• EPS NT$11.01, a 17-year high, core business profitability improving • 2025 outlook is for strong growth; 1Q25 AI server revenue to more than double • Seeing sustained demand for AI infrastructure and strong capex of major CSPs • To complete contract signing in 1-2 months for collaboration with Japan automaker---14 March 2025, Taipei, Taiwan – Hon Hai Technology Group (“Foxconn”) (TWSE:2317) today announced its full year and fourth quarter 2024 financial results. Full year 2024 net profit reached NT$152.7 billion, resulting in an earnings per share of NT$11.01, a 17-year high. At the same time, the Group announced that it will distribute a cash dividend of NT$5.80 per share, a record level since its listing in 1991. Despite significant changes in global tariff policies, geopolitical risks and monetary policies, the outlook for 2025 is for strong growth. The company also shared key developments around its three major smart platforms, as well as on AI and EV in 2025. For the full year 2024, revenue totaled NT$6.86 trillion, up 11.3% on-year. Gross profit reached NT$428.9 billion, up 10.6%; operating net profit, at NT$200.6 billion, was up 20.5%; while net profit rose 7.5% to NT$152.7 billion for the same time period. Gross profit margin, operating profit margin and net profit margin stood at 6.25%, 2.92% and 2.23% respectively, compared with 6.30%, 2.70% and 2.31% in the previous year. With profitability of the core business improving significantly, EPS for the year, at NT$11.01, was up by NT$0.76 on-year. In the October-December quarter, revenue reached NT$2.13 trillion, rising 15% on-year; gross profit reached NT$131 billion, up 15.7% from the same three months a year ago; operating net profit was NT$64.5 billion, showing an on-year gain of 31.8%; however net profit, at NT$46.3 billion, fell 12.8% for the same time period. Gross profit margin, operating profit margin and net profit margin were 6.15%, 3.03% and 2.17% respectively, compared to 6.12%, 2.64% and 2.87% in the same period a year ago, showing an increase in two of the three margins that indicated an improvement in profitability of the core business. EPS for the final three months of 2024 was NT$3.34, down NT$0.49 from the same period a year ago. Chairman of Hon Hai Technology Group (Foxconn), Young Liu, said fourth quarter revenue grew strongly both on-quarter and on-year. The record high NT$6.86 trillion in last year’sannual revenue, came as three of four primary product segments – cloud and networking; computing; components and others – all achieved strong growth. The actual performance exceeded expectations. Based on the Group's goal of an average payout ratio of no less than 40%, Chairman Liu announced a cash dividend of NT$5.80 per share to be distributed for this year, a significant increase from last year's NT$5.40 per share. The payout ratio reached 52.68%, a new high since the company was listed in 1991, and exceeding 50% for the sixth year in a row. Looking ahead for 2025, Chairman Liu said the full-year operating outlook will be for strong growth, and that first quarter performance will also turn in strong growth compared to the same period last year. The company is now positioned as a technology manufacturing platform service provider. The three major platforms – Smart Manufacturing, Smart EV, Smart City – are the important foundation for developing various technologies and businesses. These platforms are based on key components, modules, systems, software and other fundamental capabilities we have soundly established. In the future, Foxconn will introduce more generative AI to make them even more powerful. Foxconn will work with global Tier-1 partners to jointly develop and create generative AI solutions or develop large language models in-house to introduce into its three major intelligent platforms. By doing this, it will improve overall operational efficiency, and also enhance competitive advantage. The Group is using AI to produce AI; in Japan we have built a fully automated mobile phone assembly line. This is a good example of our move towards AI. In the face of market concerns regarding the impact of geopolitics and tariffs on the Group's operations, Chairman Liu said overall ICT industry demand is generally stable this year. In response to changes, we will maintain close contact with our customers and due to our comprehensive global layout make timely adjustments. Through vertical integration, a diversified footprint and close cooperation with customers, the Group will maintain its market leading position. Chairman Liu predicted that 2025 will be the first year of AI. Global demand is being sustained for AI computing infrastructure, and the capex of major CSPs will maintain strong growth this year. Looking at the first quarter, Foxconn’s AI server revenue is expected to grow more than 100% both quarter-on-quarter and year-to-year. As the GB200 enters mass production, we believe AI server revenue will improve quarter by quarter. Looking at the full year, AI server revenue will reach the trillion-dollar scale, in local currency terms, accounting for more than half of the Group’s total server share. Foxconn has beenan important co-development partner in new products with major customers, ensuring that we can participate in next generation upon next generation of product development. The company also believes that the technical methods used by DeepSeek will accelerate the popularization of AI servers for inference, thereby driving an increase in overall computing power demand. For large enterprises and CSPs, high-end AI servers and complete network and storage architectures are still indispensable, so demand for high- end AI servers will continue to grow. In the long run, as the development of edge AI gradually matures, there is likely to be more demand for upgrading personal devices. AI server training takes a long time. During this period, if there is a failure in the AI server, it will have a great impact on the training time or even the results. Foxconn has always been in a leading position in terms of quality stability. A 40% market share is our most conservative target. We are confident that we can continue to strive for a higher market share. Regarding collaboration with Japanese automakers, Chairman Liu said he anticipates “contract signing to be completed in one to two months’ time.” During the second half of the year, the MODEL B and the North American variant of the MODEL C are expected to enter mass production, while the manufacturing factory for the electric bus is scheduled to complete certification in the third quarter, expanding shipment capacity.

10 March 2025, Taipei, Taiwan – Hon Hai Research Institute announced today the launch of the first Traditional Chinese Large Language Model (LLM), setting another milestone in the development of Taiwan’s AI technology with a more efficient and lower-cost model training method completed in just four weeks. The institute, which is backed by Hon Hai Technology Group (“Foxconn”) (TWSE:2317), the world’s largest electronics manufacturer and leading technological solutions provider, said the LLM – code named FoxBrain – will be open sourced and shared publicly in the future. It was originally designed for applications used in the Group’s internal systems, covering functions such as data analysis, decision support, document collaboration, mathematics, reasoning and problem solving, and code generation. FoxBrain not only demonstrates powerful comprehension and reasoning capabilities but is also optimized for Taiwanese users' language style, showing excellent performance in mathematical and logical reasoning tests. "In recent months, the deepening of reasoning capabilities and the efficient use of GPUs have gradually become the mainstream development in the field of AI. Our FoxBrain model adopted a very efficient training strategy, focusing on optimizing the training process rather than blindly accumulating computing power,” said Dr. Yung-Hui Li, Director of the Artificial Intelligence Research Center at Hon Hai Research Institute. ”Through carefully designed training methods and resource optimization, we have successfully built a local AI model with powerful reasoning capabilities." The FoxBrain training process was powered by 120 NVIDIA H100 GPUs, scaled with NVIDIA Quantum-2 InfiniBand networking, and finished in just about four weeks. Compared with inference models recently launched in the market, the more efficient and lower-cost model training method sets a new milestone for the development of Taiwan's AI technology. FoxBrain is based on the Meta Llama 3.1 architecture with 70B parameters. In most categories among TMMLU+ test dataset, it outperforms Llama-3-Taiwan-70B of the same scale, particularly exceling in mathematics and logical reasoning (For TMMLU+ benchmark of FoxBrain, please refer to Fig.1). The following are the technical specifications and training strategies for FoxBrain: ●      Established data augmentation methods and quality assessment for 24 topic categories through proprietary technology, generating 98B tokens of high-quality pre-training data for Traditional Chinese ●      Context window length: 128 K tokens ●      Utilized 120 NVIDIA H100 GPUs for training, with total computational cost of 2,688 GPU days ●      Employed multi-node parallel training architecture to ensure high performance and stability ●      Used a unique Adaptive Reasoning Reflection technique to train the model in autonomous reasoning ●      Fig. 1：TMMLU+ benchmark results of FoxBrain, Meta-Llama-3.1-70B and Taiwan-Llama-70B   In test results, FoxBrain showed comprehensive improvements in mathematics compared to the base Meta Llama 3.1 model. It achieved significant progress in mathematical tests compared to Taiwan Llama, currently the best Traditional Chinese large model, and surpassed Meta's current models of the same class in mathematical reasoning ability. While there is still a slight gap with DeepSeek's distillation model, its performance is already very close to world-leading standards. FoxBrain's development – from data collection, cleaning and augmentation, to Continual Pre-Training, Supervised Finetuning, RLAIF, and Adaptive Reasoning Reflection – was accomplished step by step through independent research, ultimately achieving benefits approaching world-class AI models despite limited computational resources. This large language model research demonstrates that Taiwan's technology talent can compete with international counterparts in the AI model field. Although FoxBrain was originally designed for internal group applications, in the future, the Group will continue to collaborate with technology partners to expand FoxBrain's applications, share its open-source information, and promote AI in manufacturing, supply chain management, and intelligent decision-making. During model training, NVIDIA provided support through the Taipei-1 Supercomputer and technical consultation, enabling Hon Hai Research Institute to successfully complete the model pre-training with NVIDIA NeMo. FoxBrain will also become an important engine to drive the upgrade of Foxconn’s three major platforms: Smart Manufacturing. Smart EV. Smart City. The results of FoxBrain is scheduled to be shared for the first time at a major conference during NVIDIA GTC 2025 Session Talk “From Open Source to Frontier AI: Build, Customize, and Extend Foundation Models” on March 20.   About Hon Hai Research Institute The institute has five research centers. Each center has an average of 40 high technology R&D professionals, all of whom are focused on the research and development of new technologies, the strengthening of Foxconn’s technology and product innovation pipeline, efforts to support the Group’s transformation from "brawn" to "brains", and the enhancement of the competitiveness of Foxconn’s "3+3" strategy.