In context: Intel’s recent challenges are undoubtedly familiar to anyone who has paid even the slightest attention to the semiconductor market. These issues were a key reason the company recently appointed semiconductor industry veteran Lip-Bu Tan as its new CEO. Some even speculated that Intel might spin off its chip-making foundry business.
At the company’s recent Direct Connect conference, however, the new CEO made it clear that not only does Intel have no intention of doing so, but it is actually doubling down to reclaim its world-class foundry heritage.
The company used the event to update the market on its forthcoming 18A process – which Intel’s products group will use for its Panther Lake line of mobile CPUs set to launch later this year – and highlighted numerous technological advancements in both its core process technologies and advanced packaging capabilities.
While Intel didn’t announce any new foundry clients at Direct Connect, as some had hoped, it made it clear that there is a renewed focus on customer service and on delivering what potential clients want.
To that end, Intel introduced several new variations of its previously announced 18A and 14A process nodes. Both 18A-P and 18A-PT use the same GAA (Gate All Around) transistors and PowerVia backside power delivery technology as 18A, along with additional enhancements.
The most significant change is a broader range of supported voltage levels, which Intel Foundry’s new Chief Global Operations Officer, Dr. Naga Chandrasekaran, explained will enable applications beyond high-performance computing. In addition, 18A-P offers improved support for skew corner variations, referring to the range of acceptable performance levels in different areas of a chip.
Together, these and other updates provide up to an 8% improvement in performance per watt, while maintaining design rule compatibility with chips originally intended for 18A. In practical terms, most third-party designs targeting the 18A process will likely adopt 18A-P, which is expected to be available in 2026.
The 18A-PT process, scheduled for 2028, incorporates TSVs (Through Silicon Vias), making it well-suited for applications like base dies in complex chiplet designs.
For 14A, Intel shared additional technical details and outlined several key improvements. Notably, many of these changes reflect the company’s expanded support for the broader semiconductor ecosystem. In fact, CEO Lip-Bu Tan spent much of the first hour of the event engaging with partners from the EDA (Electronic Design Automation) industry, who develop chip design software.
Historically, Intel relied primarily on proprietary EDA tools, but that began to change with 18A and expanded significantly with 14A. As a result, the CEOs of Synopsys, Cadence, and Siemens spoke with Lip-Bu about optimizing their tools for Intel’s latest foundry offerings.
While this may seem like a minor shift, it marks a meaningful change that should make it easier for third parties to design chips for Intel’s manufacturing facilities. More importantly, it reflects a broader cultural shift in how Intel approaches its manufacturing business.
Intel also shared expected performance comparisons between 14A and 18A, and the numbers were impressive: a 15 – 20% improvement in performance per watt, a 1.3x increase in transistor density, and a 25 – 35% reduction in power consumption.
Intel also shared expected performance comparisons between 14A and 18A, and the numbers were impressive: a 15 – 20% improvement in performance per watt, a 1.3x increase in transistor density, and a 25 – 35% reduction in power consumption. The company emphasized that 14A – expected to enter production in 2027 – builds on learnings from 18A development, including second-generation RibbonFET transistors and enhanced backside power delivery.
Intel also announced a 14A-E variant, designed for even lower power consumption, making it ideal for next-generation mobile chip designs.
One of Intel Foundry’s most widely recognized advantages over other chipmakers lies in its advanced packaging technologies. These innovations are aimed at enabling more sophisticated chiplets that combine multiple dies – often produced by different vendors and on different process nodes.
As physical limits constrain the size of individual chips, the industry is increasingly turning to chiplet system designs to continue Moore’s Law – style improvements (1.5 – 2x every two years). In fact, the most advanced AI processors today are massive chiplets that integrate multiple dies and rely on cutting-edge packaging to function effectively.
A surprise guest at Intel Foundry #DirectConnect – Chip, an autonomous robot dog specializing in sensing, paired with AI/ML insights that help keep our fabs efficient.
Intel Foundry’s fleet of robots uses thermal, acoustic and image processing to cover 16 different inspections. pic.twitter.com/MVRU87wUIr
– Intel News (@intelnews) April 29, 2025
In this context, Intel unveiled new versions of both its EMIB chip-to-chip interconnect and its Foveros chip-stacking technologies at Direct Connect. These updates include lower-cost, higher-speed, and varied interconnect options. Intel also emphasized that these packaging technologies can be used with chip dies built at other fabs, at Intel fabs, or any combination of the two.
Despite the dominance of TSMC in advanced-node semiconductor manufacturing, Intel still faces a challenging path. However, based on the announcements made at Direct Connect, the company is clearly listening more closely to the needs of its potential customers and partners. It is also returning to its core technological strengths and working to assemble a more complete set of building blocks that could make its foundry offerings more appealing.
Given the growing awareness of semiconductor supply chain importance, ongoing geopolitical tensions, and Intel’s unique position as the only remaining US-based producer of advanced logic chips, it is critical that the company overcomes these challenges and pushes its foundry business forward.
The current uncertainty around tariffs certainly isn’t helping. But with strong leadership and solid execution, Intel appears to have as good a chance as ever to reclaim a position of prominence in semiconductor manufacturing. The road won’t be easy, and much is at stake, but the company does seem to be headed in the right direction.
Bob O’Donnell is the founder and chief analyst of TECHnalysis Research, LLC a technology consulting firm that provides strategic consulting and market research services to the technology industry and professional financial community. You can follow him on X @bobodtech
Source link