It was already a tumultuous year for the US semiconductor industry.
The semiconductor industry plays a significant role in the “AI race” where the US decided to win. Therefore, this context is worth paying attention to:
Let’s take a look at what happened since the beginning of this year.
May
Last minute flip
May 7: Just a week before the “Artificial Intelligence Spreading Framework” was established, the Trump administration plans to go a different path. According to multiple media outlets, including Axios and Bloomberg, the administration should begin on May 15th, and instead will not enforce the restrictions when working on its own framework.
April
Humanity doubles support for chip export restrictions
April 30: Support has doubled to limit human-created chip exports, including imposing further restrictions on Tier 2 countries and dedicating resources to enforcement, including some tweaks to the AI diffusion framework. An Nvidia spokesman shot back by saying, “American companies should focus on innovation rather than telling tall stories that large, heavy and sensitive electronics are somehow smuggled in “baby bumps” or “with live lobsters.”
Planning layoffs on Intel
April 22: Prior to the first quarter revenue call, Intel said it plans to fire more than 21,000 employees. This layoff was intended to streamline what CEO Rip Bhutan has been saying for a long time that Intel has had to do and helped rebuild the company’s engineering focus.
The Trump administration has also restricted chip exports
April 15: NVIDIA’s H20 AI chip was hit by export licensing requirements and was disclosed in SEC filing. The company added that it expects a $5.5 billion fee associated with this new requirement in the first quarter of fiscal year 2026. The H20 is the most advanced AI chip that Nvidia can somehow export to China. TSMC and Intel reported similar expenses in the same week.
TechCrunch Events
Berkeley, California
|
June 5th
Book now
Nvidia appears to be talking about how to get out of chip exports as well
April 9: Nvidia CEO Jensen Huang was spotted attending a dinner at Donald Trump’s Mar-A-Lago resort. At the time, NPR reported that when Huang agreed to invest in US AI data centers, it would not spare Nvidia’s H20 AI chips from export restrictions.
Suspicion of contract between Intel and TSMC
April 3: Intel and TSMC are said to have reached a tentative agreement to launch a joint chipmaking venture. The joint venture will operate Intel’s chipmaking facility, and TSMC will acquire a 20% stake in the new venture. Both companies declined to comment or confirm. If this transaction doesn’t come true, this is probably a neat preview of the potential transactions in this industry.
Intel has revolved its non-core assets and announced a new initiative
April 1: CEO Lip-Bu Tan got his job straight away. Just a few weeks after he joined Intel, the company announced that it would spin off its non-core assets so it could focus on it. He also said the company will launch new products, including custom semiconductors for its customers.
march
Intel gives a name to the new CEO
March 12: Intel announced that industry veteran and former board member Lip-Bu Tan will return to the company on March 18th as CEO. At the time of his appointment, Tan said Intel would become an “engineering-centric company” under leadership.
February
Intel’s Ohio Chip Plant is delayed again
February 28th: Intel was supposed to start operating its first chip manufacturing plant in Ohio this year. Instead, the company slowed down the construction of its second factory in February. Currently, the $28 billion semiconductor project will not be built until 2030 and may not be open until 2031.
Senators are calling for more chip export restrictions
February 3: US senators, including Elizabeth Warren (D-Mass) and Josh Hawley (R-MO), wrote a letter to Howard Lutnick, who was directed to the candidate for Secretary of Commerce, urging the Trump administration to further restrict AI chip exports. This letter was specifically mentioned in the Nvidia H20 AI chip used in training Deepseek’s R1 “inference” model.
January
Deepseek releases an open “inference” model
January 27th: Chinese AI startup Deepseek caused quite a stir in Silicon Valley when it released an open version of the “inference” model of the R1. This isn’t specifically semiconductor news, but pure alarms in the AI and semiconductor industry have sparked the release of Deepseek, which continues to have a ripple effect on the chip industry.
Presidential Order on Joe Biden’s Chip Export
January 13: With only one week left, former President Joe Biden proposed to wipe out new export restrictions on US-made AI chips. The order created a three-layer structure that determines how many US chips can be exported to each country. Under this proposal, tier 1 countries were not facing restrictions. Tier 2 countries have had chip purchase restrictions for the first time. There were additional restrictions in the three tier countries.
Anthropic’s Dario Amodei places emphasis on chip export restrictions
January 6: Humanity co-founder and CEO Dario Amodei co-written OP-Ed in the Wall Street Journal, which supports existing AI chip export restrictions and points out them as the reason behind the US. He also called for President Donald Trump to impose further restrictions and close the loopholes that allowed Chinese AI companies to continue to acquire these chips.