February 21, 2024 is a historic date for Intel, one to remember in the future (even when the first balance sheets need to be drawn up). Under the leadership of Pat Gelsinger, the Santa Clara company ushered in a new era with the official launch of Intel Foundryan initiative expressly designed to meet the needs of player who deal with artificial intelligence. This strategic move, announced by Intel’s number one, is accompanied by a series of news focused on development plans, key partners and some initiatives that look towards environmental sustainability.
Intel Foundry is not a simple rename of IFS
Ever since he took over the helm of Intel, Gelsinger has focused heavily on the importance of exploiting productive force of the company and the capacity of its factories to assist partners in the realization of theirs chip.
Three years ago now, in March 2021, the CEO of the US multinational declared that Intel would produce x86, ARM and RISC-V chips for third parties, under the banner of program IFS (Intel Foundry Services). Two years later, the multigenerational agreement between ARM and Intel for the production of chips aroused great interest. With Intel Foundry, the company now intends to position itself as the second largest semiconductor manufacturer by 2030.
In technological terms, among the production processes described by Gelsinger and “his”, Intel 14Å stands out as the first large-scale manufacturing node to use High-NA EUV lithography scanners, opening up new horizons in terms of miniaturization of transistors. With Intel Foundry, the company wants to lead the way constant innovationwith the introduction of a new node every two years.
The new namewhich inherits what has been done with the previous IFS program and tends to improve it, reflects Intel’s desire to be a protagonist in an era, the one we are living in, in which AI drives innovations.
Diversification and strategic collaborations
With Intel Foundry, the Gelsinger-led company doesn’t intend to limit itself to just one partner or a handful of names. Confirming a solid collaboration with realities such as Synopsys, Cadence, Siemens e Ansysit turned out that Microsoft chose Intel to produce its proprietary chips using Intel’s 18Å process. This partnership is catching the attention of all observers, especially since Microsoft has preferred Intel to NVidia for the manufacture of its new chips.
Intel’s CEO expressed the desire to extend collaboration also to its rival AMD. And this despite the fact that AMD had criticized Intel’s choice to invest in production plants. For its part, AMD has always declared itself satisfied with having veered towards an approach fablessthat is, not to make the chips designed by its engineers itself.
Gelsinger’s position indicates a clear desire to serve the widest range of customers possible, including direct competitors.
Gelsinger also tries to rally competitors to make their chips
Intel Foundry will allow you to make chips for various customers (therefore potentially also AMD) using Intel’s flagship process nodes, but also the full range of all its intellectual properties, including its packaging technology more evolved. An aspect that raises several question marks: how does Intel intend to deal with competitors “equipped” with its own technology? Maybe with better product design?
What is certain is that Intel has explicitly mentioned the names of companies such as AMD, Qualcomm, Google and NVidia that it would like to be “in the game” and, in particular, to rely on its factories. “I want Intel Foundry to be used by everyone. We want to help build NVidia chips, AMD chips, TPU chips for Google, and inference chips for Amazon. We want to help them and provide them with the most powerful, performing and efficient technologies to build their systems. Point“Gelsinger stated bluntly.
The collaboration with Microsoft is Intel 18Å
Confirming the choice of Intel Foundry for the creation of innovative chips under its brand, Microsoft spoke of a “stimulating innovation“. The words of Satya Nadella, CEO of the Redmond company: “we are in the midst of a platform shift that will fundamentally transform productivity for every single organization and for the entire industry“. To support this radical change, he added, “we need a reliable supply of the most advanced, high-performance, high-quality semiconductors. That’s why we’re so excited to work with Intel Foundry and have chosen a chip design that we plan to produce using Intel’s 18Å process“.
Trying to “translate” Nadella’s comment, the agreement between Microsoft and Intel could indicate both the introduction of innovative methods in chip production and new CPU architectures and paradigms. We are almost certainly talking about products that will be manufactured in large volumes and distributed on the market for several years. Production of the first Intel 18Å chips is expected to begin in the second half of 2024.
Sustainability and environmental responsibility
The race for AI cannot exclude maximum attention to environmental and safety aspects sustainability. By 2023, according to Intel leaders, 99% ofelectricity used in the company’s foundries will be of renewable origin. The company aims to achieve 100% renewable energy and zero waste to landfill by 2030. Furthermore, it is proposed to completely eliminate greenhouse gas emissions types 1 and 2 by 2040, with the ultimate goal of reducing type 3 greenhouse gas emissions to around zero by 2050.
Intel Foundry’s historic unveiling and related announcements outline an ambitious vision for the future of semiconductor production. At least at Intel.
Gelsinger’s company aspires to become the driving force in the sector, combining technological innovations, strategic collaborations and strong environmental responsibility.
With the goal of five production nodes in four years almost achieved, the new Intel seems to have really changed pace and is nothing like the old Intel which faltered for several years due to errors on the nodo a 10nmceding its crown to the Taiwanese TSMC.
Image source: Intel.