Summary of J.P. Morgan's Semiconductor Research Call Industry Overview - **Industry Focus**: Semiconductors and Data Center Capital Expenditure (Capex) - **Key Companies Discussed**: NVIDIA (NVDA), Advanced Micro Devices (AMD), Broadcom (AVGO), Marvell Technology (MRVL), Micron Technology (MU) Key Points Data Center Capex Growth - Data center capital expenditure is projected to grow by **65%+ year-over-year (Y/Y) in 2025**, an increase from the previous estimate of **55%** [1] - This growth translates to an **incremental spend of over $115 billion** in 2025 compared to 2024, up from the earlier estimate of **$75-80 billion** [1] - The expected **incremental AI-related revenue** for NVDA, AMD, AVGO, and MRVL in 2025 is projected to be **$85 billion+** [1] 2026 Projections - Data center capex is now expected to track to **50%+ growth in 2026**, revised from **30%** [1] - This implies an **incremental spend of over $150 billion** in 2026, with potential upside to **$175 billion+** if demand continues to surge [1] - The strong demand for AI compute is anticipated to push 2026 data center capex growth into the **60%+ Y/Y range** [1] Company-Specific Insights - **NVIDIA (NVDA)**: Recently announced agreements with OpenAI and Anthropic contribute to a **$500 billion+ revenue backlog** through the end of 2026 [1] - **Broadcom (AVGO)**: Market misinterpretation of management's comments on a **$73 billion backlog** due to lead times [1] - **Marvell Technology (MRVL)**: Likely upside from AWS orders for Tranium3 [1] - **Advanced Micro Devices (AMD)**: Anticipated announcements regarding significant customer engagements in the next 6-9 months [1] Market Dynamics - The spending environment is expected to support a **40-50% compound annual growth rate (CAGR)** for the AI accelerator total addressable market (TAM) off a **$200 billion base in 2025** [4] - Strong AI server spending is projected to benefit companies like NVDA, AVGO, MRVL, AMD, and Micron [4] Future Visibility - While visibility into data center capex beyond 2026 is limited, current expectations suggest a **~10% Y/Y growth** in 2027 if spending remains flat [1] Additional Insights - The report emphasizes the importance of not solely focusing on traditional hyperscalers, as significant spending is also occurring with neoclouds and sovereign AI projects [4] - The analysis indicates that the semiconductor sector is well-positioned to capitalize on the ongoing AI demand, with several companies highlighted as top picks in the AI/cloud space [4] Conclusion - The semiconductor industry, particularly in the context of data center capex, is poised for significant growth driven by AI demand, with key players like NVDA, AMD, AVGO, and MRVL expected to benefit substantially from this trend [1][4]

Summary of J.P. Morgan's Semiconductor Research Call Industry Overview - The focus is on the **semiconductor industry**, particularly in relation to **datacenter capital expenditures (capex)** and the growth of **artificial intelligence (AI)** technologies. Key Points Datacenter Capex Trends - **Long-term elevated spending** in cloud and hyperscale datacenter infrastructure is expected, indicating significant upside to future capex expectations [1] - Growth forecasts for 2025 and 2026 have been **materially increased**, supported by guidance from major U.S. cloud service providers [1][4] - **Microsoft**, **Alphabet**, and **Meta** have all raised their capex guidance for 2025, signaling strong spending momentum into 2026 [1] Specific Company Guidance - **Meta** raised its 2025 capex guidance midpoint to **$71 billion** (up **81% year-over-year**), from a previous estimate of **$69 billion** (up **76% year-over-year**) [1][4] - **Google** increased its 2025 capex guidance to **$92 billion** (up **75% year-over-year**), from **$85 billion** (up **60%+ year-over-year**) [4] - **Microsoft** reported capex of approximately **$35 billion** for the first quarter of FY26, exceeding its prior estimate of **$30 billion+** [4] AI Infrastructure Spending - The ongoing demand for AI is driving a **multi-year growth runway** for companies involved in AI infrastructure, particularly for **NVIDIA (NVDA)**, **Broadcom (AVGO)**, **Marvell (MRVL)**, **AMD**, and **Micron (MU)** [1][4] - Companies are investing heavily in AI-related technologies, with **NVIDIA** expected to benefit significantly from datacenter GPU demand [4] Market Dynamics - The **cloud backlog** for Google has grown to **$155 billion**, indicating strong momentum in cloud services [4] - The overall sentiment from the earnings reports of U.S. cloud service providers supports a **medium-term positive outlook** for capex growth in 2026 and 2027 [1] Investment Opportunities - Companies with leverage to AI infrastructure spending are expected to experience strong tailwinds, including: - **NVIDIA** (AI compute acceleration processors) - **Broadcom** (Google TPU ASIC XPU, Meta MTIA ASIC XPUs) - **Marvell** (Trainium ASIC XPUs) - **AMD** (second-largest merchant GPU supplier) - **Micron** (HBM memory, eSSD) - **Western Digital** (HDD) - **Astera Labs** (networking connectivity) [1][4] Additional Insights - The aggressive investment strategies of cloud providers are aimed at supporting the development of next-generation AI models and enhancing core business operations [1] - The anticipated capex growth in 2026 is expected to be notably larger than in 2025, reflecting a robust demand for AI capabilities [1][4]