Asia Pacific Equity Research 19 November 2024 J P M O R G A N Asian Tech Key takeaways (and transcript) from expert call on Frontier LLMs and AI use cases | --- | --- | |------------------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------| | | | | | Technology | | | Gokul Hariharan AC | | | (852) 2800-8564 gokul.hariharan@jpmo ...

Key Points Industry/Company 1. **Focus on AI Smart Marketing**: The discussion centers around AI smart marketing, specifically highlighting the performance of two companies, Huili Technology and Apple Lovely, in their Q3 earnings. 2. **Huili Technology**: The company reported revenue of $1.05 billion for the first three quarters of 2024, a 36% increase year-over-year. The company's profit was $19.16 million, a 37.4% increase year-over-year. 3. **Apple Lovely**: The company's Q3 revenue was $416 million, a 54.6% increase year-over-year and a 23.6% increase sequentially. Core Views and Arguments 1. **Programmatic Advertising Platform Growth**: Huili Technology's programmatic advertising platform, MintGrow, recorded revenue of $402.5 million, a 57.6% increase year-over-year and a 25.7% increase sequentially. 2. **Smart Bidding Product Success**: The company's smart bidding product, based on machine learning, has been well-received in the market, with its revenue contributing over 60% of MintGrow's total revenue in the first half of 2024. 3. **Expansion into Non-Gaming Categories**: The company has expanded into non-gaming categories such as e-commerce and social media, with e-commerce becoming a significant driver of growth, recording revenue of $108 million in Q3 2024, a 98.5% increase year-over-year. 4. **AI Advertising Engine**: The company's AI advertising engine, Exon 2.0, has been instrumental in driving growth, helping advertisers improve their return on investment in the advertising market. Other Important Content 1. **Focus on Ad Technology and Gaming**: The research focuses on companies involved in ad technology and gaming, with a particular emphasis on Huili Technology's strong performance. 2. **Recommendations**: The analyst recommends focusing on companies like Yidu Tianxia, Blue Focus, Shenguang Yili, and Tianai She, highlighting the potential for growth in the advertising technology and gaming sectors. 3. **AI Application**: The application of AI in advertising technology is expected to bring incremental business opportunities for companies in the sector.

Summary of AI Applications in Cybersecurity Industry Overview - The discussion focuses on the application of AI in the cybersecurity industry, highlighting its current state, future prospects, and the impact of government policies on the sector [1][2]. Key Points and Arguments 1. **Market Size and Growth**: - The Chinese cybersecurity market reached 94.5 billion in 2023 and is projected to grow to 194.2 billion by 2027 [2]. - The AI market in China was valued at 532 billion in 2023, with expectations to reach 1,530 billion by 2027 [2]. 2. **AI's Dual Role in Cybersecurity**: - AI can be utilized by attackers for malicious activities such as deepfake creation, traffic manipulation, and password cracking [2]. - Conversely, AI enhances cybersecurity defenses through fraud detection, anomaly detection, vulnerability management, and threat hunting [3]. 3. **Cost and Efficiency Improvements**: - The average cost of data breaches using AI-based cybersecurity solutions is 2.65 million, compared to 4.45 million without AI [3]. - The time to identify threats has decreased from 277 days to 177 days with AI applications [3][4]. 4. **AI Applications in Cybersecurity**: - Key applications include automated penetration testing, security operations, anomaly behavior analysis, and data visualization [4]. - Technologies employed include automated threat analysis, deep learning, and natural language processing [4]. 5. **Development of Security Models**: - The industry is moving towards the use of large AI models for cybersecurity, which analyze vast amounts of data to identify potential threats [5][6]. - Major players in the market include Qi An Xin, 360 Security, and Yaxin, which are developing large security models to enhance threat detection capabilities [6]. 6. **Current Industry Stage**: - The cybersecurity industry is currently in the second stage of development, focusing on AI as an auxiliary tool for traditional cybersecurity products [6][7]. 7. **Future Trends**: - The future of AI in cybersecurity includes automated threat analysis, adaptive security systems, and AI countering AI strategies [8]. - The ability to handle large volumes of alerts in real-time will be crucial for improving cybersecurity efficiency [8]. Important but Overlooked Content 1. **Risks and Challenges**: - Concerns regarding data security and privacy during the training and deployment of AI models, including the handling of sensitive data [9]. - The need for transparency and explainability in AI decision-making processes, as current models often operate as "black boxes" [9][10]. 2. **Vulnerabilities of AI Models**: - Potential vulnerabilities in foundational models and the security issues that may arise during their deployment [10]. 3. **Regulatory Considerations**: - The importance of compliance with regulations regarding data protection and the implications of AI-generated content [9]. This summary encapsulates the key insights from the discussion on AI applications in cybersecurity, emphasizing market dynamics, technological advancements, and the associated risks and challenges.

Key Points Industry/Company Involved - **Industry**: AI and software solutions - **Companies**: Salesforce (CRM), SAP (SAC), ServiceNow (NOW), SEED, Palo Alto Networks (PANW), Dolingo (DL), Adobe (ADBE), Freshworks (FRWS) Core Views and Arguments - **Trend 1**: Companies are seeing significant revenue growth from AI applications, particularly in cloud ERP suites and customer relationship management (CRM) systems [doc id='2']. - **Trend 2**: Software companies are continuing to invest heavily in AI, with Salesforce's AI Agent and ServiceNow's AI platform as examples [doc id='5']. - **Trend 3**: Companies are optimistic about the future of AI, with ServiceNow and Adobe providing strong revenue and profit growth outlooks [doc id='6', doc id='14']. - **Trend 4**: AI is driving growth in various sectors, including cloud services, CRM, and advertising [doc id='3', doc id='14', doc id='15']. Other Important Content - **SAP (SAC)**: Achieved revenue of 8.47 billion euros, up 99% year-over-year, driven by strong performance in cloud ERP suites and AI applications [doc id='3']. - **ServiceNow (NOW)**: Revenue increased 27.61% year-over-year, driven by strong growth in its AI platform and AI Agent [doc id='5', doc id='6']. - **SEED**: Revenue grew 7.2% year-over-year, driven by growth in its software development and collaboration tools [doc id='7']. - **Palo Alto Networks (PANW)**: Revenue grew 32% year-over-year, driven by strong growth in its AI Platform and Foundry products [doc id='8']. - **Dolingo (DL)**: Revenue grew 49% year-over-year, driven by strong growth in its AI-powered language learning platform [doc id='11']. - **Adobe (ADBE)**: Revenue grew 65.6% year-over-year, driven by strong growth in its AI-powered advertising solutions [doc id='14']. - **Freshworks (FRWS)**: Revenue grew 172.73% year-over-year, driven by strong growth in its AI-powered customer service solutions [doc id='15']. Recommendations - **Focus on AI and software solutions companies**: Salesforce, SAP, ServiceNow, SEED, Palo Alto Networks, Dolingo, Adobe, and Freshworks are recommended as leading companies in the AI and software solutions industry [doc id='17'].

然后说还在聊就是国内的模型模型会会会会会会学生开始有些变化了然后结果没想到现在全球的那个这次应用体上空拍一套这样的数据也起来了所以今天非常的高兴就是那个老朋友欢聚一趟然后要不我们现在先就是先聊一聊就是那个在你这边看你做数据网也是做了快两年啊就是那个呃就专门做AI这一块就是就是我不说我没看到我的问题啊就是为什么突然之间最近就最近一个月从十月我看从九月份那个榜开始九月十月然后很多就是那个 那我们从那个长得最快的开始吧那长得最快的应该是海螺吧就是那个海螺做视频生存的那家公司长得最快的他们视频生存的这个现在的这个步伐就是一代有一个是一个需求 比如说这个模型,无论哪一枝,它的数据表现都是这个趋势,如果有一枝新的新的新的新的新的新的新的新的新的新的新的新的 你让用户去玩这个视频,玩了以后呢,会把由这个新的模型生成的效果发射到每一顿创作,传播完了以后带来讯息。这就是新疆的这个,把视频,把文学视频这个,这个方向,他们的数据表现出来的特点。所以像比如说那个, 法轮功为什么长得这么快?也是因为他们发了一些新的科学,包括皮卡也是长得比较快的,相信也是长得比较快的他也是做了一个粘粘粘的那个什么叫粘粘粘吧,粘粘粘把那个绳杖给缠了,缠根 ...

Summary of Conference Call Notes Industry Overview - The focus of the conference call was on the application of artificial intelligence (AI) and its acceleration in product development, particularly in the Chinese market. The overall performance of the Chinese tech sector has improved significantly since September, with the industry median return reaching 14% [1][2]. Key Points and Arguments 1. **Market Recovery and Future Focus** - The recovery in the tech sector is attributed to various factors, including local government debt management, the need for technological support following Trump's election, and advancements in commercial aerospace and AI in China. The next phase will focus on fundamental turning points and government policy support [1]. 2. **Importance of State-Owned Enterprises (SOEs) Restructuring** - The restructuring of SOEs is highlighted as a critical area for improvement, particularly in addressing issues related to product chains and operational efficiency. Companies like China Television and China Electronics Technology are expected to play significant roles in this restructuring [2]. 3. **AI Performance and Trends** - The performance of AI-related companies overseas has been strong, with significant revenue growth reported. For instance, APK Loading's third-quarter revenue reached 1.2 billion, a 40% year-over-year increase, exceeding market expectations [4]. 4. **Domestic AI Developments** - In China, AI applications are also seeing growth, with platforms like ZuoBao and Baidu's WenxiaoYun surpassing 10 million users. ZuoBao's user base has exceeded 100 million, indicating a strong market presence [5][6]. 5. **AR Market Insights** - The AR market in China is experiencing rapid growth, with significant user engagement reported. For example, the 360 AR search platform has become the leading platform in China, achieving 287 million monthly searches in October, reflecting a 10.56% growth over the previous months [6][7]. 6. **Investment Recommendations** - The call emphasized the importance of focusing on AI and AR sectors, suggesting that investors should pay attention to leading companies in these fields. The recommendation includes monitoring developments in large models and the performance of key players in the AI and AR markets [7]. Additional Important Content - The call also mentioned the upcoming developments from Huawei, particularly in the automotive sector, which could present further investment opportunities [3]. - The overall sentiment is optimistic regarding the growth of AI and AR technologies, with a strong emphasis on the need for continuous monitoring of market trends and company performances [7].

Key Points Industry and Company Involved * **Industry**: AI Smart Glasses * **Company**: Baidu, NetEase Youdao,昆仑万维, Meta, Ray-Ban, Huawei, Xiaomi, OPPO, Vivo, and various startups Core Views and Arguments 1. **Rapid Development of AI Smart Glasses**: AI smart glasses are experiencing rapid development with applications in education, entertainment, and daily life. * **Application Scenarios**: Education (e.g., Baidu AI Glasses with NetEase Youdao for foreign language teaching), entertainment, and daily life (e.g., real-time translation, smart photography). * **Market Penetration**: AI search is the most popular AI application with a high market share on PC and mobile platforms, and it is expected to become one of the most widespread application scenarios for AI smart glasses. * **Hardware Development**: Meta and Ray-Ban's smart glasses have exceeded one million in sales, and domestic supply chains are mature, meeting demand. * **Challenges and Opportunities**: Weight, battery life, and image accuracy are current limitations, but technological progress is expected to improve these issues. Future development directions include medium to heavy entertainment content like AI movies and games. 2. **AI Search in the Market**: AI search has become one of the most popular AI applications with a high market share on PC and mobile platforms. It is expected to become one of the most widespread application scenarios for AI smart glasses. * **Market Share**: In October, most of the top 10 websites in China had AI search or smart assistant features. Globally, over 70% of the top 20 websites and active users were related to AI search or smart assistants. * **Competitive Landscape**: Major manufacturers are launching various AI search products and competing fiercely. 3. **Kunlun Wanwei's Layout**: Kunlun Wanwei is continuously iterating its Tian Gong AI search, exploring advertising monetization models, and investing in large models and AI search. 4. **Hardware and Supply Chain Development**: Since Meta and Ray-Ban launched the first smart glasses last September, sales have exceeded one million, and it is expected to reach nearly two million this year. The domestic supply chain is mature, meeting demand. 5. **Limitations and Future Directions**: Current limitations include weight, battery life, and image accuracy. Future development directions include medium to heavy entertainment content like AI movies and games. 6. **Short-term Application Scenarios**: Short-term applications are concentrated in tool attributes and voice/text interaction, such as real-time translation and smart photography. Advanced education, AI e-commerce, and sports scenes are worth paying attention to. 7. **Advantages of AI Smart Glasses**: Voice interaction convenience, direct visual perception, long-term wearability, and strong scalability. 8. **Major Players**: Internet companies (e.g., Meta, Baidu) and traditional AR/VR manufacturers (e.g., Lark, Look at X Real) are the main players in the market. 9. **AR Device Market**: The global AR device market grew by 63% last year, reaching about 500,000 units. The domestic market grew by over 150%. 10. **Smart Glasses as a Development Direction**: Smart glasses have a large market size and significant potential for upgrading under AI empowerment. The market is dominated by internet companies and traditional AR/VR manufacturers. 11. **Mobile Phone Manufacturers' Layout**: Huawei, Xiaomi, OPPO, and Vivo are actively researching smart glasses. They have advantages in camera and audio hardware, interconnectivity, supply chain, and sales channels. 12. **Emerging Brands and White-Label Manufacturers**: Emerging brands and white-label manufacturers are utilizing China's mature supply chain to produce smart glasses, similar to the TWS earphone market. 13. **Core Components and Costs**: The main components include the main chip, storage, camera, microphone, battery, and other core components. The main chip is the most expensive part, followed by the camera. 14. **Future Trends and Companies to Watch**: The Chinese mainland has sufficient preparation for the hardware supply chain. Companies like衡山科技, 中科蓝讯, 矩形科技, 歌尔股份, 龙旗科技, 华勤技术, 嘉禾智能医疗信息, 国光电器, 漫步者, and 博士眼镜 will play important roles in the future.

M Update China – Data Centers | Asia Pacific November 15, 2024 01:16 PM GMT Record hyperscalers' capex driven by AI Key Takeaways Alibaba and Tencent combined spent Rmb31.7bn operating capex in 3Q24, +196% YoY/66% QoQ, reaching the highest level ever seen. Both noted strong momentum for AI – AI computing is a teens (%) share of Tencent's IaaS revenue and Alicloud continues to show triple-digit AI-related growth. We think this reads positively for China's data center companies – especially GDS. Alibaba and T ...

15 November 2024 | 4:25AM GMT to LT semiconductor demand. As such, ASML reiterated its guidance for 2030 and expects revenues of €44-60bn, which at the mid-point/upper end implies a CAGR of 10%/13% across 2025-30. Key takeaways include: 1) Reiterated guide for 2030 points towards continued strength of LT demand drivers, with the stock currently trading at a modest 2030 PE multiple suggesting scope for upside, 2) ASML reiterated its 2030 semis TAM outlook of >$1trn, with a weaker mature node / NAND memory fo ...

内呢也是被称为叫AR四小龙吧然后四小龙之一啊就是我们可能而且是在在这个自家中可能是做的比较有 就是比较特立独行的一件就是如果说去形容的话因为我们的产品就一开始就做这种无限的一体式的不同于别家的这种比如说是DB方案的这种产品吧所以我们一直在产品方向上一直是被行业内外吧都会认为比较有辨识度比较有 所以我们现在的这个产品的思路也是包括发这个新品的思路其实也是延续我们这样的一个一贯的一个价值主张 现在我们是在这个月月底11月29号会在成都开我们的发布会去发布我们的两款新品今年发布了两款新品然后这两款新品其实有一款也是会延续我们整体的一个思路就是我们认为 AR眼镜或者说智能眼镜吧AI眼镜这个无所谓啊因为只是一个名字它是一个应该是手机的下一代的移动终端这个其实在这个小扎发布他这个前一阵发布他的那个产品的时候其实也提到这个事情了 跟我们的想法是一样的就是它其实是一个代替的是手机平板这一类的一个智能终端我们常用的这种智能终端的这个下一代的一个一个一个一个产品的一个形式对然后呢它的这个功能呢也是聚焦在我们比如说是 你可以类比手机Pad这种智能产品的它的这种使用场景对 这个是我们的一个产品思路将会延续只不过我们会把它做得更强大然 ...