Arm Partners with Allies to Enter the Intelligent Agent Edge Computing Domain

Overview

Morgan Stanley released a report summarizing significant progress made by Arm and its partners on a new AI computing architecture at Computex. The report indicates that AI is shifting from large models and data centers to distributed agent environments covering laptops, edge devices, robots, and cloud. Nvidia introduced the RTX Spark platform based on Arm architecture, aiming to localize AI processing capabilities and challenge Qualcomm's dominance in the Windows-on-Arm market. Meanwhile, Qualcomm unveiled an enterprise-grade AI architecture called Dragonwing, focusing on industrial and commercial use cases, strengthening its position in edge inference and distributed intelligence.

Core views

The report believes that the trend in AI development is shifting from centralized cloud computing models toward decentralized deployment on edge devices, presenting new opportunities for Arm architecture. Firstly, Nvidia’s RTX Spark platform integrates Grace CPU, Blackwell GPU technology, and unified memory, supporting up to 1 petaflop of AI performance and 128GB of unified memory. This marks Nvidia extending its CUDA ecosystem to client devices, especially PCs based on Arm architecture. If the N1/N1X chips succeed, developers will build AI applications on laptops and extend them to the cloud. Secondly, Qualcomm’s Dragonwing architecture aims to become part of enterprise AI infrastructure, applied in factory automation, drones, security cameras, and more. The design is expected to be based on Arm CPUs, possibly its own Oryon cores or older Arm processors. Lastly, for Arm, although it won't significantly impact Nvidia's profits in the short term, in the long run, as more AI agents run at the edge, Arm stands to gain increased licensing revenues from PCs and other devices. Additionally, Qualcomm using Arm CPUs for edge coordination also increases Arm’s future licensing potential.

Analysis framework

The report adopts a combination of industry dynamics tracking and technology roadmap analysis, interpreting key announcements from the Computex conference to assess Arm and its partners' strategic positioning in the AI edge computing domain. Firstly, the report analyzes the technical specifications and market positioning of Nvidia’s new platform RTX Spark, pointing out that this is Nvidia’s first major move into the Arm-based PC market, attempting to attract developers through its powerful CUDA ecosystem. Next, the report explores the application scenarios and technological foundation of Qualcomm’s Dragonwing architecture, highlighting its potential advantages in industrial and commercial AI solutions. Finally, the report combines the previously proposed 'rise of AI agents' framework to analyze Arm’s role in future AI computing, particularly emphasizing the advantages of its instruction set architecture (ISA) in power efficiency and potential licensing revenue growth.

Methodology notes

• Industry/sector analysis frameworkSupply-demand framework

  Core focus on supply in this industry

  The report evaluates competitiveness and development prospects in emerging markets by analyzing the supply capacity of Arm and its partners in AI edge computing.

• Competitive and strategic frameworkPorter's five forces

  Industry competitive landscape analysis

  The report analyzes the competitive dynamics between Nvidia and Qualcomm to reveal Arm’s strategic position within the supply chain and the market pressures it faces.

• Valuation methodsSOTP Segment Valuation

  Segment valuation method

  The report uses segment valuation to value Arm’s IP business and chip business separately, reflecting the value composition of its diversified business structure.

• Industry/sector analysis frameworkAdoption S-Curve

  Technology adoption rate analysis

  The report analyzes the penetration of Arm architecture in the PC market to predict its future growth potential in AI edge computing.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Arm Holdings plc (ARM.US)

  Benefiting from the development of AI edge computing architecture

  Strengths

  Low-power RISC architecture suitable for edge devices

  Weaknesses

  Dependent on partners' successful promotion

  Comparison

  Has cost and power efficiency advantages over x86 architecture in the PC market

  Risks

  Pressure from new entrants and existing competitors

• NVIDIA Corp (NVDA.US)

  Entering the Arm-based PC market via the RTX Spark platform

  Strengths

  Strong CUDA ecosystem and GPU technology

  Weaknesses

  Limited experience in the Arm-based PC market

  Comparison

  Compared to Qualcomm’s mobile advantage, Nvidia is more competitive in the PC segment

  Risks

  Needs to overcome competition from traditional x86 architecture

• Qualcomm Inc (QCOM.O)

  Expanding enterprise AI market through the Dragonwing architecture

  Strengths

  Deep accumulation in mobile and edge computing

  Weaknesses

  Lower market share in high-end PCs

  Comparison

  Compared to Nvidia's GPU dominance, Qualcomm excels in CPU and modem areas

  Risks

  Must respond to challenges from companies like Nvidia in AI computing

Key data

• RTX Spark Platform AI Performance1 petaflopPeak AI performance offered by Nvidia's new platform
• RTX Spark Platform Unified Memory128GBMaximum supported unified memory capacity of the platform
• Arm Target Price$202Target price derived using the SOTP valuation method
• Arm Current Stock Price$353.29Closing price as of May 29, 2026
• Arm RatingEqual-weightLatest rating given by Morgan Stanley to Arm

Impact & implications

The report believes that as AI computing shifts from the cloud to the edge, the strategic moves by Arm and its partners will reshape the competitive landscape in the semiconductor industry. Nvidia’s RTX Spark platform may break Qualcomm’s monopoly in the Windows-on-Arm market, while also creating new licensing revenue opportunities for Arm. Qualcomm’s Dragonwing architecture helps it move beyond its role as a smartphone modem supplier, expanding into broader AI application scenarios. Overall, this trend benefits Arm’s long-term development in AI edge computing.

Risks

• Threats from new entrants and existing competitors
• Heavy reliance on smartphone licensing income
• Revenue uncertainty from China joint ventures
• Pending litigation risks

What to watch

• Whether growth in Arm Neoverse and AI-related computing exceeds expectations
• Increases in licensing rates across mobile, PC, and automotive sectors
• Long-term CSS agreement signings with major clients
• Launch progress of v10 architecture or cloud GPU cores