AI demand

Despite recent stock volatility, AI demand remains strong, with experts citing energy as the primary constraint. While some companies like Meta and xAI are leasing excess capacity, the overall market faces supply shortages for compute power and components. Enterprises are shifting from “tokenmaxxing” to “valuemaxxing,” prioritizing ROI and a more rationalized approach to AI spending. This trend is expected to sustain long-term demand.

Samsung’s shares plunged 8% despite exceeding Nvidia and Apple’s profits, as investors reacted negatively to AI demand not meeting sky-high expectations. This sentiment triggered a broader sell-off in semiconductor stocks globally. Concerns about the sustainability of AI spending relative to rising memory chip prices, coupled with geopolitical shifts in chip supply chains, contributed to the market downturn. Major memory chip manufacturers and related ETFs saw significant declines.

Micron CEO Sanjay Mehrotra attributes the current memory chip shortage not solely to manufacturers, but also to past customer pricing pressures that limited investment capacity. Despite financial challenges, Micron continued strategic investments. AI-driven demand is now surging, with the supply crunch expected to persist beyond 2027. Micron is investing $200 billion in new facilities, impacting consumer electronics pricing.

Anthropic is rapidly expanding its compute infrastructure in the Asia-Pacific region to meet soaring demand for its AI models. The company is actively hiring data center professionals in Australia and Japan, focusing on engineering, operations, and deal sourcing. This strategic move underscores intense global competition for AI processing power and highlights Australia’s advantages for infrastructure development, including land, renewable energy, and political stability, while Japan offers a strong domestic AI focus and evolving grid. Securing adequate energy supply remains a key challenge across the region.

The CNBC Investing Club raised its price target on Nvidia, driven by unprecedented AI demand and an “extraordinary quarter.” CEO Jensen Huang highlighted Nvidia’s unique role in supporting all major AI models and powering hyperscalers. The company is expanding into new AI frontiers, including “physical AI” and the CPU market, potentially unlocking significant new revenue. Despite competitive pressures from custom silicon, Nvidia’s strong performance and strategic positioning solidify its AI dominance.

The proclaimed demand for AI might be inflated, with token consumption becoming a distorted metric. Anthropic is strategically pricing services based on actual usage, shifting from flat-rate to per-token billing to ensure revenue correlates with real value. This approach positions them to better navigate a potential market correction, contrasting with models potentially incentivizing inefficient consumption. Other companies are also exploring ways to measure genuine AI productivity beyond token counts.

The memory chip market faces sustained tight supply due to AI demand, with Micron, Samsung, and SK Hynix projecting shortages lasting years. Despite strong earnings, stock dips highlight investor concerns about profit sustainability amid aggressive capacity expansion. Long-term customer agreements signal clients preparing for prolonged scarcity, while some investors fear oversupply and margin erosion as new facilities come online. Analysts forecast an extended upcycle, but the market grapples with the future of high margins.

Cisco’s stock plummeted 12% due to soaring memory component costs, driven by AI demand. The global memory shortage, particularly for high-bandwidth memory needed in AI data centers, has driven up prices and constrained supply for other electronics, impacting companies like Apple and Dell. Cisco is adjusting product prices and renegotiating supplier contracts to manage these rising costs, despite reporting better-than-expected quarterly results. The company’s gross margin also saw a slight decrease, attributed to product mix and higher memory expenses.

Market rotation is accelerating, with industrials outperforming tech. Cisco’s earnings beat expectations but saw margins pressured by rising memory costs, impacting Apple and PC makers. Despite these challenges, demand for AI infrastructure and Cisco’s strategic pricing offer resilience. Apple’s Siri overhaul is staggered, but iPhone sales remain strong, supported by AI integration potential.

AMD CEO Lisa Su defends the company’s outlook amidst a surge in AI demand. She highlighted accelerating growth in the datacenter business, driven by increasing enterprise needs for AI development. Despite surpassing Q4 expectations, AMD’s stock dipped after its Q1 guidance, which still exceeded estimates. Su expressed confidence in AMD’s position, particularly with the upcoming launch of their AI system, Helios, expected to mark an inflection point in the latter half of the year.