NVIDIA Corporation ignites AI computing power revolution, Arm CEO reveals: In-house chip business may achieve $15 billion goal ahead of schedule.

Driven by the explosive growth of the AI industry, Arm (ARM.US) CEO revealed that the company's self-developed chip business may achieve its $15 billion sales target ahead of schedule. The confidence in this optimistic expectation comes from the two "deep water bombs" dropped by NVIDIA Corporation (NVDA.US) CEO Huang Renxun at the Taipei Computer Show (Computex) yesterday, which directly ignited Wall Street's frenzy for Arm on Monday, leading to a sharp rise in its stock price. First, in the cloud data center field, Huang Renxun announced in a high-profile manner that its next-generation "Vera Rubin" next-generation AI platform has entered full-scale production. This platform integrates the new Rubin GPU with NVIDIA Corporation's self-developed "Vera CPU" based on the Arm architecture in depth. Huang Renxun explicitly mentioned on site that OpenAI, Anthropic, SpaceX, and Oracle Corporation (Oracle) will be the first core supporters of this powerful Arm architecture chip. Second, on the consumer side, NVIDIA Corporation broke tradition by joining forces with Microsoft Corporation and MediaTek to officially release the first consumer-grade Windows PC based on the Arm architecture - the "super chip" RTX Spark, declaring to "reinvent the PC" with the Arm ecosystem and fully enter the traditional x86 PC domain dominated by Intel and AMD. NVIDIA Corporation's comprehensive endorsement of Arm in the "cloud" and "edge" has completely opened up the imagination space for Arm in the high-end AI computing power and trillion-edge intelligent body (AI Agent) market. Following this industry earthquake, Arm CEO Ren Haas stated in an interview on Tuesday that he was "extremely confident" in achieving the revenue target by the end of this decade. However, given the explosive industry revolution at present, the massive purchasing demand is showing a polarized eruption. Haas revealed during the exhibition that there were signs indicating that the company might achieve the performance milestone ahead of schedule. As evidence, Haas also announced in his keynote speech at Computex on Tuesday that tech giants such as ByteDance and Oracle Corporation have officially adopted Arm's new self-developed AI data center CPU. Haas frankly said, "I hope the target can be achieved ahead of schedule, as the market demand far exceeds our previous expectations, and the current order visibility is extremely good." Arm announced in March this year that it will for the first time develop and sell chips on its own, marking a significant strategic transformation from its traditional model of relying on semiconductor manufacturers to license IP technology. As part of this initiative, the company controlled by SoftBank provided aggressive performance guidance for the coming years and expects the revenue from the new chip business to eventually surpass its traditional main business of IP licensing. Meta (META.US) will be the first core customer for Arm's new self-developed chip named AGI CPU. According to the British company, the processor will have up to 136 computing cores (a measure of computing power), consume 300 watts, and will be manufactured by Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR (TSM.US). Haas reiterated on Tuesday that the decision to develop this component in-house was based on the actual demands of downstream customers. This central processing unit (CPU) is designed to work in collaboration with AI acceleration cards launched by companies such as NVIDIA Corporation, performing tasks such as server-to-server computing power scheduling, data preprocessing, and providing interactive responses to AI queries for users during actual operation.