Core Viewpoint - The company is experiencing significant growth opportunities in edge AI due to advancements in large model intelligence and lightweight technology, prompting hardware companies to enter the market [1] Group 1: Business Strategy and Product Development - The company has adjusted its business strategy since last year to promote edge AI across consumer, enterprise, and industry scenarios [1] - The company plans to launch an AI learning machine equipped with the "YunTianTianShu" multimodal large model by the end of 2024, with over 70,000 units shipped across all channels to date [1] - The company is preparing to launch the AI plush toy "LuKa Doctor" by the end of 2025, aimed at providing digital companionship for children [1] Group 2: Financial Performance - In the first half of the year, the company achieved revenue of 646 million yuan, a year-on-year increase of 123.10%, while narrowing its net loss to 206 million yuan, a 33.51% increase in loss [2] - The increase in revenue is attributed to higher sales in consumer and enterprise scenarios, while the narrowing loss is due to increased revenue and gross margin, with a gross margin of 28.5%, up 12.8 percentage points year-on-year [2] Group 3: Marketing and Sales Expenses - The company's sales expenses have significantly increased, with promotional expenses rising from 6.86 million yuan in mid-2023 to 23.30 million yuan in the first half of 2024, more than doubling [3] - The company is actively engaging in marketing efforts to support its sales strategy, despite investor concerns regarding high sales expense ratios [3] Group 4: Technological Advancements - The company has developed the Deep Edge 10 chip series, a high-performance SoC chip that supports various mainstream models and is being commercialized in robotics and edge gateway applications [3] - The company is also developing the next-generation "brain" chip series, DeepXBot, to enhance perception, cognition, decision-making, and control tasks in humanoid robots [3] - The company ranks among the top three providers of AI inference chip products and services in the industry, and is the second in China for NPU-driven AI inference chips [4] Group 5: Application and Market Expansion - The company has successfully adapted its YunTianTianShu large model training and inference machine for various applications, including government cloud platforms [5] - The company is developing multiple chips to cater to diverse application scenarios, including the DeepVerse chip series designed for large model inference applications [5] - The company is building a "robot brain" computing platform, DeepXBot, integrating high-performance NPU cores to support various algorithms [5] Group 6: Future Developments - The company is working on the fifth-generation Nova500 chip, which features significant technological breakthroughs and supports compatibility with most AI large model architectures [6]