Intel’s new roadmap to catch up with its Asian competitors by 2025 will prove critical in rebuilding the United States’ chip manufacturing capabilities in the increasingly crucial semiconductor industry. On July 26, the California-based company’s CEO Pat Gelsinger unveiled the ambitious roadmap, which aims to regain semiconductor dominance within the next five years. Gelsinger also announced that Amazon and Qualcomm would be Intel’s first two major clients for its new Intel Foundry Services, which aims to build high-end semiconductors designed by other companies and compete with foundry juggernaut Taiwan Semiconductor Manufacturing Company (TSMC). 

• Under its new roadmap, Intel aims to launch a new CPU every year between 2021 and 2025 that is more advanced than the previous year. This would culminate with the launch of the so-called Intel 20A and Intel 18A chips in 2024 and 2025, respectively, which Intel is calling the angstrom era of chips, where angstroms (0.1 nanometers) are used to denote units of measure.
• Intel is also planning to launch its first new transistor architecture since 2011, RibbonFET, along with a slew of other transistor and packaging technologies aimed at keeping up with Moore’s law for its angstrom chips. 
• In addition, Intel recently renamed its current set of chip technologies under development and mass production to Intel 7 and Intel 5 (formerly the ​​Intel 10nm Enhanced SuperFin and Intel 7nm, respectively). The move is aimed at better marketing the U.S. company’s chips against those of Taiwan’s TSMC and South Korea’s Samsung, which both used lower nanometer designations in their chip branding despite having similar transistor densities compared with Intel’s chips.

The roadmap complements the new business strategy Intel announced in March, which is aimed at regaining its leadership position in next-gen processors after falling behind TSMC and Samsung. Intel’s ‘IDM 2.0’ strategy sets out to establish a new for-hire branch of operations to build chips for companies that design but don't manufacture them (like Qualcomm, Amazon, Apple), in addition to designing and manufacturing its own chips. Such an expansion to fabrication services would mark a significant shift in Intel’s business model, which is currently solely focused on integrated device manufacturing (i.e. designing, fabricating and packaging chips internally). To realize this model, however, Intel will need to catch up with other for-hire foundries like TSMC, which does not design any of its chips. 

• Intel has struggled to introduce its 10nm and 7nm chips on the same schedule as its competitors. In 2020, the company announced that its 7nm chip (now named the Intel 5) would not reach full production until late 2022 or early 2023. By comparison, Samsung’s and TSMC’s comparable 5nm chips entered full production in 2020 and are now being used for many advanced processors designed for smartphones like the iPhone. 
• After a 15-year partnership with Intel, Apple stopped using the firm’s processors for its Mac computers last year, due in part to Intel falling behind some of its competitors. 
• To help increase its capacity, Intel announced a $20 billion plan in March to build two new fabrication plants in Arizona. The Wall Street Journal also reported in July that Intel was in talks to acquire GlobalFoundries, the United States’ largest contract chipmaker, for $30 billion. 

The United States’ ability to achieve its strategic goal of reasserting control over the entire semiconductor value chain by boosting domestic chip manufacturing will hinge in part on Intel’s success in implementing its IDM 2.0 model and associated technology roadmap. Like its predecessor, the administration of U.S. President Joe Biden has identified the hollowing out of U.S. chip manufacturing capabilities, particularly for logic chips, as a strategic risk. The rising tech competition with China (where the semiconductor industry is a focal point), as well as the ongoing global chip shortage (which has forced many automakers to reduce their production), has only reinforced Washington’s view that more of the production supply chain needs to be reshored. To that end, the U.S. government is seeking to woo TSMC and Samsung to build and invest in more fabrication plants in the United States. But Intel’s success in its new business model and roadmap will also be essential in ensuring U.S. dominance over the global tech sector in the future.  

• Total U.S. semiconductor manufacturing capacity is at least 25% behind Japan, South Korea, Taiwan and China. The decline is particularly acute when it comes to the most advanced logic chips (i.e. CPUs, GPUs). 
• Unlike other market segments — such as memory and analog chips, where the U.S.-based companies Micron and Texas Instruments are major global players, respectively — the logic chip segment is becoming increasingly specialized and segmented between foundries (i.e. TSMC) and fabless chip design companies (i.e. Nvidia, Qualcomm). Integrated device manufacturers who do both, meanwhile, are far less dominant. 
• While the United States remains a hegemon for logic chip design, the pure-play foundry segment of the industry has been dominated by TSMC, Samsung and Taiwan’s United Microelectronics Corporation (UMC). China’s SMIC is the world’s fourth-largest contract chipmaker that has set out to catch up to TSMC, Samsung and UMC.
• In June, the U.S. Senate passed the U.S. Innovation and Competition Act, which includes $52 billion in support for domestic chip manufacturing as a part of a $250 billion package aimed at countering China’s technological ambitions. 
• TSMC announced in 2020 that it would invest $10-12 billion in a fabrication plant in Arizona. Samsung is also currently in talks to build a $17 billion new plant in the United States.