"CPU Renaissance" Sweeps the Globe! AI Intelligent Bodies Spark a Super Cycle for CPUs, Two x86 Giants Join Hands with ARM to Attend a Super Bull Market.

Against the backdrop of a surge in CPU demand in data centers, Citigroup, a financial giant on Wall Street, significantly raised its 12-month target stock prices for the two x86 architecture CPU super giants - Intel Corporation (INTC.US) and AMD (AMD.US) on Monday. This financial giant also substantially raised its expectations for the outlook of the data center CPU and overall CPU market size in its latest research report. ARM instruction set architecture owner Arm (ARM.US) has also seen its stock price rise rapidly recently and has received target stock price increases from several major Wall Street banks, highlighting the huge advantage of the ARM architecture in energy efficiency and low power consumption in the AI data center field, which is highly favored by institutional investors. With the launch of Anthropic's Claude Cowork and autonomous task execution super AI agent tools like OpenClaw exploding in 2026, the wave of AI agents is rapidly sweeping the globe, and the bottleneck of AI computing architecture is shifting from GPU core throughput based on matrix multiplication to data center CPUs core based on control flow, task scheduling, and memory/IO coordination. High-performance CPUs for ultra-large-scale AI data centers are facing a serious supply shortage. As the AI wave sweeps the globe, the main investment theme in AI computing power is shifting from "a single point AI computing power competition centered around GPUs" to "a full-stack AI computing system driven by AI agents." The next round of super-alpha returns will no longer be exclusive to the AI GPU/AI ASIC field, but will also systematically expand to include CPU, storage, PCB, liquid cooling systems, ABF substrates, and a wide range of wafer foundries and other full-stack AI computing infrastructure. In this shift in AI narrative, CPUs, optical interconnects, and memory chips may be the biggest winners. In the past two years, the AI narrative has been dominated by GPUs, with CPUs appearing as "supporting actors" in the AI arms race; however, with the open-source nature of autonomous AI workflows like OpenClaw leading to a significant increase in inference workloads, data arrangement, task scheduling, memory access, network communication, and multi-tools invocation, the market has realized that without powerful CPUs as the system's centerpiece, GPU clusters cannot operate efficiently. This fundamentally brings CPUs back to the forefront from being "underestimated infrastructure" to being at the center of the chip stage with a clear "Renaissance" style retro wave. Although the geopolitical storm in the Middle East is not yet over, Wall Street's bullish sentiment towards the global stock market is becoming more intense. After going through the initial series of intense selling turmoil, Wall Street institutional investment forces seem to be filtering out the "war-related noise," shifting away from viewing the war as the "core variable determining market direction" as they did in early March, and are largely "ignoring the sounds of war." Several financial giants on Wall Street directly attribute the current upward resilience of the stock market to the continuous upward revision of corporate profit expectations, especially the strong profit expectations of technology companies closely associated with the exploding demand for AI computing infrastructure. The AI computing infrastructure boom enters the "CPU Renaissance moment" Wall Street analysts are expanding the narrative around AI computing infrastructure from "GPU dominance/single-core driven" to "AI GPU/ASIC + CPU + HBM/DRAM/NAND storage chips + optoelectronic interconnect led high-speed data center connection system" full-stack computing revaluation. Citigroup recently significantly raised target prices for Intel Corporation and AMD, and significantly raised CPU market size expectations. In a report to clients led by senior analyst Atif Malik at Citigroup, the analyst team wrote: "We believe that the data center server CPU [Total Addressable Market, or CPU TAM] could increase significantly from $29.3 billion in 2025 to $131.5 billion in 2030, with a potential Compound Annual Growth Rate (CAGR) of up to 35%. We expect that the general CPU systems market dominated by traditional cloud computing and high-performance computing will grow at a CAGR of 20% to reach $50.9 billion by 2030; AI head nodes are expected to grow at a CAGR of 21% to reach $21.1 billion by 2030; and the data center CPU market dominated by AI agents like OpenClaw is expected to grow at an astonishing 185% CAGR, to reach $59.4 billion by 2030, with a projected 45% share of the overall market by 2030, becoming the fastest-growing segment." "We expect that by 2030, general-purpose CPUs will account for about 39% of the Total Addressable Market, AI head nodes will account for 16% in 2030, and intelligent data center server CPUs like AI agents will account for a whopping 45% of the potential total market size in 2030. For reference, Arm previously forecasted the potential Total Addressable Market for CPUs as approximately $100 billion by 2030, while AMD previously forecasted a Total Addressable Market of $120 billion by 2030 before our latest expectations. Additionally, by 2030, we expect that Intel Corporation will have approximately 47% market share, AMD will have 34% market share, and ARM/other vendors will increase from less than 5% in 2025 to about 19% market share." Atif Malik's analyst team wrote. Citigroup's latest outlook highlights that the introduction of autonomous AI automated workflows (i.e., AI agents) brings not only linear growth but a substantial increase in common computing density. For the 12-month target stock prices for the two x86 architecture CPU super giants, Citigroup raised Intel Corporation's target stock price from $95 to a significant $130 and AMD's target stock price from $358 to a substantial $460. By comparison, as of Monday's close, Intel Corporation's stock price was around $108, with a year-to-date gain of 200%, and AMD's stock price was around $420, with a year-to-date gain of 97%. Regarding the bullish logic for Arm by Wall Street, the core logic has shifted from being a "smartphone IP licensing company" to one of the core beneficiaries of the AI data center CPU and Agentic AI infrastructure mega trend. As for the highest target price, a well-known Wall Street investment firm Evercore ISI recently raised its target stock price for Arm to the highest level on Wall Street at $240, with another very aggressive new target price of $300 from Bernstein. The core reasoning of these institutions' bullish stance undoubtedly focuses on Arm's shift from being a traditional IP licensor for smartphones and lightweight consumer electronics to a direct participant in AI data center silicon chips and supercomputing platforms. The average target price from 22 analysts on TipRanks is around $255; as of Monday's close, Arm's stock price was at $215. It is understood that the U.S. cloud computing and e-commerce giant Amazon.com, Inc. (AMZN.US) and Facebook's parent company Meta Platforms Inc. (META.US) have reached a multi-billion dollar long-term agreement. This social media giant will lease hundreds of thousands of Amazon.com, Inc.'s own ARM architecture general data center server CPU chips for its large-scale new AI data centers to meet the massive artificial intelligence inference workload for Facebook and Instagram social media users. Graviton is Amazon.com, Inc.'s own ARM architecture general server CPU for AWS cloud computing business, mainly responsible for general computing, scheduling, data preprocessing/post-processing, service arrangement, and some AI inference-related scheduling and coordination tasks in AI data centers. Arm is considered one of the biggest winners in the global artificial intelligence frenzy. NVIDIA Corporation's self-developed Grace CPU is based on the ARM architecture, Amazon.com, Inc.'s self-developed Graviton data center server processor also uses ARM architecture, and similar products include Google Axion Processors built on ARM Neoverse created by Alphabet Inc. Class C as well as Microsoft Corporation's Azure Cobalt 100 self-developed ARM architecture data center CPU. The ARM architecture is evolving from being "the king of smartphones" to becoming one of the cornerstones of AI cloud era computing infrastructure. The reduced instruction set architecture used by ARM makes servers based on its design have significant advantages in energy efficiency and low power consumption compared to Intel Corporation's x86 architecture when performing AI inference/training tasks. This feature makes the ARM architecture particularly suitable for data center servers, where it can efficiently cooperate with AI GPUs to meet the almost endless demand for AI inference/training computing power. The "AI bull market" narrative can't be suppressed by the Middle East conflict! GPUs are no longer the only game in town. The trend of AI agents detonates CPUs. In the context of the Middle East conflict, investors who have experienced the latest round of global stock market turmoil, from phase lows to new highs, undoubtedly realize the most profound lesson - not to treat every GEO Group Inc political headline as the trend itself. In recent weeks, the market has put the core logics such as "Trump backing down at the last moment," "significant upwards revision trend of tech sector performance driven by the AI computing infrastructure chain," and "the ultimate return of the US, Iran, and Israel to the negotiating table" at the top of the GEO Group Inc political chessboard. In the AI agent chain, a significant amount of workload is not only consumed by GPU token generation, but also by CPU-dominated tasks such as Python interpretation, web scraping, database retrieval, RAG index access, lexical processing, task queue scheduling, RPC/IPC communication, KV state updates, and more. This means that what determines the user experience is increasingly not just the peak computing power of a single GPU, but whether the CPU has enough cores, thread concurrency, cache levels, memory bandwidth, PCIe/CXL/interconnect scheduling capability to support high-frequency tool calls and high-density task switches. Once the CPU cores, memory subsystem, or I/O scheduling are inadequate, the GPU, despite its nominal computing power abundance, will collapse due to data preparation, task coordination, and system waits. Therefore, without a doubt, the bottleneck of the AI computing architecture is shifting from GPUs with core throughput based on matrix multiplication to data center CPUs with core control flow, task scheduling, and memory/IO coordination at the center. This change's root cause lies in the essential shift in workload paradigms. CPUs are no longer just general computing chips but are becoming the control plane processors, system orchestration engines, and resource scheduling centers of the AI agent era. The fact that CPUs are no longer just "supporting roles" but have become the new bottleneck determining system throughput, latency, and resource utilization is not an emotional judgment but rather a necessary result of the evolution of AI workloads from "inference computing problems" to "complex systems engineering problems." Recent forecasts by Morgan Stanley show that the explosion of AI agents signifies a structural shift from computation to orchestration, leading to an estimated incremental market space for CPUs of $32.5 billion to $60 billion by 2030, and dramatically expanding the Total Addressable Market for server-level CPUs to $82.5 billion to $110 billion. A forecast report by TrendForce shows that in the age of AI agents, the CPU:GPU ratio may shift from the traditional 1:4 to 1:8 in AI data centers to 1:1 to 1:2, a significant revaluation. GF Securities, a well-known investment research firm, recently published a research report stating, "Data center server CPUs are playing out a supercycle." GF Securities suggests that while x86 architecture remains mainstream, the company believes that servers CPUs based on ARM architecture may also see accelerated growth through NVIDIA Corporation's self-developed ARM architecture server CPUs - Vera series computing cluster and dedicated application integrated circuits (i.e., AI ASICs), including Amazon.com, Inc.'s self-developed Graviton 5 and Alphabet Inc. Class C (GOOGL.US)'s self-developed Axion 2 - all of which are built on ARM architecture. GF Securities analysts added, "NVIDIA Corporation has introduced a separate Vera CPU rack (256 CPUs) for data center customers, early adopters including Alibaba Group Holding Limited Sponsored ADR, CoreWeave, Meta, and Oracle (Oracle Corporation). Within the Alphabet Inc. Class C ecosystem, the GPU/CPU ratio of the TPU computing system is being adjusted from 4:1 to 4:2. With the stronger demand outlook for TPU computing power after 2026, we expect shipments of Axion 2 to accelerate further. At the same time, Amazon.com, Inc.'s AWS self-developed Graviton 4/5 is also gaining traction through the Teton Max rack (18 CPUs) and Meta's standalone Graviton rack. For Arm as a company, AGI CPU demand is raised to $20 billion by the company." Wall Street's current bullish logic about Arm is shifting from being a "smartphone IP licensing company" to one of the core beneficiaries of the AI data center CPU and Agentic AI infrastructure mega trend. In terms of the highest target price, a well-known Wall Street investment firm Evercore ISI recently raised its target stock price for Arm to the highest level on Wall Street at $240, with another very aggressive new target price of $300 from Bernstein. The core reasoning of these institutions' bullish stance undoubtedly focuses on Arm's shift from being a traditional IP licensor for smartphones and lightweight consumer electronics to being a direct participant in AI data center silicon chips and supercomputing platforms. The average target price from 22 analysts on TipRanks is around $255; as of Monday's close, Arm's stock price was at $215.