The narrative of Intel Corporation (INTC.US) transitioning from an AI edge role to a "full-stack AI computing platform" is becoming increasingly bullish.

Against the backdrop of a surge in CPU demand in data centers and the growth trajectory of advanced chip manufacturing processes such as 18A, Wall Street financial giants have shown increasing bullish sentiment towards Intel Corporation(INTC.US), a super giant in x86 architecture CPUs. This has helped drive Intel Corporation's stock price up by 145% since April, surpassing a market value of over $500 billion. Thanks to the continued bullish sentiment towards Intel Corporation on Wall Street and positive outlook from executives, Intel Corporation's stock price surged over 8% to $116 in pre-market trading on Wednesday. The company's executives emphasized that Intel Corporation's production capacity for its advanced semiconductors Intel 3 and 18A nodes is significantly increasing, and the demand for Intel Corporation's CPUs in the global data center sector continues to explode. According to Intel Corporation CFO David Zinsner, speaking at the Bank of America Corp's 2026 Global Technology Conference on June 2, everything in the short term revolves around supply. They are ramping up chip supply, and within the next few quarters, the supply of 3A and 18A nodes will see a meaningful increase to meet the basic demand they have seen from a supply perspective. The day before, Intel Corporation's CEO, Liwu Chen, stated at the Computex conference in Taipei that Intel Corporation is transitioning from being seen as a "PC CPU supplier" to covering PC, edge, data center, AI rack, custom chips, and advanced packaging as a full-stack AI computing platform. The strategic focus is first on the 18A process and Xeon 6 Plus (Xeon 6+), which is based on the 18A process. Wall Street research firm Melius Research recently raised Intel Corporation's target price from $100 to $150, highlighting the combined factors of the data center CPU demand boom and the improving profit outlook for the 18A advanced chip manufacturing process that are driving Intel Corporation's super bull market narrative. Melius Research analyst Ben Reitzes stated that their core logic is not simply betting on a PC recovery but repricing Intel Corporation as the most core beneficiary of CPUs in the AI inference and intelligent computing era, challenging Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR's advanced process/outsourcing narrative. The x86 architecture interprets "classic never goes out of style": CFO boldly announces that Intel Corporation is not afraid of ARM! From Intel 7 to 18A advanced process, advanced chip manufacturing is on the rise Intel Corporation CFO Zinsner also talked about the plans to transition from Intel 7 to 3 and even 18A nodes in the coming years. Zinsner stated, "In fact, even this year, we will see a significant increase in Intel 7 wafer starts, at least initially; and then maybe by next year, I think we could start to bring it down a bit and let Intel 3 and 18A fill the huge supply gap." Zinsner also pointed out that the 18A on the laptop side is ramping up rapidly and is their fastest product ramp in at least the past five years. This executive also expressed confidence in competing with ARM architecture products and other solutions in the client and data center market. Zinsner said, "And in the data center side, I would say we have strong competitiveness. In some cases, we have a very strong multi-product line. In other cases, especially in the case of multi-threading, not so good." "Yes, of course, we have been attentive to ARM-based solutions, in the client area, it's good competition. I think, given our large market share and strong performance, we have shown that we can compete well with these ARM-based solutions. So I am very confident in our product portfolio's ability to continue to dominate the market." Zinsner also discussed the near-endless demand for Intel Corporation's CPUs in the global AI data center construction frenzy and noted that the company will have no problem achieving significant revenue and profit growth in the data center sector. Zinsner said, "Obviously, as we transition from training to inference, from inference to agentic AI and multi-agent systems, the CPU: GPU ratio is significantly increasing. Therefore, this will only drive a lot of CPU demand. To be honest, the scale is already large enough. I mean, we have enough demand, and if we can execute the supply ramp well, there should be no problem in substantially increasing sales in the data center sector." The chief financial officer also pointed out that due to the rapid popularity of AI intelligent agents worldwide, the CPU market has indeed experienced uncontrollable "explosive growth." Furthermore, Zinsner discussed a series of restructuring measures taken by CEO Liwu Chen. Zinsner said, "We used to have 12 levels of management. He compressed it to 6 levels of management. We used to, I think at the peak, have over 400 vice presidents, he compressed it to under 200. We used to have over 100,000 employees. He compressed it to less than 80,000 employees." As media reports suggest that Intel Corporation has requested PC manufacturers to use chip products manufactured using its 18A advanced chip manufacturing process, Wall Street investment giant Wedbush Securities believes that this move is a positive signal indicating that this semiconductor manufacturing giant is prioritizing expanding its profit margins. Wedbush Securities analyst Matt Bryson wrote in a report to clients on Wednesday, "In our view, this strategy is very reasonable, as Intel Corporation should prioritize using older and already expanded process nodes to produce higher-margin Xeon data center-level CPUs, while Intel Corporation's ability to enable new capacity (as it has existing clean room space) is a strategic advantage, enabling it to leverage this advantage." "In our view, the question is instead, how strong will the actual performance of the 18A node and the chip products manufactured based on this be (recent hints from CEO Liwu Chen suggest rapid improvement in yield)." At the 54th Global Technology, Media, and Telecommunications Conference hosted by JPMorgan Chase on Wall Street in May, Intel Corporation CEO Liwu Chen stated that Intel 18A (an advanced chip manufacturing process below 2nm to 1.8nm) has supported Panther Lake mass production with a monthly yield increase of about 7%, exceeding Intel Corporation's internal expectations. According to the latest developments revealed by Liwu Chen, the 0.5 PDK for Intel 14A has been released, and a 0.9 PDK is scheduled to be released to external customers in October. The team has also started long-term planning for the 10A and 7A process nodes. Liwu Chen also stated that as the focus of AI compute infrastructure shifts from training to inference, CPUs are becoming increasingly important and indispensable in the AI era, with the ratio of CPUs to GPUs accelerating from 1:8 to 1:1, or even up to 4:1. In addition, Intel Corporation's business plan shows that it is actively seeking ASIC business, offering customized AI CPU or AI GPU chip solutions. Intel Corporation is launching an attack on the "AI compute industry leader" position with Xeon 6 Plus, Rack Scale, and EMIB-T The core of this narrative repair for Intel Corporation is not just a simple "comeback of an old-school CPU manufacturer" but an attempt to reposition itself at the center of the AI infrastructure value chain: as AI transitions from large model training to inference, intelligent agents, multi-intelligent agents, and reinforcement learning, the system bottleneck is no longer determined solely by GPU computing power. Tasks such as task scheduling, state management, tool invocation, database access, network requests, code execution, and security isolation will significantly increase and be handled by general-purpose CPUs. Intel Corporation CFO David Zinsner emphasized at the Bank of America Corp Global Technology Conference that as AI transitions from training to inference, and then to Agentic AI and multi-agent systems, the CPU/GPU ratio is significantly increasing, leading to an "explosive growth" in demand for data center CPUs; these positive factors are also the underlying market logic for Melius Research analyst Ben Reitzes' $150 highest target price. Li Wu Chen stated at the Computex conference in Taipei that Intel Corporation is transitioning from being seen as a "PC CPU supplier" to covering PC, edge, data center, AI rack, custom chips, and advanced packaging as a full-stack AI computing platform, with a strategic focus on the 18A process and Xeon 6 Plus (Xeon 6+). This also means that 18A is not just a PC-side story but a key proof of Intel Corporation's re-entry into the advanced chip competition in the data center. In the Agentic AI framework, Intel Corporation is trying to shape the concept of "CPU orchestrates the show" as a new anchor in AI infrastructure pricing. Traditional AI inference can heavily rely on GPUs for matrix computation, but intelligent agent workflows will split a task into planning, retrieval, file read/write, rule verification, code execution, tool invocation, and result verification stages, which naturally depend more on general-purpose CPUs, memory access, I/O, and system scheduling. In traditional inference scenarios, the GPU to CPU resource ratio is about 7:1, but in intelligent agent workflows, it may approach 1:1, or even weigh more towards CPUs; this does not mean that GPUs are unimportant, but that as AI data centers transition from a "GPU island" to a heterogeneous system era of CPU+GPU+ASIC+optical interconnect networks+storage, Intel Corporation is betting on reallocating system-level computing power after the complexification of AI applications. Xeon 6 Plus, Rack Scale Blueprint, and collaborations with partners such as Foxconn and SambaNova demonstrate that Intel Corporation is shifting from selling chips to selling "deployable AI rack solutions." This aligns with the trend of NVIDIA Corporation's DGX/GB series, AMD's rack-level solutions, and cloud providers' in-house AI cluster solutions: enterprise customers are increasingly unwilling to assemble chips, networks, storage, cooling, and software stacks themselves, and instead prefer to purchase validated cabinet-level infrastructures. If Intel Corporation can leverage the x86 ecosystem, ODM partners, server customers, and open standards to enter the AI cabinet market, there is an opportunity to transform CPU demand revival into system-level AI revenue like NVIDIA Corporation's Blackwell rack clusters, rather than just improving the ASP of single processors. "Decoupling inference" is perhaps the most noteworthy engineering direction in Intel Corporation's AI narrative. SambaNova RTU, Intel Corporation Xeon, and NVIDIA Corporation's GPU collaborative work essentially breaks down the AI inference pipeline to the hardware best suited for execution: GPUs handle large model prompts or highly parallel operators, RTUs handle specific token generation, CPUs handle agent scheduling, tool execution, and workflow control. If such heterogeneous combinations can achieve a 2 to 3 times end-to-end performance improvement in real enterprise scenarios, then Intel Corporation can bypass the disadvantageous battlefield of "head-on competition with NVIDIA Corporation in GPUs" and emphasize its irreplaceability in AI system control, CPU scheduling, and server platform layers. Intel Corporation is aimed at the most critical bottleneck in the AI supply chain - advanced packaging Currently, Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR's CoWoS advanced packaging is one of the core production capacity bottlenecks for integrating NVIDIA Corporation's GPUs, ASICs, and HBM, causing delays in AI chip delivery not only in wafer manufacturing but also in advanced packaging. If Intel Corporation's EMIB-T can enter the next generation TPU supply chain evaluation by Alphabet Inc. Class C, it means that Intel Corporation's contract manufacturing and packaging business may receive actual verification from major cloud providers on a large scale. Although final orders will depend on yield, cost, and supply chain strategy, this also indicates that Intel Corporation is vying for higher-value segments of AI chip design, manufacturing, and packaging with EMIB+Foveros advanced packaging and 18A and custom silicon businesses. From Alphabet Inc. Class C's TPU and Telefonaktiebolaget LM Ericsson Sponsored ADR Class B's next-generation communication chips to potential TPU packaging support, Intel Corporation's custom chip strategy is filling in the relatively weak link of the past few years. Major cloud providers have proven that the AI era will not only have general-purpose GPUs, but also TPUs, Trainium, ASICs, DPUs/IPUs, network chips, and specialized accelerators, so Intel Corporation's advanced packaging business has the potential to become a mainstream high-end manufacturing outsourcing entry point for the customized AI compute infrastructure supply chain of global cloud computing giants. If Intel Corporation can offer "cutting-edge chip design support+manufacturing+advanced packaging+system-level reference platforms," it won't just compete with AMD for x86 server market share, but will also compete for the mainstream high-end manufacturing outsourcing entry point of global cloud computing giants' customized AI compute infrastructure supply chains. This is also the high elasticity assumption implied in Melius Research's star analyst Ben Reitzes' $150 target price: the market will no longer value Intel Corporation solely based on traditional PC/server CPU cycles, but will assign it a comprehensive option value for a part of the AI contract manufacturing, AI advanced packaging system, and AI compute infrastructure system platform.