Investment bank Wedbush believes Intel still has the ability to implement price increases without negatively impacting market demand.
$Intel (INTC.US)$ Confirmed price increases for certain client and server CPUs, citing rising supply chain costs and demand exceeding supply. Price hikes for client processors range from $30 to $50, while data center products have seen increases of hundreds or even thousands of dollars. An Intel spokesperson stated that recent price adjustments reflect current market dynamics, including higher supply chain costs and strong demand for Core Ultra 200S Plus processors.
In response, investment bank Wedbush believes Intel still has room to raise prices without negatively impacting demand. In a note to clients, Wedbush analyst Matt Bryson wrote: 'Given the ongoing shortage in server CPU supply, we believe Intel still has pricing power without dampening demand.'
He added: 'The real question is whether OEMs’ actual purchase prices (PX) will rise in tandem with official list prices, or if Intel’s latest move only affects distributor and retail channel pricing—which accounts for only a small portion of Intel’s business.'
Despite recent pullbacks in Intel’s share price, the stock is still up nearly 200% year-to-date. Bolstered by surging demand for data center server CPUs and progress in its foundry business, Intel—once mired in difficulties over the past few years—is regaining investor favor.
Demand for data center chips to support the large-scale construction of AI infrastructure is driving sales of Intel's flagship Xeon server processors. These general-purpose semiconductors—central processing units (CPUs)—are regaining prominence as they help convert AI software into revenue-generating services.
Over the past two years, the narrative around the AI supply chain has been almost entirely dominated by GPUs, while CPUs have had minimal presence in AI servers. The reason lies in the fact that during the training era, the primary bottleneck was parallel computing capacity: GPUs handled the heaviest matrix operations, while CPUs were largely relegated to general-purpose control and basic scheduling tasks.
However, with the explosive growth of AI agent and reinforcement learning (RL) workloads, the strategic importance of CPUs in data centers is undergoing a structural reassessment. The essence of an AI agent is not about generating longer answers, but rather decomposing a single request into an entire workflow. Models are no longer just producing a response—they are executing a process. Once AI shifts from 'a single computation' to 'running a workflow,' system reliance on CPUs rises significantly. This is because many critical workloads are ill-suited for GPUs—for example, task orchestration, thread scheduling, process management, sandboxed execution, pre- and post-processing, cache coordination, and state maintenance—all of which are quintessential CPU responsibilities. Particularly in multi-agent collaborative scenarios, where multiple agents run concurrently, invoke tools interchangeably, and share states, demands on CPU core count, thread count, single-core performance, and memory management capabilities have all increased substantially.
Dylan Patel, lead analyst at the prominent semiconductor research firm SemiAnalysis, stated plainly in an in-depth interview on April 8 that AI workloads are evolving from simple text generation toward complex agents and reinforcement learning, and that CPUs are now facing 'an extremely severe capacity shortage.'
The latest report from market research firm TrendForce corroborates this view: currently, the CPU-to-GPU ratio in AI data centers ranges from 1:4 to 1:8, but in the era of agentic AI, this ratio is expected to shift toward 1:1 to 1:2. In terms of market size, Creative Strategies forecasts that the data center CPU market will grow from $25 billion in 2026 to $60 billion by 2030; factoring in agentic AI-related demand, the market could approach $100 billion.
Intel’s recent server CPU price hike carries significant market implications—it validates 'demand outstripping supply' rather than merely reflecting cost pass-through. While select SKUs of the consumer-grade Core Ultra 200S Plus saw price increases of $30–$50, the more critical development is the larger hikes for data center Xeon CPUs. According to Tom’s Hardware, a specialized tech hardware media platform, the new suggested customer price for the data center Xeon 6980P is $13,955, up from $12,460 in 2025—a $1,495 increase. Moreover, Intel did not uniformly raise prices across the entire Arrow Lake family; instead, it selectively increased prices on specific SKUs that customers are willing to pay a premium for. This essentially represents an upward shift in product mix and pricing weight amid constrained supply.
However, it should be noted that actual transaction prices for data center hardware often differ from published list prices, as final pricing is influenced by factors such as order volume and strategic relationships between suppliers and customers. Therefore, although Intel has significantly raised the recommended customer prices for its data center Xeon CPUs, the impact on its average selling price for this quarter and the full year remains to be seen.