Arm Holdings shares fell 5 percent to $225.43, erasing over $12 billion from its $252 billion market valuation, after the company warned of softness in the smartphone market and challenges securing manufacturing capacity for its new AI chip beyond the first $1 billion of demand despite projecting more than $2 billion in revenue across fiscal 2027 and 2028.
The smartphone warning is the immediate headwind. Arm designs power virtually every smartphone globally, and the sector faces memory chip shortages driving up electronics prices and slowing sales. CEO Rene Haas called smartphone unit growth "slightly negative" in the last quarter, primarily affecting the lower end of the market. Licensing revenue beat expectations, driven by new designs for AI infrastructure, but royalty revenue, which depends on actual device shipments, came in lighter. Qualcomm's dour smartphone outlook added pressure, as memory shortages weigh on the entire ecosystem.
The AI chip supply concern is the structural issue. Arm secured capacity for the first $1 billion of demand for its new data center processor, produced on TSMC's 3-nanometer process with advanced packaging. Beyond that, Haas said the company has yet to lock in manufacturing capacity, wafers, and testing equipment. The chip targets AI agents and data center workloads, a high-growth category that analysts expect to drive billions in revenue. Arm's shares had doubled year to date on AI optimism, but the supply constraint reminded investors that chip availability remains the binding bottleneck even for architecture leaders.
For SF founders and investors, Arm's update shows the AI hardware boom is not lifting every semiconductor name equally. Arm's architecture dominates mobile, edge devices, cloud CPUs, and custom silicon from Apple, Qualcomm, AWS, and Google. The smartphone cyclicality affects royalties from high-volume consumer devices. AI exposure comes through licensing for data center designs and custom chips. The supply constraint for Arm's own AI chip highlights that even architecture providers face manufacturing bottlenecks as Nvidia, AMD, and hyperscalers compete for TSMC capacity.
The cyclical consumer-device slowdown versus AI demand tension is the core narrative. Smartphones generate 70 percent of Arm's royalty revenue through high-volume shipments. Memory shortages and softening demand at the low end create near-term pressure. Data center AI designs provide higher per-unit royalties and longer design cycles, but volume is lower and ramp-up is slower. Arm's beat-and-raise quarter still met skepticism because investors wanted confirmation that AI licensing and royalties would offset smartphone weakness. The $2 billion AI chip revenue projection over two years is credible but requires supply chain execution.
Arm's architecture position gives it pricing power in a fragmented chip stack. Nvidia owns GPUs and CUDA. Apple designs custom silicon on Arm for iPhones and Macs. Qualcomm powers Android devices. AWS Graviton and Google Axion run cloud workloads on Arm cores. Hyperscalers increasingly design custom CPUs to reduce Nvidia dependency. Arm collects royalties on every chip shipped, giving it exposure to the entire ecosystem. The question is whether Arm can capture more value as the stack diversifies. Licensing revenue beat expectations, suggesting customers are paying for new AI designs. Royalty growth lags as smartphone shipments slow.
Startups and investors should treat Arm as a leveraged bet on the entire semiconductor ecosystem rather than a pure AI chip play. The architecture touches mobile, edge, cloud, and custom silicon. Smartphone cyclicality creates volatility, but AI data center growth provides the offset. Supply constraints remind everyone that TSMC capacity remains the bottleneck.
Also read: AWS Bedrock AgentCore Payments makes stablecoin micropayments the default for autonomous agents • Block's 40 percent AI-driven layoffs make Dorsey's automation thesis a Wall Street reality • Perplexity's Mac app turns search loyalty into a desktop agent wedge before Apple and OpenAI close the gap
