Intel previews Computex 2026 lineup for all mobile devices, desktops, and servers as 18A processor becomes base calling card

Intel will arrive at Computex 2026 in Taipei on June 2 with something it hasn’t had in a decade: a product across all computer categories built on a single production line. Panther Lake, a laptop chip introduced at CES in January, expands to include the Arc G3 and Arc G3 Extreme processors designed for the gaming handheld market. Nova Lake, a 52-core desktop chip with a new socket and a new CPU architecture, is expected to launch in the second half. Clearwater Forest, a 288-core server processor that was shipped at MWC in March, joins Xeon’s lineup of data centers and cloud computing. All of them are built on or designed around the Intel 18A, a 1.8-nanometer process node that includes RibbonFET gate transistors that are compatible with PowerVia’s rear power delivery and represent the most advanced manufacturing capabilities produced entirely in the United States. CEO Lip-Bu Tan will deliver the keynote address. The site is 40 kilometers from TSMC headquarters. The message is not subtle.

Products

Panther Lake was introduced as the Core Ultra Series 3 at CES in January and is already shipping in more than 200 laptops. The chip delivers a total platform TOPS of 180, including 120 TOPS from its Xe3 integrated GPU and 50 TOPS from the NPU 5 neural processing unit, and claims a 60 percent improvement in multi-threaded performance over its predecessor at the same power. The Computex expansion brings Panther Lake to gaming laptops with the Arc G3 platform: a 14-core design with two performance cores, eight efficiency cores, and four low-power cores paired with a 10 or 12-core Xe3 GPU in an adjustable power envelope of 25 to 80 watts. MSI, OneXPlayer, GPD, and Acer are expected to show off handheld devices using Arc G3 chips at the event, with reports suggesting a Microsoft Xbox-branded handheld could make an appearance.

Nova Lake, called Core Ultra Series 4, is Intel’s next desktop platform and will be previewed at Computex before launching in late 2026. The chip scales from 8 to 52 cores using the new Coyote Cove cores and high-performance Arctic Wolf cores, introduces the LGA 1954 socket, and includes Xe3 graphics, Thunderbolt 5, and Wi-Fi 7. The power range goes from 35 to 175 watts, indicating a high-performance desktop design that includes both high-end desktops. Nova Lake adopts what Intel calls “last-level cache”, a design approach inspired by AMD’s success with large L3 caches that prioritize keeping data close to CPU cores. Intel’s first-quarter earnings reveal that AI-driven CPU demand is real: data center and AI revenue grew 22 percent year-on-year to $5.1 billion as the agency’s AI workload shifts processing needs back to CPUs and away from the GPU-only model that defines the training period.

The server

The Clearwater Forest, officially unveiled at MWC in March as the Xeon 6+, is Intel’s most ambitious processor. It packs 288 high-performance Darkmont cores across 12 computing chiplets manufactured at 18A, assembled using Foveros Direct 3D packaging on core tiles built on Intel 3. The IPC is 17 percent higher than the previous generation, and the chip targets cloud computing and dense computing workloads that grow from in the continuation of the work of production to the development of A. The shift to agent AI is driving the need for predictive computing at all major cloud providers: Meta has made more than $140 billion in chip purchases from Nvidia, AMD, and Amazon, and the predictive workload of those chips continuously requires CPU resources for processing, memory management, and real-time decision-making required by autonomous AI agents.

Intel’s server story at Computex also includes updates on Crescent Island, its dedicated accelerator, and Jaguar Shores, a rack-scale computing platform designed for the AI ​​data center in the late 2020s. Neither product has been officially launched, but both are expected to receive architectural details in Tan’s keynote. The inference accelerator is Intel’s attempt to compete directly with Nvidia’s well-made products instead of conceding the AI ​​acceleration market entirely. Whether Intel can build a competitive inference chip while simultaneously focusing its innovation business and launching three client platforms is an operational question Computex won’t answer but can’t help but raise.

The process

The cable that connects every product at Computex is 18A. Panther Lake is the first locally built consumer chip. Clearwater Forest is the primary server host. The Arc G3 handheld processors are the first gaming-focused silicon. Nova Lake will be the first desktop chip, although reports indicate that more than 90 percent of Nova Lake’s computer chips will be produced by TSMC with its N2 process instead of Intel’s fabric, which is a concession to the fact that Intel’s manufacturing capacity is not enough to provide for internal needs and external customers at the same time.

That agreement is important because the 18A node is not just a manufacturing process. It’s a product that Intel sells to Apple, Amazon, Musk’s Terafab, and every other company that has signed or is negotiating an innovation agreement. Intel recently hired Qualcomm veteran Alex Katouzian to lead a new Client Computing and Physical AI group, a sign that the company sees local AI, the kind of processing that runs on PCs, handhelds, and edge devices instead of cloud data centers, as the next wave of chip demand. Computex’s product line is a proof of concept: if 18A can produce competitive chips for all laptops, mobile devices, desktops, and servers, the foundry’s pitch to foreign customers becomes more credible than a road slide.

The market

GitHub suspended new Copilot subscriptions after the agent’s AI workflow consumed more computing than users paid for, an early sign that the economy of agent AI will push processing to local hardware. When AI agents run continuously on a cloud infrastructure, costs are proportional to usage and ultimately manageable with low subscription rates. When those agents run locally, on a laptop with 180 TOPS of AI processing power or a desktop with 52 cores and a large cache, the economics shift from cloud costs per query to one-time hardware purchases. Intel’s bet is that the AI ​​PC is not a marketing label but an architectural necessity: the agency era needs a local computer, and Intel’s chips are designed to provide it.

The competition is real. Qualcomm’s Snapdragon X Elite has established itself in the thin and light Windows market that uses high-end power. AMD’s Ryzen AI 300 and 400 series compete directly with Panther Lake for laptops and Arrow Lake Refresh for desktops. Apple’s M-series processors remain the benchmark for integrated performance in the consumer market. Nvidia’s server GPUs are selling for a million dollars each in China despite export controls, showing a level of demand that Intel’s data center products have never produced. Intel’s advantage at Computex is not that its chips are the best in any one category. It is because it has chips in all categories, all of which are produced in the area of ​​the process that also serves as the basis of the business of innovation, and the business of innovation is the reason that Apple is in negotiations, Musk is building a fabric of 25 billion dollars, and the US government owns 10 percent of the company.

Poles

Computex has been an Intel event for decades. The show takes place in Taipei, the heart of the global semiconductor supply chain, and Intel often uses it to announce products that define each generation of PC computing. The difference in 2026 is that Intel is no longer just a chip designer launching products made from its own fabric. It is a foundry operator that competes with the host country’s most important company for the right to produce chips for other people. Tan’s keynote will be watched not only for what Intel announces about its products but also what those products reveal about the 18A’s readiness to serve foreign customers. Every Panther Lake laptop that runs flawlessly, every Clearwater Forest server that meets its performance claims, every Arc G3 handheld that runs inside its thermal envelope is a data point for Apple, Amazon, and every other company testing to trust Intel with its silicon.

In 2016, Intel was the world’s largest semiconductor company by revenue. By 2024, it had dropped to eighth place, behind Nvidia, TSMC, Samsung, Broadcom, Qualcomm, AMD, and Texas Instruments. Its manufacturing process is down to two generations after TSMC. Its CEO was forced by a board that had lost confidence in the change. The stock reached 18 dollars. Fourteen months later, Intel is on a roll, its customer base is strong including Apple and Musk’s Terafab, and it will be going to Computex with a product in every category for the first time in a decade. The transformation is real, but also not complete: the site loses 2.4 billion dollars per quarter, foreign revenue is 174 million dollars compared to 20 billion for TSMC, and 90 percent of the desktop chip that Intel is previewing at Computex will be made by a competitor that is trying to take it away. The 18A node is Intel’s answer to all those problems. Computex is where it starts to prove that the answer works.