本網站使用瀏覽器紀錄 (Cookies) 來提供您最好的使用體驗，我們使用的 Cookie 也包括了第三方 Cookie。相關資訊請訪問我們的隱私權與 Cookie 政策。如果您選擇繼續瀏覽或關閉這個提示，便表示您已接受我們的網站使用條款。 關閉
Flagship W4 Earbuds Become the Most Accurate Interpreter Device Ever Released LAS VEGAS, Jan. 5, 2026 /PRNewswire/ -- Timekettle, the global leader in AI-powered cross-language communication, today announces a major system-wide upgrade that redefines what real-time translation can achieve. Central to this leap forward is Timekettle's new SOTA (State of the Art) Translation Engine Selector, a behind-the-scenes but transformative capability that dynamically identifies the optimal translation model for each language pair and conversational scenario in real time. SOTA Translation Engine Selector While invisible to the user, the impact is unmistakable: the fastest, most precise, and most natural translations Timekettle has ever delivered, regardless of accent, grammar structure, or speaking style. This upgrade arrives alongside an enhanced Bone-Conduction & Hybrid Algorithm, now offering even cleaner, more accurate voice capture by improving the purity of the user's vocal signal before it reaches the translation pipeline. Together, these advancements make the flagship W4 Interpreter Earbuds the most accurate in company history, surpassing the already industry-leading performance showcased at IFA 2025 that leaves its competition lost in translation. Since, hearing is believing, this innovation along with the full Timekettle portfolio can be experienced throughout CES at the Timekettle booth [LVCC, North Hall — 9163]. A Leap Forward in Machine Understanding: SOTA Translation Engine Selector Different language pairs behave like entirely different systems—English to Korean follows different logic than German to English; Spanish to Chinese requires different segmentation, prediction, and contextual inference than Japanese to French. Timekettle's SOTA Translation Engine Selector resolves this fundamental problem by: Detecting the language pair and context in real time Selecting the specialized AI engine best suited to that specific pair Optimizing accuracy, fluency, and naturalness for each exchange This is equivalent to assigning a "native specialist" to every conversation. This capability is made possible by Timekettle's advanced Large Language Model and the ongoing evolution of its Babel OS platform now updated with new machine-learning layers trained on millions of linguistic samples. The result: translations that sound less like a machine "attempting" to interpret—and more like a system that genuinely understands the intent, tone, and structure of what is being said. To Timekettle users, this means that no matter what languages they speak, Timekettle can always match them with the world's best translation engines to serve their needs, enabling them to enjoy the most accurate cross-language communication experience without fear, awkwardness, or hesitation; to truly express themselves boldly and communicate naturally. Next-Generation Bone-Conduction & Hybrid Algorithm: Superior Voice Capture for Superior Accuracy Building on its proprietary bone-voiceprint sensor technology, the enhanced 2026 algorithm upgrade extracts cleaner vocal vibrations to improve signal purity, isolates noise more effectively in chaotic environments, and more accurately identifies speakers during multi-voice conversations. These improvements work together to deliver far clearer acoustic input—the foundation of any high-precision translation—and ensure that the system captures meaning with exceptional fidelity. Because AI translation quality is only as strong as the audio it receives, this upgrade dramatically elevates performance across business meetings, travel scenarios, classrooms, and multilingual events. Rolling Out Across the Entire Timekettle Portfolio Beginning in early 2026, the SOTA Translation Engine Selector, updated Hybrid Algorithm, and extended Babel OS enhancements will roll out across the full Timekettle product family, including: W4 Interpreter Earbuds: Now the most accurate consumer translation earbuds Timekettle has produced W4 Pro: Elevated for professional meetings and enterprise use T1 AI Translator: The world's fastest offline translator now combines Edge AI with the SOTA Engine Selector to boost accuracy beyond its already industry-leading offline performance and ultra-fast 0.2-second response time. WT2 Edge, M3 Earbuds, and X1 Interpreter Hub: Each gaining improvements in context prediction, segmentation, and tonal translation Shared Capabilities Across the Timekettle Portfolio Across its expanding lineup of earbuds and handheld translators, Timekettle delivers a unified experience built around seamless, natural, real-time communication. All devices support instant translation across 43 languages and 96 accents, covering more than 95% of global regions to enable smooth, uninterrupted conversations anywhere in the world. Users benefit from all-day practicality as well, with long-lasting battery performance designed for travel, business, and extended multilingual interactions. Timekettle's conversation modes are consistent across the portfolio, offering flexible ways to communicate in different settings. One-on-One Mode allows two people to share earbuds for fluid, face-to-face dialogue. Listen & Play Mode records spoken audio through the app and provides translations that can be reviewed or replayed later. Speak Mode enables users to talk into their device and have translated speech broadcast clearly through the phone's speaker, ideal for group discussions, classroom settings, tours, or presentations. Whether using earbuds like the W4 or handheld devices such as the T1, the Timekettle ecosystem is designed to make multilingual communication effortless, intuitive, and adaptable to nearly any scenario. "Accuracy is the result of perfect collaboration between hardware and software," said Leal Tian, Founder & CEO of Timekettle. "With the introduction of our SOTA Translation Engine Selector and the next-generation bone-conduction algorithm, we've elevated that collaboration to a new level. W4 becomes the most accurate interpreter device we've ever created—and this upgrade strengthens every product across our portfolio." Availability The 2026 SOTA Translation Engine Selector upgrade and enhanced Hybrid Algorithm will begin rolling out via over-the-air updates across supported Timekettle devices starting January 2026, with the full portfolio supported throughout Q1. The Timekettle portfolio is available to buy at www.timekettle.co and Amazon. Prices include W4 Interpreter Earbuds ($349), W4 Pro ($449), T1 AI Translator ($299.99), WT2 Edge ($349.99), M3 Earbuds ($149.99), and X1 Interpreter Hub ($699.99). About Timekettle Since 2016, Timekettle has led innovation in AI-driven cross-language communication, with award-winning devices used by travelers, businesses, educators, and multilingual teams worldwide. Its products—including the W4 and W4 Pro Interpreter Earbuds, WT2 Edge/W3 series, X1 Interpreter Hub, and T1 AI Translator—deliver seamless, natural, and secure communication across nearly all global regions.
Next-generation AI-powered devices and experiences to be showcased at the company's exclusive exhibition through January 7 SYDNEY, Jan. 5, 2026 /PRNewswire/ -- Samsung Electronics Co., Ltd. today unveiled its "Companion to AI Living" vision at The First Look, its CES® 2026 event held in the Latour Ballroom at the Wynn Las Vegas. The focus of the event was on AI as Samsung's philosophy; a foundation that connects the company's R&D product development, operations, and user experience. TM Roh, CEO and Head of Samsung's Device eXperience (DX) Division, opened The First Look by describing the company's AI and how, due to its vast, AI-enabled, connected ecosystem, Samsung can provide users with an elevated experience in their daily life. This approach gives users a chance to access more than just the basics from their technology and instead, provides opportunities to find more meaningful moments everywhere. "Samsung is building a more unified, more personal experience across mobile, visual display, home appliances and services," said CEO TM Roh. "With our global connected ecosystem, and by embedding AI across categories, Samsung is leading the way to offer more meaningful everyday AI experiences." Entertainment Companion: Extending the Experience Beyond Simple Viewing SW Yong, President and Head of the Visual Display (VD) Business at Samsung Electronics, and Sukhmani Mohta, Chief Marketing and Partnerships Officer, VD Business at Samsung Electronics America took the stage to describe how Samsung's displays are combining hardware excellence and visual intelligence to provide a true entertainment companion. Drawing from twenty years of leading the TV industry, Samsung has built a full AI TV lineup that provides an entirely new way for users to interact with their TV. The centre of the display lineup is the 130-inch Micro RGB, which represents a monumental leap in scale and picture quality. The 130-inch Micro RGB marks a new era of colour, featuring the widest and most detailed spectrum ever seen in Samsung TVs while its Timeless Frame design helps to minimise distractions and allows the picture to take centre stage with understated elegance. A micro-sized RGB light source drives the incredible picture quality, with each micro-sized red, green, and blue diode shining independently to produce colour in its purest, most natural form. Micro RGB AI Engine Pro[1] enables precise control of RGB colors and creates overwhelmingly vivid picture quality in virtually every scene. Elevating this next-level viewing experience, Vision AI Companion (VAC) uses AI technology to work alongside users as an entertainment companion to enhance viewing. Samsung also offers intuitive modes to personalise the viewing experience. For soccer fans, AI Soccer Mode Pro delivers a more exciting gameday experience through AI-driven picture and sound tuning to near stadium-level quality. AI Sound Controller Pro[2] lets you raise or lower the volume of the crowd, commentary, or background music, providing a personalised listening experience for TV shows and movies. Users can simply make verbal requests, and any TV equipped with VAC – which includes Micro RGB, OLED, Neo QLED, and UHD TV[3] – contextually carries out those requests. The Movingstyle is designed to move easily throughout the home and kitchen appliances, for a multi-device experience achieved through complete ecosystem integration. Samsung has led the global soundbar market for 11 consecutive years[4]. This year, it is introducing two new WiFi speakers, the Music Studio 5 and 7, to further expand its integrated ecosystem. These models support a wider range of sound system combinations, enhance audiovisual quality, and improve the aesthetics of any space. Each model shares a timeless, dot design concept by renowned designer Erwan Bouroullec, inspired by a universal symbol in music and art, and grounded in Samsung's signature design language. Samsung also unveiled a host of new products that blend beautifully with users' homes and aesthetics. The new, ultra-thin Samsung OLED S95H features a refined bezel that gives it art gallery elegance, and Samsung's new portable projector, The Freestyle+, supported by VAC, allows users to view content on walls and ceilings, as well as uneven surfaces like corners and curtains. The 2026 TV lineup[5] supports HDR10+ ADVANCED, delivering enhanced brightness, genre-based optimisation, intelligent motion smoothing, advanced local tone mapping, and improved gaming experience. Samsung also unveiled its most advanced Odyssey gaming monitor lineup yet, introducing five new models that push the boundaries of resolution, refresh rate, and immersive visual performance. Led by Samsung's first 6K 3D Odyssey G9, the 2026 lineup debuts Samsung's first display technologies for gamers and creators, including the next-generation Odyssey G6 and three new Odyssey G8 models. At the centre of all these display innovations is the most powerful Tizen OS[6] yet. Users can now enjoy seven years of Tizen OS upgrades, ensuring that TVs continue to evolve long after they have entered the home. Simon Howe, Director – AV, Samsung Australia said "With leading innovation and scale spanning TVs, audio, home appliances, mobile experiences, and services, Samsung is uniquely positioned to deliver trusted, reliable, and cutting edge, AI-powered products that adapt to your routines and support everyday life in subtle, meaningful ways. Through optimised form factors and ongoing partnerships, Samsung has transformed products into companions, providing personalised, natural ways to interact with devices to meet your needs easily." Home Companion: Connected Smart Appliances That Guide You Through Your Day Cheolgi Kim, Executive Vice President and Head of Digital Appliances (DA) Division at Samsung Electronics, and Elizabeth Anderson, Head of Integrated Marketing, DA at Samsung Electronics America (SEA) outlined Samsung's vision to evolve from providing home appliances to true home companions that work to eliminate the stress of daily chores. The AI-enabled Family Hub refrigerator is the centerpiece of the home and now, with an upgrade to AI Vision built with Google Gemini, it refines living for an AI future. With this update, AI Vision unlocks existing limitations in recognising food items, seamlessly tracking what is placed into and taken out of the refrigerator, making meal planning and food management simpler than ever. The Family Hub has won 10 CES Innovation Awards, while Samsung's AI-enabled fridges have received the award for the past three years. Through a gamified "What's for Today?"[7], select refrigerators provide recipe recommendations based on what is in the refrigerator or also random recommendations, reducing the stress caused by choosing what to cook. When selected, recipes show up through SmartThings Food, where users receive a step-by-step guide to help them get started instantly. The selected recipe can also be sent to connected cooking appliances to begin the process seamlessly. Video to Recipe adds even more simplicity, providing users with recommended cooking videos and converting those videos into easy-to-follow steps, allowing users to follow along while they cook without pausing the video or backtracking. Samsung also unveiled FoodNote, a new weekly report that recaps users' food intake patterns, from most-used ingredients and recipe recommendations, and which items it's time to restock. Additionally, Now Brief[8] includes more widgets on the Family Hub screen and with Voice ID[9], can distinguish between family members to display content relevant to each individual. Ultimately, these features work together to provide helpful insights and information throughout the week. In the laundry room, the Bespoke AI Laundry Combo removes the need to transfer loads of laundry, solving a major pain point for households. This year's model comes with enhanced features like a faster super speed cycle and enhanced drying performance. Moreover, Samsung's new Bespoke AI AirDresser is here to solve another common problem with your clothes. It features Auto Wrinkle Care, which blasts strong air and steam jets to smooth out shirts. All users have to do is hang their shirt and wait, reducing the burden in busy mornings. Around the house, the Bespoke AI Jet Bot Steam Ultra, powered by a Qualcomm Dragonwing processor, features an Active Stereo 3D Sensor to recognise liquids like coffee, juice, or even transparent liquids like water. While its camera helps with navigation, it also allows the robot vacuum to serve as a monitoring device for when users are away from home. Plus, with a smarter Bixby, users can speak conversationally to their robot vacuum to carry out tasks with ease. For the deep level of connectivity and integration that forms the backbone of a holistic, AI-powered experience, Samsung's Bespoke AI appliances have received a CES Innovation Award. Care Companion: Shifting From Reactive to Proactive Care Finally, Praveen Raja, Vice President and Head of Digital Health at Samsung Research America (SRA), introduced Samsung's long-term vision for intelligent care enabled by Samsung's integrated device ecosystem, redefining care as a reactive need to a proactive opportunity. For example, Samsung aims to provide a personalised wellness experience by deepening integration within its ecosystem, offering effective exercise, sleep coaching and suggesting appropriate recipes based on ingredients available in connected refrigerators. Samsung Knox and Knox Matrix serve as the foundation of this hyper-personalised ecosystem, safeguarding user data at every turn. As AI is constantly evolving, so are Knox and Knox Matrix. To maintain security despite constant changes, Samsung's security systems are consistently identifying AI risks by advancing to protect data in AI training processes and approving models through red team analysis. The Samsung Exhibition Zone at CES is available to the public Jan. 4-7. For more information, visit the Samsung Newsroom. About Samsung Electronics Co., Ltd. Samsung inspires the world and shapes the future with transformative ideas and technologies. The company is redefining the worlds of TVs, digital signage, smartphones, wearables, tablets, home appliances and network systems, as well as memory, system LSI and foundry. Samsung is also advancing medical imaging technologies, HVAC solutions and robotics, while creating innovative automotive and audio products through Harman. With its SmartThings ecosystem, open collaboration with partners, and integration of AI across its portfolio, Samsung delivers a seamless and intelligent connected experience. For the latest news, please visit the Samsung Newsroom at news.samsung.com. [1] *Viewing experience may vary according to types of content and format. Upscaling may not apply to PC connection and Game Mode. [2] * Performance may vary depending on installation environment and surroundings. [3] Vision AI Companion is available on all 2026 models from UHD up and select 2025 models. [4] Source – Omdia, Feb 2025. Based on overall TV market share by manufacturer on an annual unit & revenue basis [5] HDR10+ ADVANCED will be available on the following models: QN990H, R95H, S95H, S90H, R850H, S85H, LS03HW, LS03H, and QN80H. [6] Internet connection and Samsung Account required. Data and subscription charges may apply. Apps and UI may be subject to change without notice. Agreement of the Smart Hub Terms and Conditions and Privacy Policy is required before use. Tizen OS upgrades will be provided for 7 years, starting from release year 2025. [7] A Wi-Fi connection and a Samsung account are required. [8] A Wi-Fi connection and a Samsung, Google, or Microsoft account may be required. You must use the same log-in account on each of the devices. Availability of certain apps may vary by country, region, service provider, network environment, or device, and may change without notice. [9] Each user must register for a Samsung Account on screen appliances in advance. Up to six accounts can be registered on a single device. To use Voice ID, users must have agreed to the Bixby terms and conditions. Voice ID registration is available on select appliances, TVs and mobiles, and only one voice can be registered per account.
HONG KONG, Dec. 30, 2025 /PRNewswire/ -- Insilico Medicine (3696.HK), a clinical-stage drug discovery and development company driven by generative artificial intelligence (AI), is successfully listed on the Hong Kong Stock Exchange today, becoming the first AI-driven biotech company to go public in Main Board under Chapter 8.05 listing rules of the HKEX. This initial public offering (IPO) raised a total of HKD 2.277 billion, achieving the largest biotech IPO in Hong Kong this year, as in the size of fundraising. Insilico Medicine "With this massively oversubscribed listing we set several world's firsts further confirming Insilico's validated leadership position in AI-powered drug discovery and development and the strength of our platform and pipeline. Insilico is dedicated to extending human productive longevity and this listing provides us with more resources to advance our mission", says Alex Zhavoronkov, PhD, Founder and CEO, Chief Business Officer of Insilico Medicine. "Over the past few years, we set very clear industry benchmarks demonstrating that AI can help make drug discovery faster, cheaper, and deliver higher success rates in preclinical and early clinical development. We have validated the end-to-end capabilities of AI-empowered programs from novel target discovery to molecular generation, and then to preclinical and clinical stages. Going forward, we will continue to increase investment in our AI platform and innovative pipeline, accelerate the advancement of differentiated innovative programs into clinic, and bring truly accessible, affordable, and breakthrough treatment solutions to patients worldwide." "Today marks not only an important step forward for Insilico Medicine, but also a new starting point for our deep integration of AI with life sciences and reshaping of the drug development paradigm", says Feng Ren, PhD, Co-CEO and Chief Scientific Officer of Insilico Medicine. "With our self-developed AI platform, we have identified innovative targets and novel molecules with first-in-class or best-in-class potential across multiple disease areas, and several programs have already entered clinical development. We believe that the value of AI extends far beyond cost reduction and efficiency gains; it lies in continuously pushing the boundaries of foundational innovation. Support from the capital markets will help us further enhance our AI platform, expand our pipeline, improve R&D efficiency, and collaborate with global partners to accelerate the translation of innovative medicines. Guided by a long‑term vision and driven by technology, we are committed to transform AI into real productivity that improves human health." Global Capital Endorsement: AI‑driven Biotech Lands on HKEX As a global pioneer AI‑driven biotechnology company, Insilico's Hong Kong listing, jointly sponsored by Morgan Stanley, CICC and GF Securities, attracted strong interest and active participation from both local and international investors. A total of 94,690,500 shares were offered globally, with 10% Hong Kong public offering, which was oversubscribed by approximately 1427.37 times, locking in subscription funds of over HKD 328.349 billion and setting a record for Hong Kong public offering subscription amount among non‑18A healthcare IPOs in Hong Kong during the year. The international offering accounted for 90% of the total, and was oversubscribed by 26.27 times, marking the most oversubscribed case in international placement among non‑18A healthcare IPOs in Hong Kong during the year. In terms of cornerstone, Insilico introduced 15 cornerstone investors globally, including Lilly, Tencent, Temasek, Schroders, UBS AM, Oaktree, E Fund and Taikang Life Insurance, among others, forming an all-star line‑up covering global pharmaceutical leaders, internet giants, international sovereign funds and large asset managers, as well as leading Chinese mutual funds and insurances funds. Notably, Lilly and Tencent for the first time subscribed as cornerstone investors in a biotechnology company, highlighting cross‑industry leaders' recognition of and confidence in the AI‑driven R&D business. Additionally, Oaktree Capital, a long‑term U.S. investment institution, returned to the Hong Kong biotech market as a cornerstone investor for the first time this year, reflecting its continued optimism about the long‑term value of the capital market and innovative pharmaceuticals. Furthermore, several international institutional investors also made their first entry into the Hong Kong capital market through this offering, further enhancing Hong Kong's position on the global capital map for healthcare and technology. Racing Towards AI-first Innovation: From GANs Pioneer to Pharmaceutical Superintelligence Insilico is at the forefront of generative AI application in life science research, aiming to tackle key scientific and technological challenges in life sciences from first principles using cutting‑edge generative AI. Since its inception, Insilico Medicine is committed to transparency and reproducibility, publishing hundreds of academic papers and presenting at frontier academic and industry conferences. In 2016, Insilico published a Nature Communications paper, proposing a novel computational method, iPANDA, to address core challenges in biomarker discovery and target-disease associations in omics data analysis. This pathway-based method has since evolved into multiple deep-learning generative algorithms behind Insilico's AI‑powered biological research powering the PandaOmics platform. In 2019, Insilico's publication in Nature Biotechnology described how its proprietary deep generative model, GENTRL, was used to discover a potent inhibitor of the kinase target DDR1. It took only 21 days from target identification to the generation of a molecule with drug‑like properties. This study is regarded as one of the landmark events in the rise of AI‑driven drug discovery and laid an important academic foundation for the widespread adoption of generative AI in drug discovery and development. These two early academic papers, along with a series of other published fundamental research, laid a solid foundation for Insilico to build Pharma.AI, its integrated end‑to‑end platform spanning biology, generative chemistry, clinical development and scientific research. In 2025 alone, Insilico Medicine scientists published 9 papers in Nature-family journals and 10 papers in the American Chemical Society's Journal of Medicinal Chemistry. Since 2020, the Pharma.AI platform has been launched and commercialized as a modular software suite, and its collaboration network has expanded globally. As of the latest practicable date, Insilico has entered into software licensing collaborations with 13 of the world's top 20 largest pharmaceutical companies. By continuously iterating its algorithms, building proprietary multimodal data warehouses, and improving infrastructure and tooling, Insilico has consistently optimized the performance of the Pharma.AI platform and extended its capabilities into emerging areas such as sustainable chemistry and agriculture, driving deep integration of AI‑driven innovation across the entire life science value chain. In December 2022, Life Star 1, the industry's first fully automated biology laboratory independently constructed by Insilico, was officially put into operation. Highly integrated with Insilico's AI platform, the lab closes the loop from in silico design to experimental validation, enabling a highly efficient "lights‑out factory" capable of unmanned operation. While supporting wet‑lab validation for both internal and partnered projects, it continuously accumulates high‑quality proprietary datasets to further power the evolution of the AI platform. In September 2025, the upgraded Life Star 2 was officially launched with significantly enhanced cross‑island equipment coordination and multi‑workflow parallel processing capabilities, establishing a more convenient, stable and highly scalable intelligent lab system through refined allocation algorithms and response mechanisms. In 2025, Insilico first proposed the concept of Pharmaceutical Superintelligence, defining it as the next stage of AI‑driven drug R&D, which aims at a fully AI‑driven, autonomous, and intelligent system for drug discovery and development. Insilico's experimentally-validated purpose-built models can now be used to make other frontier multimodal models including third-party models more intelligent and capable in multiple scientific discovery tasks. Global Leading AI R&D Capabilities: From First AIDD Milestone to Global Benchmarks By integrating its proprietary AI‑enabled drug discovery platform with advanced automated laboratory capabilities, Insilico has significantly improved the efficiency of drug R&D in practice. It has not only delivered pioneering innovation milestones, but also established a benchmark for AI‑driven drug discovery. Compared with the average of 4.5 years typically required for traditional early‑stage drug R&D, Insilico reduced the average time from program initiation to preclinical candidate (PCC) nomination to 12–18 months across more than 20 in‑house programs between 2021 and 2024, with only 60–200 molecules synthesized and tested per program. The most representative example is Rentosertib (ISM001‑055), a novel mechanism‑of‑action candidate for idiopathic pulmonary fibrosis (IPF) empowered by AI. This program is considered the most advanced first‑in‑class AI‑discovered drug currently in development. Insilico completed the early discovery phase in just 18 months and identified the PCC after screening and testing only 78 molecules. The AI‑driven R&D journey of this program, from target discovery and molecule design through to Phase I clinical studies, was published in Nature Biotechnology in March 2024, providing the first systematic demonstration of an AI‑enabled, end‑to‑end drug discovery process from scratch. Also in 2024, Insilico completed the China Phase IIa clinical trial for Rentosertib, with favorable safety and tolerability demonstrated, as well as an emerging dose‑dependent efficacy trend in IPF patients, indicating the potential to reverse the disease progression. The results were published in Nature Medicine in May 2025 and are regarded as the first clinical proof‑of‑concept milestone in the AI drug discovery field, providing valid practical evidence for AI‑driven drug discovery. In addition, Insilico continues to extend its specialized AI‑driven capabilities into more areas of high unmet medical needs, building a structured portfolio of more than 30 programs centered around key therapeutic areas including fibrosis, oncology, immunology, inflammation, cardiometabolics and central nervous system (CNS) disease. Among all the innovative pipelines, 10 have received IND clearance, 7 of them are currently in active clinical development. Meanwhile, AI‑driven R&D collaborations and pipeline out‑licensing are expanding. As of the latest practicable date, Insilico has achieved 3 out‑licensing deals with global pharmaceutical companies including Exelixis and Menarini, with the maximum total deal value up to$2.1 billion. Furthermore, co‑development partnerships with global pharma leaders including Fosun Pharma, Sanofi and Lilly provide solid validation and powerful boost for the value of AI applications in the global drug discovery industry. AI&DD Dual‑engine Drive: Global Appearance Attract Long‑term Support for Industry Paradigm Shift Since its founding in the United States in 2014, Insilico Medicine has expanded its business to four continents with the pioneering AI&DD dual‑engine management model, driving an in-depth combination of AI algorithm innovation and drug development expertise to achieve efficient and balanced growth. As of the latest practicable date, under the leadership of its two Co‑CEOs, Insilico boasts an R&D team of 249 experienced scientists, with master and doctorate degree holders accounting for 87% of this highly international team. Under this dual‑engine management framework, the two core leaders bring their unique strengths to work in close synergy. Alex Zhavoronkov, PhD, Founder, Chief Executive Officer and Chief Business Officer of Insilico, focuses on the application of cutting‑edge AI technologies and is responsible for overall corporate strategy and AI innovation. Dr. Zhavoronkov is widely recognized for his academic influence and research productivity, being listed 3 times among the Clarivate Highly Cited Researchers, with an H‑index of 72 assessed by Google Scholar. The drug discovery and development team is led by Feng Ren, PhD, Co-CEO and Chief Scientific Officer of Insilico Medicine, an experienced drug hunter, who is responsible for guiding R&D exploration directions and overseeing the execution of pipeline asset development. Dr. Ren has devoted nearly 20 years to innovative drug R&D and has authored more than 80 publications in peer-reviewed journals, and holds approximately 120 patents in total. Committed to open-source academic research and knowledge sharing, Insilico maintains active communication with the industry with continuous publication about technological breakthroughs and industrial milestones. The company authors more than 300 peer‑reviewed papers and holds over 700 patents or applications. Supported by its outstanding contribution, Insilico was recognized in the Nature Index 2024 list of the world's top 100 global corporate institutions for biological sciences and natural sciences publications. In 2025, one of Insilico's studies was selected among the top 10 advances by Nature Biotechnology, and featured on the journal's December cover. Following the IPO, Insilico plans to allocate approximately 48% of the net proceeds to fund further clinical research and development of our key clinical stage pipeline drug candidates, 20% to fund the research and development, for early-stage drug discovery and development, 15% for development of new generative AI models and the associated validation research work, 12% for the further development and expansion of our automated lab, and 5% for working capital and other general corporate purposes. While consolidating the strengths of its "AIDD dual‑engine" model, the company will continue to integrate global resources and seek collaboration opportunities to accelerate the fundamental paradigm shift across the biopharmaceutical industry. About Insilico Medicine Insilico Medicine is a pioneering global biotechnology company dedicated to integrating artificial intelligence and automation technologies to accelerate drug discovery, drive innovation in the life sciences, and extend healthy longevity to people on the planet. The company was listed on the Main Board of the Hong Kong Stock Exchange on December 30, 2025, under the stock code 03696.HK. By integrating AI and automation technologies and deep in-house drug discovery capabilities, Insilico is delivering innovative drug solutions for unmet needs including fibrosis, oncology, immunology, pain, and obesity and metabolic disorders. Additionally, Insilico extends the reach of Pharma.AI across diverse industries, such as advanced materials, agriculture, nutritional products and veterinary medicine. For more information, please visit www.insilico.com
Editor's Summary Ganzin will introduce AURORA IIE, its latest ASIC-based eye-tracking solution designed for next-generation AR/AI glasses and VR/MR headset. AURORA IIE moves eye-tracking computation from an NPU to a proprietary Eye Processing Unit 2 (EPU2), reducing power consumption to 1/4 of NPU-based solutions, enabling true always-on eye-tracking. Always-on eye-tracking allows gaze interaction and continuous capture of eye-based physiological indicators. AURORA IIE supports up to 120 Hz throughput, empowering advanced VR/MR devices, medical applications, and academic research. Ideal architecture: one ASIC processor (EPU2), two eye sensors, and two IR LEDs — maintaining the same ultra-compact design philosophy of Ganzin's previous AURORA IIS platform. With the introduction of AURORA IIE, Ganzin now delivers a complete, scalable eye-tracking product line — from software IP to turnkey NPU solutions and finally an ultra-low-power ASIC platform ready for mass production. NEW TAIPEI CITY, Dec. 30, 2025 /PRNewswire/ -- Ganzin, a pioneer in glint-free wearable eye-tracking technology, is excited to announce the global debut of its next-generation AURORA IIE eye-tracking solution at CES 2026. Building on the success of last year's AURORA IIS NPU solution, AURORA IIE introduces a custom-designed ASIC processor--Eye Processing Unit 2 (EPU2)--dedicated solely to eye-tracking computation. For years, the high computational power required for eye-tracking has prevented AR/AI glasses from integrating this feature in a truly power-efficient manner. While AURORA IIS successfully offloaded processing from the host platform to a compact NPU, enabling substantial power savings, the move to ASIC marks another breakthrough. EPU2: The Future of the Edge AI Computing for Wearable Eye Tracking AURORA IIE is powered by Ganzin's in-house developed ASIC, EPU2, engineered specifically for low-power, high-performance eye-tracking on wearable platforms. Compared to last year's NPU-based architecture, the new ASIC achieves: 1/4 the power consumption of an NPU Always-on operation for gaze interaction and biometric sensing Ultra-compact form factor optimized for AR/AI smartglasses 120 Hz processing throughput, meeting the demands of VR/MR devices, medical institutions, and research labs Ganzin AURORA IIE A Seamless, Minimalist Hardware Architecture The AURORA IIE solution maintains the simplicity of Ganzin's signature design: One dedicated ASIC processor, 3.6 x 3.6mm Two miniature eye sensors Two ultra-small IR LEDs Flexible Product Portfolio and Business Models for Seamless Integration With the launch of AURORA IIE, Ganzin now offers a fully scalable eye-tracking product portfolio tailored to different development stages and integration needs: AURORA-II — Software IP Licensing A proven, highly flexible eye-tracking software stack that delivers robust performance across diverse hardware platforms. Ideal for partners seeking maximum design freedom and customization capability. AURORA-IIS — NPU-Based Turnkey Solution A production-ready reference design leveraging partner NPUs to accelerate system integration and shorten time-to-market, while maintaining excellent flexibility, power efficiency, and performance. AURORA-IIE — ASIC-Based EPU2 Platform Ganzin's next-generation ASIC solution featuring the EPU2 processor, delivering ultra-low-power, always-on eye-tracking in a compact, procurement-ready hardware form factor for mass-production devices. Together, the AURORA family enables customers to adopt the right solution at every stage — from early prototyping to large-scale commercial deployment. Leadership Statement "AURORA IIE represents the next evolutionary stage of wearable eye-tracking," said Dr. Shao-Yi Chien, CEO of Ganzin. "Our custom ASIC, EPU2, enables unprecedented power efficiency and allows always-on operation, opening the door to new applications in interaction, wellness, and research." Experience AURORA IIE at CES 2026 Ganzin will host private demonstrations of the AURORA IIE solution at its executive suite in the Westgate Hotel during CES 2026. About Ganzin Ganzin is a global leader in lightweight, compact, and energy-efficient eye-tracking technology for next-generation wearable and immersive devices. With a mission to create the most practical, easy-to-use eye-tracking solution for everyone, Ganzin provides hardware and software total solution that enable seamless integration into AR/AI glasses, VR/MR headsets.
6,000 attendees from 60 countries gathered at HKCEC to explore the latest innovations SIGGRAPH Asia 2025 fueled discussions on the future of computer graphics with leaders from Vicon, Adobe, Tencent, Pixar's Renderman, and more Artists and creators were celebrated with awards for their outstanding achievements Kuala Lumpur named host city of SIGGRAPH Asia 2026 in Historic Debut HONG KONG, Dec. 23, 2025 /PRNewswire/ -- SIGGRAPH Asia 2025, themed "Generative Renaissance," has successfully come to a close at the Hong Kong Convention and Exhibition Centre. Over 6,000 attendees from 60 countries gathered to explore the latest innovations in computer graphics and interactive techniques. "I extend my deepest gratitude to the entire program committee and the innovators who traveled from around the world to participate in SIGGRAPH Asia 2025," said Taku Komura, Conference Chair and Professor at The University of Hong Kong. "This collective effort has showcased our city's vibrant ecosystem and reinforced Hong Kong's role as a hub for technological advancement." "As we celebrate the success of SIGGRAPH Asia 2025, we recognize the creative minds behind every pixel, animation, and innovation," said Prakash Ramajillu, SIGGRAPH Asia Chief Staff Executive & General Manager, Koelnmesse Pte Ltd. "Our community of technologists and innovators continues to push the boundaries of what's possible and we're honored to be part of your journey." Announcing SIGGRAPH Asia 2026: A Landmark First in MalaysiaThe Hong Kong edition of SIGGRAPH Asia 2025 concluded with a warm welcome from the 2026 Conference Chair, Frank Guan of the Singapore Institute of Technology. The 19th edition of the Conference and Trade Exhibition will take place at the Kuala Lumpur Convention Centre in Malaysia from 1-4 December 2026, under the theme "Weaving the Future." 2025 Keynote Spotlight: Influential Talks from SIGGRAPH Asia HK EditionSIGGRAPH Asia 2025 featured engaging keynotes that sparked excitement among attendees. Harry Shum discussed the economic potential of low altitude airspace, setting the stage for innovative opportunities. Jamie Shotton highlighted the rise of synthetic data in AI, showcasing its transformative impact on autonomous vehicles. Dr. Chunchao Guo captivated the audience with his insights on crafting intelligent 3D worlds, particularly through the groundbreaking HunyuanWorld model. These thought leaders inspired a forward-looking dialogue about the future of technology and creativity. The 2025 Featured Sessions highlighted industry leaders in animation, VFX, virtual production, and digital storytelling, with standout presentations that captivated audiences. "Alien Abduction! The Truth Behind the Making of Pixar's Elio," from the talented Pixar Animation Studios team: Claudia Chung Sanii, David Luoh, Emron Grover, Jordan Rempel, Renee Tam, and Sajan Skaria. The all-star roster of speakers continued with Paul Debevec (Eyeline), Richard Chuang (d1n0), and Momo Wang (Animation Director and Creator of Tuzki), Scott Ross (Co-founder of Digital Domain), Raman Hui (Monster Hunt, Shrek), and Chris Bremble (Base Media, Ne Zha 2). The lineup was further enriched by contributions from Shalini De Mello (NVIDIA) and Alex Woo (IN YOUR DREAMS, Kuku Studios), making for a dynamic series of sessions. Celebrating Excellence: Technical Papers Awards at SIGGRAPH Asia 2025The Technical Papers Awards at SIGGRAPH Asia 2025 highlighted cutting-edge research in computer graphics. In addition, several Honorable Mentions recognized outstanding contributions, including works on UAV simulation, color calibration for 3D printers, and perceptual quality metrics, highlighting the depth of talent in the field. The Test of Time Awards celebrated influential past works, reinforcing the ongoing legacy of excellence in the technical papers program. This event continues to be a cornerstone of SIGGRAPH Asia. See full list of winners HERE. Exhibitor Talks at SIGGRAPH Asia 2025The Exhibitor Talks at SIGGRAPH Asia 2025 showcased exciting advancements in technology and creativity. Notable presentations included "To Infinity … and RenderMan XPU!" by Wendy Wirthlin and Dylan Sisson from Pixar Animation Studios, and "Crafting the Visual Language of K-Pop Demon Hunters" by Elise Testa and Jiwoon Kim from Foundry. Andrei Krovopuskov from Faceform presented "Ultra-Accurate 4D Facial Capture with Stereo HMC," while the "Tencent Workshop on Neural Graphics & Generative AI" explored cutting-edge AI developments. The "Adobe Substance 3D Keynote and Demo" featured insights from Pierre Bosset and industry guests, and Andrew Ray from Vicon discussed "VTubing: How Precision Motion Technology Powers the Next Era of Creativity."Exhibition Highlights at SIGGRAPH Asia 2025SIGGRAPH Asia 2025 featured over 70 exhibitors and sponsors, showcasing innovations from Adobe, FORUM8, Foundry, Huawei Technologies, Meta's Reality Labs Research, Morpho Inc., NVIDIA, Pixar's RenderMan, Tencent, Vicon and more. With 600+ expert speakers and hundreds of sessions across Technical Papers, Featured Sessions, Art Papers, Courses, Educator's Forum, Games, Posters, Technical Communications, the Art Gallery, XR, Emerging Technologies, Real-Time Live! and the Computer Animation Festival, the event provided a wealth of knowledge and inspiration. For latest updates, please visit: https://asia.siggraph.org/2025/ About ACM SIGGRAPH ACM SIGGRAPH is an international community of researchers, artists, developers, filmmakers, scientists, and business professionals with a shared interest in computer graphics and interactive techniques. A special interest group of the Association for Computing Machinery (ACM), the world's first and largest computing society, our mission is to nurture, champion, and connect like-minded researchers and practitioners to catalyze innovation in computer graphics and interactive techniques. ACM SIGGRAPH offers two of the premier technology conferences in the world, one in North America and one in Asia. Attended by tens of thousands of computer graphics professionals, SIGGRAPH and SIGGRAPH Asia are highly respected venues for the presentation of new computer graphics technology and research. For more information, please visit www.siggraph.org. About Koelnmesse Koelnmesse Pte Ltd is one of the world's largest trade fair companies. Its more than 80 trade fairs and exhibitions have the broadest international scope in the industry, as 60 percent of the exhibitors and 40 percent of the visitors come from outside Germany. The Koelnmesse events include leading global trade fairs for 25 sectors, such as Imm Cologne, Anuga, IDS, INTERMOT, Interzum Cologne, gamescom, and the International Hardware Fair Cologne. Koelnmesse is ACM SIGGRAPH's event organizer for the last 17 editions of SIGGRAPH Asia. For more information, please visit www.koelnmesse.asia. For further press information, interviews and on-site press event, please contact: Wani Diwarkar Paulene Lynch PR Bespoke PR Bespoke E: wani@prbespoke.com E: paulene@prbespoke.com M: +65 9832 0643 M: +65 9057 4101
Together, MINISFORUM and AMD are transforming the landscape of intelligent computing, setting new standards for power, efficiency, and creativity. BEIJING, Dec. 22, 2025 /PRNewswire/ -- On December 10th, MINISFORUM and AMD are proud to announce a high-level AI Application Product Experience Event, "Leading Intelligent Computing, Renewing the Core for the Future," in Beijing, China. This unique event, co-hosted by MINISFORUM and AMD, highlights two relatively new flagship AI computing products: the AI Mini Workstation MS-S1 MAX and the AI NAS N5 Pro, while exploring the wide-reaching impact of AI-powered innovation for enterprises and individuals. (Left to Right) Xie Huoping, Senior Product Manager of MINISFORUM; Jonathan Liao, Director of Consumer Marketing, Greater China, AMD; Tim Li, Director of Emerging Business Department, Greater China, AMD; Jovi Chi, Senior Director of Marketing, Greater China, AMD; Roy Jiang, Chairman of MINISFORUM; Gai Bintao, Channel Director of MINISFORUM; Jiang Xue, Product Director of Workstation & NAS, MINISFORUM; Emily Yeh, Director of Brand & Marketing, MINISFORUM Through live demonstrations, onsite product experiences, and in-depth discussions, attendees witnessed how MINISFORUM and AMD are working together to redefine the frontier of intelligent computing with next-generation performance, robust AI capabilities, and flexible deployment for real-world productivity. "Our vision at MINISFORUM is to accelerate the transformation of intelligent computing for everyone," said Roy Jiang, Chairman of MINISFORUM. "Partnering with AMD enables us to leverage world-class AI processors, empowering users to reach new levels of productivity and creativity. The rapid evolution of global AI trends motivates our continuous product innovation and drives real value for businesses and personal applications." Jovi Chi, Senior Director of Marketing at AMD, stated: "The cooperation between AMD and MINISFORUM fully demonstrates how AMD's technologies empower endpoint and edge AI. Powered by Ryzen™ AI Max and Ryzen™ AI series processors, MINISFORUM's products have achieved cutting-edge performance, accelerated AI workflows, and supported multi-scenario applications across industries. We look forward to advancing AI+ endpoint and edge computing together with MINISFORUM." AI Mini Workstation MS-S1 MAX: 4-Node Cluster for Local LLMs Anywhere The MS-S1 MAX is equipped with the advanced AMD Ryzen™ AI Max+ 395 CPU (16 cores, 32 threads) and Radeon™ 8060S GPU, making it a compact, ultimate-performance AI mini workstation. Supporting up to 128GB LPDDR5x-8000MT/s quad-channel unified memory and featuring high-speed USB4 V2, dual 10GbE LAN, Wi-Fi 7, and Bluetooth 5.4, the MS-S1 MAX can scale from a desktop to rack deployments. It can locally run large models such as Llama 109B and DeepSeek-R1 70B, and cluster multiple units to handle the DEEPSEEK 671B Q4 LLM. Real-world applications include content creation, research, rapid AI inference, and edge computing. Compared with a traditional 5U RTX 5090 server for large-model inference, when handling the DEEPSEEK 671B Q4 LLM, an MS-S1 MAX four-node cluster offers about 50% less volumn, 80% lower power consumption, and up to 77% cost savings. Each node draws only 180W at full load, so the entire cluster operates at just 0.72 KW, much less than 4-5 KW of 5U RTX 5090 server at full load, making it an exceptionally efficient solution for on-premises large-model inference. MS-S1 Max is available on MINISFORUM official brand shop, and currently on Christmas sale at $2,639.90. MS-S1 Max with AMD Ryzen AI MAX+ 395 chip, 4-node cluster display AI NAS N5 Pro: The First NAS to Realize AI PC Operation Powered by AMD Ryzen™ AI 9 HX PRO 370 (12 cores, up to 5.1GHz) and up to 80 TOPS neural processing power, the MINISFORUM N5 Pro redefines secure and intelligent storage for homes and small teams. It supports up to 144TB hybrid storage (5 HDDs + 3 SSDs), up to 96GB DDR5 ECC RAM, and advanced expansion via native PCIe x16 and OCuLink interfaces. The N5 Pro can be an AI PC while securing data privacy at the same time, and runs on MinisCloud OS for local LLM deployment, enhanced privacy, and versatile AI-powered tasks—including an intelligent photo library, secure media center, work vault, and smart home data solutions. MINISFORUM and AMD remain committed to advancing the future of intelligent computing through collaboration and innovation in the integration of AI PC in compact form factor. Together, they continue to set new benchmarks for performance, security, and adaptability in next-generation enterprise and personal AI solutions. AI NAS N5 Pro is available on MINISFORUM official brand shop, and currently on Christmas sale at $1,019. About MINISFORUM Founded in 2018, MINISFORUM is dedicated to "Bringing technology into everyday life." The brand applies its AI research to PC design, production, and manufacturing, offering high-performance computer solutions across AI Mini PCs, AI Mini Workstations, AI Mini Gaming PCs, AI NAS, and accessories. Now MINISFORUM has over 4 million users worldwide, with a presence in nearly 100 countries. For more information, please visit: https://www.minisforum.com/ © 2025 MINISFORUM. All rights reserved.
A12 藝術空間
Edge AI
請先登入後才能發佈新聞。
還不是會員嗎？立即 加入台灣產經新聞網會員 ，使用免費新聞發佈服務。 （服務項目） （投稿規範）
