The AI PC marks a return to the consumer market after a decade, heralding the most significant innovation in the PC industry in 40 years.
NVIDIA and Microsoft have officially launched their first self-developed Arm-architecture PC processor, the N1X, which utilizes TSMC's 3nm process and integrates a Blackwell GPU with up to 200 TOPS of local AI computing power. Jensen Huang has stated its importance is "comparable to the reinvention of the mobile phone into today's smartphone." This represents the first complete redesign of the PC product line in four decades, with OEMs like Lenovo, Asus, and HP fully on board. The first batch of AI PCs equipped with the N1X chip is set to launch in the fall, marking a milestone inflection point for AI deployment at the edge.
On the software front, the IGV index surged 21% in a single month, disproving the "AI disruption" narrative as earnings provided validation. The rollout of enterprise agents in the second half of the year is emerging as a core catalyst.
The U.S. stock IGV software index gained over 21% in May, its best monthly performance since October 2001. AI infrastructure companies like Snowflake and MongoDB reported earnings that exceeded expectations across the board, prompting the market to reassess the value proposition of software firms in the AI era. The SaaS sector had previously suffered an excessive correction of about 30% in Q1 due to fears of "generative AI disruption." As the earnings season validated performance, these overly pessimistic expectations are being corrected. The accelerated adoption of enterprise agents is anticipated in the second half of 2026, with databases, public cloud, and foundational software representing the most certain directions for capital recovery and valuation re-rating.
The convergence of these two trends suggests that a dual strategy of investing in both computing power and applications is the optimal approach for the next phase of AI investment. AI PCs are generating new demand for edge computing power, while the software recovery is initiating a valuation re-rating on the application side, with each reinforcing the other. For general investors, there is no need to choose exclusively between "betting on hardware" or "betting on software." The E Fund Global Artificial Intelligence ETF (03489) provides exposure to leading computing power and application stocks across Hong Kong and U.S. markets, spanning edge hardware, AI infrastructure, and commercialized use cases, making it a rare tool for a one-stop investment in the dual themes of AI "computing power + application" at this stage.