Jensen Huang, NVIDIA's founder and CEO, publicly expressed regret that the company never acquired Arm, the chip design firm whose stock has surged 263% in 2026 as ByteDance and Oracle adopt its new data center CPU. The candid admission came during a lighthearted on-stage conversation with Arm CEO Rene Haas at Computex 2026, revealing a missed opportunity that could have reshaped the semiconductor industry.

Why Did NVIDIA Miss Out on Acquiring Arm?

During their conversation, Huang joked about the timing and circumstances that prevented the deal. When Haas mentioned that "if the two companies merged, we would become the largest company in the world," Huang responded with clear nostalgia, saying he had once tried to work with Arm and attempted cooperation again. "But it's okay. I'm still sad," Huang remarked, acknowledging the strategic opportunity that slipped away.

When Haas

The two executives shared a warm rapport, joking about why the acquisition never happened. When Haas pointed out that Huang "used to be a shareholder, and then you sold your shares," Huang quipped back, "Yes, yes. Oh, I needed cash." The exchange highlighted how different circumstances and business priorities at the time made the acquisition impossible.

How Is Arm Becoming a Powerhouse in AI Infrastructure?

Arm's momentum in the AI era is undeniable. At Computex, Arm CEO Rene Haas announced that ByteDance and Oracle have adopted Arm's self-developed data center CPU chip, called Arm AGI. Last month, Arm doubled its demand forecast for this processor, expecting it to generate $2 billion in revenue during fiscal years 2027 and 2028, with projections reaching $15 billion annually within five years.

The timing of these announcements coincided with NVIDIA's own product launches. On Monday, NVIDIA released the RTX Spark superchip and Vera data center CPU, both based on Arm architecture. That same night, Arm's stock price soared, and by Tuesday's close, it had risen 16%. Year-to-date, Arm's stock has accumulated a 263% increase, reflecting investor confidence in the company's position in the AI infrastructure market.

Huang acknowledged Arm's competitive advantages during their conversation. He praised the company's business model, noting that "one of Arm's greatest advantages is not having to worry about the supply chain." He explained that Arm's supply chain relies on electrons and intellectual property distribution, which can scale infinitely without physical constraints. "So I love its business model," Huang said, underscoring why acquiring Arm would have been strategically valuable.

Huang

What Makes Arm's Position Unique in the AI Chip Market?

Arm's advantage extends beyond just supply chain flexibility. Haas revealed that the company faces fewer export restrictions than traditional AI chip manufacturers. When discussing US export controls on AI chips, Haas stated that "it's almost impossible" for the US to prevent the export of AI CPUs to China because AI CPUs are widely used across many applications, making it difficult to determine which specific CPUs are intended for AI purposes. This regulatory flexibility gives Arm a competitive edge in global markets.

Haas

The semiconductor industry is experiencing unprecedented demand. Haas mentioned conversations with TSMC leadership, noting that executives at the world's largest chip manufacturer "have never seen the semiconductor industry cycle be so prosperous for four consecutive years." This sustained growth is driving adoption of diverse chip architectures, including Arm-based designs that compete with traditional x86 processors.

Haas

How Is NVIDIA Adapting Its Strategy With Arm Architecture?

Rather than acquiring Arm, NVIDIA has embraced partnership with the company. NVIDIA's new RTX Spark superchip demonstrates this collaboration, featuring a 20-core CPU with excellent single-thread performance. The chip incorporates NVIDIA's proprietary technologies, including CUDA for accelerated computing and CUDA Tile for tensor core processing, integrated into a single processor.

NVIDIA also developed a new data format called NVFP4 to compress large-scale language models, allowing more AI models to fit into system memory. This innovation reflects how NVIDIA is optimizing Arm-based systems for AI workloads, even without owning the underlying architecture. The company's strategy now focuses on layering its software and acceleration technologies on top of Arm's CPU designs.

Steps to Understanding NVIDIA's AI PC Strategy

• Local Agent Processing: NVIDIA's Arm-based PCs will run AI agents autonomously on the device itself, reducing dependence on cloud services and enabling offline functionality even when disconnected from the internet.
• Cloud Integration When Needed: For tasks requiring external resources, these PCs can call cloud APIs on demand, balancing local processing with cloud capabilities for optimal performance and cost efficiency.
• Operating System Importance: Contrary to claims that "software is dead," Huang emphasized that operating systems remain critical, as AI agents rely on system tools, APIs, and Skills files to unlock the full functionality of installed applications.

Huang painted an optimistic picture of the market opportunity ahead. He explained that the computer industry has historically been limited by the number of people using computers. With autonomous AI agents, robots, and autonomous vehicles, the addressable market could expand dramatically. "Instead of billions of people using computers, there will be tens of billions, and perhaps even more agents, robots, and autonomous vehicles using computers," Huang said. He concluded, "I feel that by now, the outcome is already determined. The scale of this multi-trillion-dollar industry may be ten times larger, and we're on the way".

The conversation between Huang and Haas revealed a semiconductor industry in flux, where traditional hierarchies are shifting. NVIDIA remains dominant in AI accelerators, but Arm's flexible licensing model and lack of supply chain constraints position it as an increasingly attractive alternative for companies seeking to diversify their chip suppliers. For Huang, the missed acquisition of Arm represents a strategic decision that, in hindsight, may have cost NVIDIA significant influence over the architecture powering the next generation of AI infrastructure.