Summary of Conference Call on Greater China Semiconductors Industry Overview - **Industry**: Greater China Semiconductors - **Focus**: Semiconductor foundry market, particularly SMIC and Hua Hong Key Points Demand and Growth Projections - **AI GPU Demand**: The proliferation of AI applications in China, supported by government policies, is expected to significantly boost domestic leading-edge foundry demand over the next two years [1][2] - **Revenue Forecasts**: Updated revenue forecasts for China AI GPU are Rmb113 billion for 2026 and Rmb180 billion for 2027, reflecting a 62% CAGR from 2024 to 2027 [2][20] Supply Dynamics - **SMIC's Expansion**: SMIC is expanding its leading-edge fab capacity, which is anticipated to alleviate equipment bottlenecks. The forecast includes a total capacity of 22kwpm for 7nm and under by 2025, increasing to 42kwpm by 2026 [1][10] - **Local Supply**: Local suppliers like Naura and AMEC are gradually replacing previously bottlenecked tools, enhancing China's ability to produce AI GPU chips [1][10] Competitive Landscape - **Mature Node Demand**: Demand for mature nodes remains weak, with oversupply in capacity. However, there is still demand from smartphone SoCs and autonomous driving semiconductors that could offset potential GPU demand weaknesses [3] - **Hua Hong's Position**: Despite raising wafer prices, Hua Hong's profitability appears weaker compared to SMIC and UMC, with an EBITDA margin of 30% in Q2 2025 compared to SMIC's 47% and UMC's 41% [3][9] Stock Recommendations - **SMIC**: Upgraded to Overweight (OW) with a price target of HK$80, reflecting strong domestic AI demand and improved gross margins [4][8] - **Hua Hong**: Downgraded to Underweight (UW) due to concerns over the sustainability of its mature node business and inventory build-up [4][9] Strategic Insights - **Self-Sufficiency in Semiconductors**: China's semiconductor self-sufficiency ratio is projected to rise to 30% by 2027, driven by advancements in local production capabilities and government support [52][60] - **AI Localization**: SMIC is positioned to benefit from strong domestic AI localization demand, supported by government initiatives and the need for advanced node manufacturing [89] Risks and Considerations - **Market Risks**: Potential risks include the possibility of local CSPs purchasing more AI chips from foreign vendors, which could impact SMIC's utilization rates and market share [94] - **Performance Variability**: The performance of local AI chips, particularly from Huawei, may face challenges compared to global competitors like NVIDIA [36][90] Additional Insights - **Huawei's Developments**: Huawei is advancing its AI chip capabilities with new product launches and improvements in interconnect bandwidth, which may enhance its competitive position in the market [35][37][38] This summary encapsulates the critical insights from the conference call regarding the semiconductor industry in Greater China, focusing on demand forecasts, supply dynamics, competitive positioning, and strategic recommendations for key players like SMIC and Hua Hong.

Group 1 - The Federal Reserve is widely expected to lower interest rates this week, contributing to a positive performance in the US stock market, with the S&P 500 rising by 0.47%, the Dow Jones by 0.11%, and the Nasdaq by 0.94% [1] - Strong performance in technology stocks, with Tesla experiencing a three-day increase of over 7%, and Google nearing a market capitalization of $3 trillion with a nearly 4.5% rise [2] - The Nasdaq Golden Dragon China Index rose by 0.87%, approaching its January 2022 high, with notable gains in Chinese concept stocks such as Mindong Hong Group up over 35% and Xiaoma Zhixing up about 11% [2] Group 2 - US Treasury yields fell, with the 10-year yield dropping by 3.45 basis points, while the US dollar index decreased by 0.35%, falling below the support level of 97 [3] - A weaker dollar contributed to gold prices rising by over 1.2%, approaching $3,700, marking a new high [4] Group 3 - The US and China have reached a basic framework consensus to properly address the TikTok issue during trade talks [5][10] - Nvidia is under further investigation in China for violating antitrust laws [6][12] Group 4 - Elon Musk increased his stake in Tesla by acquiring 2.5687 million shares, valued at approximately $1 billion [8] - The UK and US governments are set to sign the Atlantic Advanced Nuclear Partnership agreement during the upcoming state visit of the US President, aimed at accelerating the construction of nuclear power plants by companies in both countries [9] Group 5 - The Chinese government is promoting the construction of a unified national market, which is crucial for high-quality development and gaining an advantage in international competition [17] - The Guangzhou government has issued a plan for the large-scale application of Vehicle-to-Grid (V2G) technology, aiming for significant milestones in energy management by 2027 [16] Group 6 - Nvidia is pushing suppliers to develop a new micro-channel liquid cooling plate (MLCP) technology, which is 3 to 5 times more expensive than existing cooling solutions [7][18] - Huawei is set to hold a conference to fully open-source its CANN architecture, which is seen as a significant step in changing the AI chip landscape in China [19] Group 7 - Several companies are making significant acquisitions, including New Dazheng planning to acquire at least 51% of Jiaxin Liheng Facility Management [22] - Longpan Technology signed a procurement agreement with CATL for lithium iron phosphate cathode materials, with a total sales amount exceeding 6 billion yuan [24]

Core Viewpoint - Tencent has sufficient AI GPU chips to meet future model training needs and does not require additional purchases of NVIDIA H20 chips, indicating a strategic shift towards non-Western AI chips for inference tasks [1][3][4]. Group 1: AI Chip Inventory and Strategy - Tencent has a strong inventory of GPUs, sufficient for ongoing model upgrades, marking the third consecutive quarter the company has stated it does not need more GPUs [3]. - The company is exploring alternatives to Western GPU manufacturers for inference chips, evaluating various AI accelerators that can be legally imported or sourced within China [4]. - Tencent is enhancing the utilization of existing computing power through software optimizations, aiming to double AI inference efficiency, effectively doubling GPU capacity [4]. Group 2: Regulatory and Market Conditions - NVIDIA's AI chips have not yet received security approval from the Chinese government, leading to caution among domestic tech giants regarding the procurement of NVIDIA H20 and AMD MI308 chips [4]. - The company is adapting to the uncertainty of GPU supply by focusing on revenue growth and expanding in CPU, database, storage, and CDN infrastructure [6]. Group 3: Financial Performance and Investment - Tencent's capital expenditure in Q2 reached 191.1 billion RMB, a 119% year-on-year increase, covering infrastructure, data center expansion, and AI software platform development [6]. - Despite a 30% decrease in capital expenditure from Q1, the company maintains a strong financial position, with Q2 revenue of 1,845 billion RMB, a 15% increase, and net profit of 648 billion RMB, an 11% growth [6].