Artificial intelligence has become the strongest growth core engine of this round of semiconductor industry.NVIDIA's performance has exploded due to its exclusive dividends from AI computing chips. At the same time, a number of companies covering the entire chain of wafer foundry, special-purpose chips, and component distribution have simultaneously taken advantage of the AI trend to grow rapidly, forming a complete AI industry growth curve across upstream and downstream.

Nvidia announced impressive financial results for the second fiscal quarter of fiscal year 2025, with single-quarter revenue reaching$30.04 billion, a year-on-year increase of 122% and a month-on-month increase of 15%. The gross profit margin is as high as 75.1%, leading the industry in profitability.As Blackwell's new high-end computing chips are confirmed to be mass-produced and shipped in the fourth quarter, concerns about delays are eliminated, which will further enable AI server expansion, accelerate the commercialization of large models, and continue to consolidate NVIDIA's monopoly in the global high-end computing market.

TSMC’s Q2 revenue was approximately US$20.8 billion, a year-on-year increase of 40.1% and a month-on-month increase of 13.6%; net profit was approximately US$7.66 billion, both year-on-year and month-on-month, higher, and gross profit margins were stable and high.In terms of process structure, the three advanced processes of 3nm, 5nm, and 7nm accounted for 67% of total revenue; high-performance computing HPC revenue accounted for 52%, becoming the largest source of revenue, and AI demand has significantly driven demand.With core AI customers such as NVIDIA, AMD, Qualcomm, MediaTek, and Apple in hand, as AI PCs and end-side AI mobile phones accelerate the penetration, the demand for advanced process wafer production continues to expand.At the same time, TSMC officially proposedFoundry 2.0Strategy, integrate advanced packaging and testing into the business system, rely on CoWoS packaging technology to deeply integrate AI chip manufacturing; cross-border collaboration with SK Hynix for HBM4 basic die packaging, enter the high-bandwidth memory track, move from single foundry to AI full-process services, and strengthen the core voice of the industry chain.Looking forward to Q3, performance will maintain a high growth trend, 2nm process and advanced packaging will increase volume, and continue to undertake the long-term incremental demand for AI.

As the supplier of the underlying core chips for AI servers, Broadcom is an irreplaceable and indispensable role in the construction of computing power clusters.Revenue in the second quarter of fiscal year 2024 was US$12.487 billion, a year-on-year increase of 43%; of which AI-related products generated single-quarter revenue of US$3.1 billion,A surge of 280% year-on-year, the growth potential is strong.The product matrix covers Ethernet switching chips, SerDes communication chips, XPU/TPU computing chips, power management, radio frequency and optocoupler devices, providing one-stop hardware solutions for AI servers.It is highly complementary to NVIDIA: NVIDIA provides core GPU computing power, and Broadcom is responsible for network transmission, interconnection architecture and hardware supporting, jointly supporting the construction of large-scale AI data centers and fully benefiting from the continued expansion of global AI servers.

toDalianda, Wen YeAs a core distributor in the Asia-Pacific region, relying on the advantages of regional server and PC industry clusters, Q2 performance has increased significantly simultaneously.Benefiting from the rise in AI server shipments, the moderate recovery of the PC market, and the scale expansion brought about by Wenye's acquisition of Future Electronics, the orders and revenue of the two major distribution leaders have increased simultaneously.Industry predictions are unanimously optimistic about the market situation in the second half of the year. Q3 is expected to usher in the peak of shipments throughout the year, with both volume and price rising.With the gradual mass production of AI PCs and end-side AI processors, there will be a massive demand for components in the future. Distributors will rely on material reserves, solution support, and supply and demand allocation capabilities to continue to undertake the downstream increments of the AI industry chain and open up the key links between original factories and terminals.

AI has penetrated the entire industry chain of computing power, manufacturing, network, terminals, and distribution, and has become the main axis of the long-term development of the semiconductor industry.In addition to NVIDIA, companies in advanced foundries, high-speed interconnect chips, storage packaging, component distribution and other segmented tracks have all benefited deeply.In the second half of the year, with the large-scale implementation of new-generation AI chips, AI PCs, and large-scale model applications, the prosperity of the entire industry chain will continue to rise, and the dividends of AI trends will further spread to more subdivided IC fields.