymizer's results demonstrate the company's leadership in AI system software. Skymizer's software stack is re-targetable to a wide range of hardware platforms, and it can achieve high performance on even the most resource-constrained devices. ONNC has been applied in many neural network accelerators, including recommendation systems for data centers, general accelerators for data centers, smart phones, surveillance systems, and microcontrollers. Skymizer is committed to providing its customers with the best possible tools and services, and ONNC is a crucial part of that commitment.For more information, please visit https://skymizer.com/ and request a demonstration.For the best ARM Cortex-M4 MCU solution, please visit https://www.nuvoton.com/.

Not slowed down by the global pandemic for these past couple of years, Xperi maintains a steady step to complete a Spin-Off to keep its core business secure and stable over technical innovations. For better differentiating the intellectual property (IP) business, Xperi Holding Corporation separates the company's IP licensing business to Adeia. And Xperi branding will continue focusing on the product business of the company. Talking about the IP licensing business of Adeia, there are lots of the business connection related to semiconductor industry covering IC packaging technologies, which are tightly linked with Taiwan supply chains.Xperi creates a leading consumer and entertainment technology business including several sub brands such as DTS, HD Radio, IMAX Enhanced, TiVo. After completed the merger procedures with TiVo in 2020, Xperi actively participates with Smart TV business ranging from content providers to consumer electronics device makers and automotive manufacturers leveraging TiVo Operating System. This powerful platform is strengthening the collaboration with Taiwan electronics manufacturing service providers, OEM/ODM makers and semiconductor chipmakers. In Computex2023 exhibition, Xperi joined with the partners to demonstrate a complete range of integrated solutions to expect the further market development in the next 6-month.IMEX Enhanced is fast forwarding to the PC market with rapid growthAaron Chen, Xperi General Manager of Taiwan and APAC, talks the product highlights in Computex2023. The first product focus is IMAX Enhanced program series demonstration – IMAX Enhanced is a partnership between IMAX and DTS to bring IMAX quality content and sound (provided by DTS) to the home. There are several major TV global brands joined IMAX Enhanced camp to enable consumers to more easily find, watch, and enjoy entertainment. SONY BRAVIA CORE is one of them to showcase the IMAX Enhanced TV product series in Computex2023.One thing he mentioned is the new HP Envy series laptop PC just announced in late April 2023 offering a seamless way for consumers to enjoy the benefits of IMAX Enhanced movie titles. This is the first IMAX Enhanced certified PC and represents a major milestone in the world of mobile home entertainment. There are more IMAX Enhanced certified notebook PC brands coming to be released in the next few months. Looking at business milestones, IMAX Enhanced was first available on TVs and AV receivers, Soundbar systems and loudspeakers before it expanded into mobile and now to laptops.DTS Play-Fi enables consumer desire to enjoy entertainment anywhere, anytime, on any deviceMoreover, the wireless speakers and Soundbar systems embrace DTS Play-Fi technology to show the values in the crowded consumer electronics market sector. For instance, the Philips 5.1 channels DTS Play-Fi Home Theater systems bring wireless surround sound capabilities to consumers' homes. These systems create seamless integration with wireless speakers by eliminating wires. The Play-Fi ecosystem continues to grow and add new partners including Soundbar systems, All-in-one PCs, display monitors, surround sound earbuds and Surround Speakers. This approach will be extended to new production matrix having a unique industry platform to address an ever-increasing consumer desire to enjoy entertainment anywhere, anytime, on any device.These new product development trends are encouraging the Taiwan IC vendors and chip makers quickly start to design and implement new SoC chip solutions integrated DTS Play-Fi technology to further expanding the market reach for entertainment use casesTiVo operating system gears automotive streaming marketXperi launches TiVo operating system to Smart TV Makers and welcomes Vestel as the first TV OEM partner. Vestel, one of the top three European TV brands, is a Turkish home and professional appliances manufacturer aiming to expand European TV and streaming business powered by TiVo Smart TVs. Notably, Xperi aims to serve large target markets, smart TV, IP TV and the automotive industry, along with an installed base of approximately 40 million TVs in North America market. One recent announcement from Xperi mentioned BMW has selected TiVo platform to provide customers access to a growing number of video content providers, including linear and on demand streaming services. The partnership between Xperi and BMW is targeting to introduce all sorts of streaming media to the cabin to the new BMW 5 Series in 2024. BMW's deployment of the Powered by TiVo platform will include a variety of country-specific content, offering news, movies and updates through Over-The-Air messaging technology.The integrated solutions create high quality meeting experiencesAnother technology is DTS:X Ultra, which includes several Virtualizer Enhancement software stacks for aiming to integrate with advanced telephony and video conferencing systems. These features are going to look at the further collaboration with laptop PC's CPU makers. The idea is leveraging the build-in sensing technology of new processors to create seamless integration with commercial laptops. With advanced technologies evolving to play an important role in driving future growth for electronics industries, Xperi invests dedicated engineering support teams in Taiwan to work with local electronics manufacturing partners and IC chipset design vendors to provide robust product solutions and looks forward to establishing successful, mutually profitable relationships.

How to leverage digital technologies, such as artificial intelligence (AI), to enhance the operational efficiency and product quality of factory production lines, reduce labor and costs, then achieve the urgent goal of net zero for sustainability is the most crucial concern for manufacturing companies today. To achieve smart manufacturing, it is essential to rely on collaborative efforts from interdisciplinary industry players, combining their expertise and resources. And all the discussions in the "Smart Manufacturing Summit" during Computex 2023 are around related topics.In the Summit, Speakers from ASUS, SIEMENS, and NVIDIA shared their solutions and relevant experiences in supporting the transformation and upgrading of the manufacturing industry.Digital Transformation is imminentThe world today is changing as it has never changed before. The rapidly changing market poses urgent challenges for industries. Customers not only focus on basic requirements such as product safety, intelligent features, and high quality but also have new demands, such as personalized. Whether it's buying a pair of sneakers or a car, most consumers want something unique. If manufacturers can provide personalized solutions based on customer preferences, they gain a competitive advantage in the market. This means that manufacturing companies accustomed to standardized mass production must make changes to their operational models.Manufacturers also face challenges from the goals of sustainability. According to statistics from World Economic Forum, global industrial production accounts for 37% of total energy consumption and contributes 30% of global greenhouse gas emissions. "Our resources are finite, so the industry must do more with less. Digitalization, automation, and the intelligent use of data are key to meet this challenge."Ramkrishna sees "Digital transformation" and "Net Zero transformation" are interconnected priorities for countries and enterprises to cope with the challenges today. "10 years ago I was talking to my customers in Taiwan about the topic of digital transformation. But somehow this topic never really accelerated. For the last one or two years this is massively accelerating, but what has been even more interesting or more positive is this topic around the net zero transformation.""For example, some of our customers in Taiwan are small and medium machine builders. Their customers in the US or Europe are asking them questions about how efficient the machine is, and what the common footprint of it. Because this is CapEx, which is gonna be sitting somewhere there for the next 25 years. " But just getting the data and getting transparent to the first step is already a big bottleneck for them, taking digital transformation is imminent.What Ramkrishna sees is everybody's loving IT and wants to get there, but everybody's running on OT. And the interesting part is sort of these two guys are not talking to each other very often. As an automation supplier, this is also the biggest challenge, and getting these guys connected is what SIEMENS is doing.SIEMENS is assisting Taiwanese companies in addressing these challenges from three directions: first, transform Taiwan's industries business model, balancing sustainability and circularity with other business objectives; second, tackling the complexity of Taiwan's industrial ecosystem, breaking down silos across industrial processes and ecosystems; third, assisting local customers to get from data to decisions, turning many sources of data into actionable sustainability indicators.Combining digital twin technology with software-as-a-Service (SaaS) on the cloud, and application development solutions, SIEMENS' Xcelerator open platform help manufacturers reduce costs while continuously optimizing. The company also offers the SIMATIC series industrial computers (IPC) Tensor Box, which incorporates NVIDIA GPUs and integrates Edge AI capabilities. Combined with their own 5G router and ASUS' 5G network infrastructure, it enables tighter integration between IT and OT, facilitating the digitization and net-zero transformation of industrial processes.Challenges of Enabling Smart ManufacturingAs Alber Wu, Division Director of ASUS' Server Business Unit mentioned, the first challenge to enable smart manufacturing is smooth integration between IT and OT. However, people from IT and OT departments have different mindsets, goals, and priorities, have the alignment between them is critical. Another challenge is having the convergence between lots of the systems on both the IT and OT sides."That's why we provide solutions from the cloud to the edge," he highlighted ASUS' industrial-grade servers for Edge AI inference and 5G network applications, operating at -5°C to 55°C, support GPUs in short-depth chassis, tool-less and cable-less design. "Those edge servers can help to provide better convergence, and they are scalable, effective, also flexible for end customers. We aim to provide the seamless integration of IT and OT systems.""In the past years, cloud computing is a big topic, and we see a lot of hyperscale data centers. We run workloads in the cloud. We talk about public cloud, private cloud, hybrid cloud. But now Edge computing is booming because of AI technology and 5G technology. It brings the benefit of real-time, reduced latency per source of bandwidth, to make the device more connected at edge side and far rage." Cherrie Lien, Senior Director of Product Marketing in ASUS Server Business Unit added.She also mentioned ASUS' software for supporting edge to cloud automation, such as ASUS Control Center, a centralized IT management tool of monitoring and controlling all types of ASUS' computing systems. And AISVision, an AIO toolkit uses ASUS AI techniques for the application of computer vision, offers a flexible training and inference architecture, for building the smart manufacturing scenario.Virtual Factory is the Solution for Smart ManufacturingAI plays such a critical role in smart manufacturing, it might be the reason why both SIEMENS and ASUS are in collaboration with their partner, NVIDIA. The company's platform has become essential in the digital transformation journey of numerous manufacturers. "But don't make it wrong, this is more of a collaboration platform. We work with SIEMENS, we work with ASUS, and we work with the whole ecosystem, to make it a transformation platform for all." Andrew Liu, Senior Data Scientist of NVIDIA, emphasized that it is not an achievement by NVIDIA alone, but a collaborative effort with ecosystem partners.Liu said the global manufacturing industry is a $45 trillion business, with 10 million factories, and the enterprises are facing unprecedented external pressures, including high inflation, supply chain disruptions, mass labor shortages, raising energy prices, and all the sustainability requirements. And NVIDIA believes that virtual factories are key for enabling sustainable manufacturing."We already saw a lot of big names, such as BMW, Mercedes-Benz, and GM, started to build up their virtual factories, if you are a vender trying to deal with them, trying to build the digital twin of your factory would be one way you can have the same language, and consistency." Liu added, "you build a factory not just for yourself, you build a factory for your customers, right? And you need to build a factory more sustainable, collaborate with so many different stakeholders with so many different perspectives. If you can have a virtual factory, you can have a much rapid experimentation about how the factory going to lay out, what kind of robotics fit in what kind of operation, and what AI automation needed to tap into your factory to make it more efficient. You can accelerate your decision-making process, reduce the weight, and optimize the process, the idea, or the vision."For all the reasons above, NVIDIA wants to work with customers to build software-defined, data-driven factories. "First, we need to make them connected. Whether you deploy the factory on the cloud or the hybrid environment, it needs to be cloud-native, based on new technologies like Microsoft's container." Liu mentioned NVIDIA AI Enterprise, cloud-native AI software with an extensive library of full-stack software, including AI solution workflows, frameworks, pre-trained models, and infrastructure optimization that accelerates the data science pipeline and streamlines the development and deployment of predictive AI models to automate essential processes and gain rapid insights from data.And NVIDIA Omniverse, an AI-enabled, physically simulated, and industrial-scale virtual-world engine that enables full-fidelity live digital twins. Last October, NVIDIA announced to partner with SIEMENS, to connect Omniverse and SIEMENS Xcelerator platforms to enable the industrial metaverse with physics-based digital models from SIEMENS and real-time AI from NVIDIA, speed companies' process of making decisions.NVIDIA and SIEMENS share a common vision that digital twins that can deliver productivity and process improvements across the production and product lifecycles. And companies of all sizes will be able to employ digital twins with real-time performance data; create innovative industrial IoT solutions; leverage actionable insights from analytics at the edge or in the cloud; and tackle the engineering challenges of tomorrow by making visually rich, immersive simulations more accessible.Holistic Industrial Security is EssentialLast, but not least, there is one major concern that worries users the most about building a smart, connected factory: cybersecurity."The more you connect, the more cybersecurity threats you have. But the answer is not to connect lesser, you have to connect a lot more." SIEMENS' Ramkrishna explained, in factories, there're a lot of old machines, if you are not connecting all those, the risks are much higher, but this also gets a lot more complicated. "This is where we are helping our customers with different methods. We're using the so-called defense-in-depth approach."SIEMENS believes, only a holistic industrial security concept can be effective against cyber threats, it needs considerations on three dimensions. First is the security of plant itself, needs physical access protection, security policies, and monitoring; the second is network security, the solutions include setup cell protection, perimeter network, trusted zones, or usage of firewalls and VPN tools; the third is system integrity, practices are hardening systems, building strategies of patch management, authentication, and access protection.Ramkrishna concluded, "We consult our customers on how they should look at their plant security, their network security, making sure that everything is working there. All our products are developed according to the latest protocols to make sure that it is no threats are there. We're considering our customers a lot on this topic and it's the most important part of our products, it is not something that even selling."Robert Chin, General Manager, Server Business Unit, ASUSAditya Ramkrishna, General Manager of Digital Industries, SIEMENS TaiwanCherrie Lien, Senior Director, Server Business Unit, ASUSAndrew Liu, Senior Data Scientist, NVIDIA