Market Overview - The main board led the gains with notable increases in stocks such as Shengshi Technology (+10.01%), Dahua Intelligent (+9.94%), and Wuhan Fanggu (+7.39%) [1] - The ChiNext board saw significant growth, particularly in Guangzhi Technology (+19.99%) and Huina Technology (+15.05%) [1] - The Sci-Tech Innovation board was also active, with Weide Information (+13.00%) and Fuchuang Precision (+8.08%) leading the way [1] Semiconductor Industry Developments - Yangzhou Kangying Semiconductor has officially opened with a total investment of 500 million, focusing on SSD and PSSD products [1] - Yangzhou Jinxin Micro's 6-inch chip factory is set to achieve mass production in May, with an expected annual output of 360,000 chips and additional sales revenue of 300 million [1] - Foxconn and HCL Group's joint chip factory project in India has received cabinet approval, with a total investment of approximately 435 million, focusing on display driver chips [1] - NVIDIA's CEO Jensen Huang indicated a strategic reassessment of the Chinese market due to U.S. government restrictions on the Hopper architecture H20 chip exports [1] Company Announcements - Lifan Holdings announced the completion of a share reduction by a major shareholder, reducing holdings by 922,432 shares, now holding 9,665,804 shares (10.48%) [2] - Guangge Technology reported a significant reduction in shares by a major shareholder, decreasing from 7.26% to 5.00% after selling 1,490,323 shares [2] - Sitwei announced a reduction in shares by Brizan Holdings, lowering its stake from 5.82% to 5.00% [2] - Hu塑科技 declared a cash dividend of 2.0 yuan per 10 shares, totaling 12 million, based on a total share capital of 60 million [2] Flash Memory Market Insights - International Flash spot prices have generally increased, with MLC chips seeing a rise of 1.89%, reaching an average price of 5.063 USD for MLC 64Gb 8GBx8 [3] - Applied Materials reported Q2 FY25 revenue of 7.1 billion, a 7% year-over-year increase, slightly below market expectations [3] - The telecom sector is increasingly focusing on data center interconnect technology, with an expected market value growth of 14.3% in 2025, surpassing 40 billion [3]

Financial Performance - Revenues grew to $17.4 million in Q1 2025, up approximately 190% sequentially and 150% year-over-year[6, 8, 23] - Gross margin reached a record 40% due to positive NRE contribution[6, 8] - Cash used in operations decreased to $20.7 million in Q1 2025, compared to $23.0 million in Q1 2024[8, 23, 24] - The company's cash position grew to $85.4 million[8, 23] NRE Payment Plan - The NRE payment plan expanded from approximately $80 million to approximately $95 million, with the majority of payments expected in 2025-2026[5, 8] - The company is targeting $20-50 million in new NRE bookings[20] Manufacturing and Partnerships - A partnership with Fabrinet was announced to launch mass production of the InnovizTwo platform[5] - Fabrinet will act as the manufacturing partner for the InnovizTwo product platform[10] Program Updates and Engagements - The company is accelerating the delivery of LiDARs to VW to support the expansion of autonomous shuttle operations with hundreds of ID Buzz AD vehicles ahead of fleet deployments[5, 14] - There is growing momentum in the relationship with NVIDIA and the Hyperion platform, with progress being made with a major OEM evaluating Hyperion[5, 16] Targets - The company is aiming for revenue between $50-60 million, more than double 2024 levels[20]

Group 1 - The core index, the China Securities Hong Kong and US Listed Global Smart Car Theme Index, reported a value of 5461.69 points, with a recent one-month increase of 8.39%, a three-month decrease of 10.04%, and a year-to-date decline of 9.83% [1] - The index comprises 50 listed companies involved in smart automotive sectors such as perception positioning, decision planning, control execution, smart cockpit, and vehicle networking, reflecting the overall performance of these companies in the Hong Kong and US markets [1] - The index is based on a starting point of 1000.0 points as of December 30, 2016 [1] Group 2 - The top ten holdings of the index include Li Auto-W (9.52%), NVIDIA Corp (9.44%), Xpeng Motors-W (8.84%), Tesla Motors Inc (8.59%), Marvell Technology Group Ltd (6.26%), Rivian Automotive Inc. (5.6%), Lucid Group Inc (3.64%), Aurora Innovation Inc. (3.61%), Baidu Group-SW (3.27%), and Analog Devices Inc (3.19%) [2] - The index's holdings are primarily listed on the Nasdaq Global Select Market (60.44%), followed by the Hong Kong Stock Exchange (29.66%), Nasdaq Capital Market (5.02%), and New York Stock Exchange (4.88%) [2] - In terms of industry distribution, passenger vehicles account for 23.02%, digital media for 3.27%, optical optoelectronics for 3.05%, and automotive parts and tires for 0.32% [2] Group 3 - The index samples are adjusted semi-annually, with adjustments occurring on the next trading day after the second Friday of June and December each year [3] - Weight factors are adjusted in accordance with the sample changes, and generally remain fixed until the next scheduled adjustment [3] - Special circumstances may lead to temporary adjustments, including the removal of samples that are delisted or undergo mergers, acquisitions, or splits [3]

Market Performance - The main board saw significant gains with Wuhan Fanggu rising by 10.04%, Aopu Optoelectronics by 10.01%, and Tianjian Technology by 10.00% [1] - The ChiNext board led with Kela Software increasing by 13.98%, Hanbang High-tech by 6.55%, and Shuo Beid by 5.97% [1] - The Sci-Tech Innovation board was also up, with Aerospace Nanhu gaining 11.11%, Holaiwo by 8.77%, and Guobo Electronics by 5.66% [1] Domestic News - Yangjie Technology has initiated a 1 billion yuan project for SiC automotive-grade power semiconductor module packaging, focusing on IGBT modules and SiC MOSFETs, aiming for import substitution and international competitiveness [1] - Rockchip announced the launch of the new AI vision chip RV1126B, which has passed mass production testing and will begin bulk supply this month, featuring a 4-core Arm Cortex-A53 CPU and a self-developed 3TOPS NPU [1] - TCL Huaxing is supplying Lenovo's Moto Razr 60 series with the latest generation of display materials, achieving a 13% reduction in power consumption compared to the previous generation [1] - Nanchip Technology has launched a vehicle-grade high-speed CAN/CAN FD transceiver SC25042Q, entering the CAN transceiver market with strong anti-electromagnetic interference capabilities [1] Company Announcements - Changguang Huaxin announced it has received government subsidies totaling 10.3997 million yuan, with 8.1775 million yuan related to revenue [2] - Hongsoft Technology is planning a share transfer of 8,023,408 shares, representing 2.00% of its total equity, by HomeRun Capital Management Limited [2] - Dahua Technology has signed a maximum guarantee contract with ICBC for 10 million yuan to provide joint liability guarantees for its subsidiary Zhejiang Huayi Technology [2] - Weston reported a reduction in shareholding by a major shareholder, with a decrease from 7.00% to 5.79% after selling 1,066,500 shares [2] Overseas News - Intel is in discussions with NVIDIA, Microsoft, and Google regarding its 18A process, which may serve as a strong alternative to TSMC's N2 process [3] - OpenAI has agreed to acquire the AI coding tool Windsurf for approximately 3 billion USD, which allows developers to navigate and modify code using AI [3] - Apple's silicon design team is developing new chips for future devices, including its first smart glasses and more powerful Macs [3] - Infineon has released a new generation of 750V silicon carbide MOSFETs for automotive and industrial applications, with a conduction resistance range of 7mΩ to 140mΩ [3]

Summary of Rigetti Computing Conference Call Company Overview - Rigetti Computing is a pioneer in full-stack quantum computing, operating superconducting quantum computers since 2017 through a cloud services platform and selling on-premise quantum computers since 2021 [2][8] - The company developed the first multichip processor in the industry and manufactures superconducting quantum processing units (QPUs) in-house at a dedicated facility [2] Core Industry Insights - Quantum computing is an emerging technology that fundamentally differs from classical computing by using qubits instead of bits, allowing for exponential computing capabilities and lower energy consumption [5][6] - Rigetti focuses on gate-based quantum computing, which is expected to dominate the industry, while acknowledging niche applications for other modalities like annealing [12][14] Technological Advantages - The advantages of superconducting technology include scalability and gate speed, with Rigetti's gate speeds comparable to current CPU and GPU technologies [16][17] - Recent advancements have improved two-qubit gate fidelity to 99.5%, with a target of reaching 99.8% to enable commercial applications [24][25] Roadmap and Future Goals - Rigetti aims to demonstrate a quantum computer with over 100 qubits and maintain high fidelity in 2024, with plans to increase qubit count and fidelity in subsequent years [27] - The company is pursuing a modular chiplet architecture to facilitate scaling, which has shown promising results in initial deployments [28][34] Quantum Networking and Error Correction - Rigetti believes in a hybrid computing model where quantum computers coexist with classical systems, rather than requiring a separate quantum network [41][42] - The company is collaborating with Riverlane to develop real-time error correction codes, which are critical for future quantum computing applications [46][51] Strategic Partnerships - Rigetti has formed a strategic partnership with Quanta Computer, which includes a $35 million investment and a commitment of $250 million over five years for non-quantum hardware development [82][86] - This partnership is aimed at enhancing Rigetti's manufacturing capabilities and reducing R&D costs as the business scales [85] Government Funding and Legislative Updates - The National Quantum Computing Act is pending reauthorization, which is crucial for long-term funding and support for quantum initiatives [90][92] - Rigetti has secured funding from DARPA and the Air Force Research Lab, indicating strong government interest in quantum technologies [93] Conclusion - Rigetti Computing is positioned at the forefront of quantum computing technology, focusing on scalability, fidelity, and strategic partnerships to drive future growth and commercialization [95]

Group 1 - Preferred Networks is recognized as a "true AI" company due to its reliance on deep learning, NLP, and generative models, along with its self-developed models and AI frameworks [1][3][4] - The company has a strong product versatility, offering solutions across various sectors including industrial automation, healthcare, and education, with over 435 global patents [5][6] - Despite its initial ambitions for international expansion, Preferred Networks has reverted to a domestic focus, raising concerns for other Japanese tech firms considering overseas ventures [2][10] Group 2 - Preferred Networks was founded in 2014 and developed the deep learning framework Chainer, which was once positioned alongside TensorFlow and PyTorch [3][11] - The company has shifted its strategy to collaborate with major Japanese corporations like Toyota and Nissan, focusing on customized AI systems rather than pursuing a broader international presence [13][18] - The company has established a new subsidiary, Preferred Elements, aimed at foundational technology development, indicating a potential shift towards a more open approach [14][16] Group 3 - PKSHA Technology, another prominent Japanese AI firm, has shown strong profitability with significant revenue growth, serving various industries including retail and finance [24][25][26] - Unlike Preferred Networks, PKSHA retains ambitions for international collaboration, partnering with companies like Microsoft and Tencent [26] - The early establishment of AI companies in Japan, such as PKSHA and Preferred Networks, was driven by a combination of engineering talent and industry demand for automation [28][30] Group 4 - The Japanese AI industry is characterized by a closed-loop system where startups primarily serve large domestic corporations, limiting their growth potential and innovation [44][45] - The government and large companies emphasize project-based AI solutions, which diminishes the drive for exploratory or innovative AI developments [44][45] - Cultural factors contribute to the lack of ambition for developing universal AI platforms, contrasting with the more aggressive approaches seen in other countries [30][43]

Core Insights - Sify Infinit Spaces Limited has announced that its new AI-ready data center facilities in Chennai and Noida are now part of the NVIDIA DGX-Ready Data Center program and certified for liquid cooling, following the certification of its Mumbai, Rabale facility as the first in India [1][3]. Company Overview - Sify Technologies Limited is recognized as India's leading Digital ICT solutions provider, offering a comprehensive range of services including data center, cloud, networks, security, and digital services [1][4]. - The company has received multiple Golden Peacock awards for Corporate Governance, highlighting its commitment to high standards in business practices [4]. Infrastructure and Capabilities - Sify's data center deployment solution supports up to 200 KW/rack capacity and complies with ASHRAE's 30 years ambient temperature guidelines, ensuring robust performance for AI workloads [2]. - The company operates state-of-the-art data centers and has the largest MPLS network in India, making it a preferred choice for startups, SMEs, and large enterprises [5]. Market Position and Reach - More than 10,000 businesses across various sectors utilize Sify's integrated services of Data Centers, Networks, and Security, enabling seamless operations in over 1,700 cities in India [6]. - Sify also has an international presence in North America, the United Kingdom, and Singapore, expanding its global service capabilities [6].

Summary of Conference Call on ABF Substrate Industry Industry Overview - The ABF (Ajinomoto Build-up Film) substrate market is expected to experience oversupply conditions until the second half of 2026, with a projected 7% oversupply in 2025 [1][11][25]. - The total addressable market (TAM) for ABF substrates has been revised down by 12% and 10% for 2025 and 2026, respectively, due to lower-than-expected AI server GPU substrate shipments and increased pricing pressure [2][22]. Key Insights - **Demand Dynamics**: The demand for AI server substrates has been cut by approximately 50% for 2025, reflecting a significant decrease in AI server rack shipments, which are expected to decline by 35% [2][60]. - **Pricing Pressure**: The average selling price (ASP) for AI GPU substrates has decreased from over US$120 earlier in the year to US$100, indicating a challenging pricing environment [2][60]. - **Impact on Suppliers**: Tier-2 suppliers like Unimicron and Kinsus are expected to face significant challenges due to their exposure to legacy ABF products and weaker-than-expected spillover orders from tier-1 suppliers [2][9][24]. Supplier Performance - **Unimicron**: Downgraded to Neutral from Buy due to weaker AI spillover orders and a slow PC demand outlook. Target price revised from NT$170 to NT$90 [9][10]. - **Kinsus**: Downgraded to Sell from Neutral, with a new target price of NT$63, reflecting less favorable growth in the mid-to-low-end ABF substrate market [9][10]. - **Ibiden**: Maintained Buy rating with a target price of ¥6200, benefiting from its leading position in the high-end substrate market [10][22]. - **Zhen Ding Technology (ZDT)**: Initiated with a Buy rating, expected to benefit from ABF demand expansion from Chinese IC design houses [8]. Future Catalysts - The mid-to-long-term outlook for Taiwan ABF suppliers is expected to improve with the localization of AI IC production in China, projected to grow at a CAGR of 53% from 2025 to 2027 [3][4]. - Adoption of CPO (Chip-on-Package) technology is anticipated to double the consumption of switch IC ABF substrates compared to general 2.5D packaging substrates [3]. Capacity and Utilization - The overall industry capacity is expected to increase by approximately 40% from 2024 to 2027, with a significant portion of new capacity planned before 2024 [29][53]. - Utilization rates for ABF substrate suppliers are projected to improve in 2025, but overall industry utilization will remain below 80% [27][24]. Risks and Challenges - Potential risks include higher electricity costs in Taiwan and the impact of tariffs, which could increase costs as a percentage of revenue significantly [46][45]. - The overall cash margins for ABF substrate suppliers are expected to remain under pressure until at least the second half of 2025, with a potential recovery starting in 2026 [44][59]. Conclusion - The ABF substrate industry is currently facing a challenging environment with oversupply and pricing pressures, particularly affecting tier-2 suppliers. However, long-term growth opportunities exist through technological advancements and increased demand from AI-related applications. The market is expected to stabilize post-2026, with significant growth driven by AI and server upgrades.

Core Insights - Firefly Neuroscience, Inc. has acquired Evoke Neuroscience, Inc., marking a significant step towards developing a proprietary foundation model of the human brain using its FDA-cleared Brain Network Analytics (BNA™) technology [1][2]. Company Overview - Firefly is an AI company focused on improving brain health outcomes for patients with neurological and mental disorders, utilizing its BNA™ technology for diagnostics and treatment monitoring [3][4]. - The company has built a comprehensive database of brain wave tests and secured patent protection over the past 15 years, now targeting pharmaceutical companies and medical practitioners for commercial use [3]. Acquisition Details - The acquisition of Evoke will enhance Firefly's assets, including a more than two-fold increase in proprietary brain scans, a three-fold increase in patents, and a ten-fold increase in commercial sites [2]. - The purchase terms include a total payment of $6 million, split 50% in cash and 50% in Firefly's common stock priced at $3.50 per share, along with a potential $500,000 earn-out for Evoke's investors based on future revenue performance [5]. Technological Synergies - The integration of Evoke's technologies is expected to create significant technological, clinical, and commercial synergies, enhancing Firefly's capabilities in brain health assessment [2]. - Firefly's BNA™ technology utilizes AI and machine learning on a proprietary database of over 17,000 patients, providing comprehensive insights into brain function and improving diagnostic accuracy for mental and cognitive disorders [4]. Market Position - Firefly possesses the largest known proprietary database of over 180,000 standardized EEG/ERP assessment records and has 27 granted patents, positioning the company strongly in the brain health technology market [5].

April 21, 2025 11:41 PM GMT Asia Technology | Asia Pacific M Foundation Tuesday TMT: TSMC, Tariff Impact on Technology, and Japan SPE Join our Technology analysts Shawn Kim, Charlie Chan, Tetsuya Wadaki, and Derrick Yang as they discuss TSMC's takeaways post results, further potential tariff impacts on Technology, and Japan semicap equipment updates. | Morgan Stanley Asia Limited+ | | | --- | --- | | Shawn Kim | | | Equity Analyst | | | Shawn.Kim@morganstanley.com | +852 3963-1005 | | Morgan Stanley Taiwan ...