US Semiconductor Industry Resurgence: Micron's Strategic Moves Amid Global Chip Race

Chip Makers are Betting Big on US

The US offers a secure environment compared to regions with geopolitical tensions for a steady and dependable supply of these essential components. The global semiconductor industry has experienced a significant change over the past few years, driven by disruptions caused by the pandemic, changes in consumption behaviour, and rising geopolitical tensions. As the world increasingly relies on state-of-the-art technologies, securing a stable chip supply has taken centre stage. The onset of the COVID-19 pandemic five years ago turned the global semiconductor industry upside down. Unlike past economic disruptions, this crisis simultaneously hit both supply and demand. Factories in key hubs like China shut down, leading to chaos amongst chip makers. The crisis highlighted the risks of overreliance on geographically concentrated supply chains and pushed companies to diversify and localise production. Notably, the US has emerged as a key player by using government incentives and private investments to alter the global landscape of semiconductors. Recognising the need to secure its semiconductor supply, the US government introduced substantial incentives to encourage domestic semiconductor manufacturing in 2022. This was an attempt to lower costs, create more localised jobs, strengthen the supply chain in the US and counter China's advancements in chip production. The CHIPS and Science Act, passed in 2022, allocated $52.7 billion to boost US semiconductor research, manufacturing, and workforce development. The Act also allocated $1.5 billion to advance wireless technologies, support emerging fields like AI and biotech through new National Science Foundation (NSF) initiatives, and fund regional innovation hubs to boost local economies. Following the Act, companies announced nearly $50 billion in additional investments in American semiconductor manufacturing, which included a $40 billion investment by Micron in memory chip manufacturing and $4.2 billion through a partnership between Qualcomm and GlobalFoundries. In April last year, the US finalised a $6.6 billion subsidy for Taiwan Semiconductor Manufacturing Company (TSMC)'s production facility in Phoenix, Arizona. Moreover, as part of the CHIPS and Science Act, Samsung Electronics was awarded up to $6.4 billion from the US government to build its new complex in Taylor. Several factors make the US a sought-after hub for semiconductor manufacturing. The country offers a secure environment compared to regions with geopolitical tensions for a steady and dependable supply of these essential components. Several AI chip companies, including TSMC and Samsung, have been shifting their focus to building manufacturing facilities in the US. TSMC's Arizona facility has achieved a 4% better yield than the semiconductor giant's manufacturing sites in Taiwan. Notably, states like Arizona and Texas provide ideal climate conditions for chip manufacturing facilities. Access to advanced research institutions, skilled labour, and strong infrastructure supports high-tech manufacturing. Locating production facilities closer to major clients facilitates better collaboration and reduces logistical complexities. For example, TSMC's Arizona plant is expected to serve clients such as NVIDIA, Apple and AMD, enhancing operational efficiency. During the COVID-19 pandemic, the demand for semiconductors used in cars, phones, and gadgets fell sharply as people tightened their budgets. At the same time, chips for cloud computing and home internet equipment saw a surge in demand, driven by the rise of remote work and online activities. The AI revolution has ushered in a new wave of demand for advanced semiconductors. At the core of AI are processors like GPUs, CPUs, and accelerators from NVIDIA, AMD, and Intel, powering models like generative AI. Data centres, transformed by companies like Dell and HPE, provide the infrastructure for AI to scale. Meanwhile, while innovators like ASML and TSMC create smaller, faster, and more efficient chips, Qualcomm and ARM's custom designs extend AI into devices and industries. Chip production shines with innovators like ASML and TSMC, which create smaller, faster, and more efficient chips. Qualcomm and ARM's custom designs extend AI into devices and industries. Meanwhile, companies like Micron and Western Digital provide robust data storage solutions, and design tools from Synopsys and Cadence speed up innovation. These advancements have enabled AI to move from theory to real-world applications. Michael Hurlston, CEO at Synaptics Inc., recently claimed to make the chip design even more efficient and cost-effective by shrinking AI models and optimising software to fit within a minimal memory footprint. In retrospect, while the government's benefits may have helped attract the attention of these giants, they could also have adversely affected their supply chains. With the rise in regulations surrounding the export of AI chips to China, Washington plans to also limit semiconductor shipments to some countries accused of supplying to Beijing in order to Close China's Backdoor Access.

US logic lost its way, can American memory prevail?

With Intel's foundry future in doubt, Micron takes center stage in US Chips Act push Comment The US once led the world in the development and manufacturing of semiconductors and integrated circuits. And the $280 billion US CHIPS and Science Act sought to reestablish the country as a semiconductor manufacturing superpower and alleviate reliance on foreign fabs. Two years after the massive funding bill was passed, Intel's ambitious foundry strategy is in doubt following yet another executive shakeup and calls to spin off the ailing manufacturing unit. Intel may have been central to efforts to establish a domestic alternative to foreign fabs, but it's not the only major US chipmaker to fit the bill. While Intel waits to chart a new course following former-CEO Pat Gelsinger's sudden departure last year, US memory champion Micron is already well on its way to positioning itself to take on its larger rivals SK Hynix and Samsung Electronics. Unlike Intel, which has yet to prove the commercial viability of its mass market process tech, Micron has already demonstrated the ability to produce leading-edge memory modules. And while substantially smaller than its South Korean competitors, it still ranks among the largest memory vendors in the world. With more than $6 billion in US CHIPS Act funding committed to its cause, Micron is pressing forward with an aggressive plan to expand its manufacturing capacity. To date, Micron has pledged upwards of $100 billion to scale up fab capacity across New York, Idaho, and Virginia over the next 20 years. In its latest announcement last week, Micron revealed plans to spend $2.17 billion to expand its memory plant in Manassas, Virginia, a project it now expects to rake in another $275 million in CHIPS funding, up to $70 million in state funds, and even more in tax-breaks. And while Samsung Electronics and SK Hynix have also secured CHIPS act funding, neither companies have disclosed plans to manufacture memory modules at these plants. Samsung's Taylor, Texas plant is slated to produce logic chips based on 4nm and 2nm process technologies. SK Hynix, meanwhile, plans to invest $3.87 billion in a memory packaging facility in West Lafayette, Indiana. Memory has taken on a new importance over the past years on account of, you guessed it: AI. Memory modules take many forms ranging from the NAND flash used in thumb drives and SSDs to the DRAM used in system memory for everything from smartphones to notebooks. Micron controls somewhere between 20 and 25 percent of the DRAM market and just under 12 percent of the NAND flash market. However, in 2025, everyone's eyes are on a specific subset of DRAM called high-bandwidth memory (HBM), which is an essential component to the GPUs and AI accelerators at the heart of the generative AI boom. The higher capacity and the more bandwidth these modules are capable of delivering, the faster these chips can run ever larger models. Here, Micron lags its larger rivals by a wide margin having lost ground due to its decision to skip the HBM3 generation and jump straight to HBM3e which is only now making its way onto the market in products like Nvidia's H200. According to the industry watchers at TrendForce, between 2022 and 2024, Micron's share of the HBM market fell from around 10 percent to 5.1 percent. SK Hynix and Samsung currently sit at 52.5 percent and 42.4 percent, respectively. Despite this, Micron's says it's already sold out of HBM through late 2025, with CEO Sanjay Mehrotra optimistic the company will continue to grow its share in 2025. Even with billions of dollars in CHIPS Act grants at its disposal, Micron remains at the center of a geopolitical storm. Not long ago, China accounted for more than half of the chipmaker's revenues. That figure has slipped precipitously in recent years after Beijing banned Micron products over vague "security risks caused by hidden product problems," in an apparent act of retaliation for US sanctions levied against YMTC. But while Micron was among the first to suffer collateral damage as a result of the US-China trade war, it may not be the last. In December, the Biden administration specifically targeted HBM in its latest salvo of the trade war. Under the new rules, the latest in the US crusade to stall Chinese AI development, HBM vendors will need to obtain special permission from the US Commerce Department to sell HBM in the region. It remains to be seen how big an impact this will have on HBM heavyweights SK Hynix or Samsung. Further complicating the situation is incoming president Donald Trump who is due to make his return to the White House later this month. The President-elect has made it no secret he isn't a fan of the CHIPS act or foreign chipmakers, like TSMC, which have profited from it. During an appearance on the Joe Rogan Experience podcast ahead of the election, Trump condemned the CHIPS Act, calling it "so bad," and attacked Taiwan. "They stole our chip business," Trump said. Shortly after Speaker of the House Mike Johnson suggested the GOP would likely move to repeal the $280 billion funding bill. Johnson quickly walked back the comments after Republican congressman Brandon Williams of New York praised Micron's investments in New York and said he wouldn't support efforts to undermine the investment. Johnson later suggested there could be "legislation to further streamline and improve the primary purpose of the bill -- to eliminate its costly regulations and Green New Deal requirements." In the wake of the comments, the Biden administration has wasted little time formally awarding the CHIPS Act subsidies to recipients. But these subsidies represent only a fraction of the CHIPS and Science Act's value, with the majority coming in the form of a 25 percent tax credit on advanced semiconductor manufacturing facilities. On the flip side, Trump's comments regarding foreign chipmakers and his administration's "America-First" agenda suggest Micron along with other US chip companies could be spared any revisions to the bill once he takes office. The president-elect's agenda includes leveling tariffs on imports from allies and rivals alike with the goal of spurring domestic manufacturing by making it more expensive to build products using foreign goods. Tariffs on South Korean goods could drive US chip companies like AMD and Nvidia, which make extensive use of HBM in their products, to Micron. While this might be to Micron's benefit - assuming it can scale capacity fast enough - the consensus, as we've previously discussed, is that tariffs will simply result in higher prices for end customers. ®