© JasonDoiy / Getty Images
The semiconductor stocks might have endured a bumpier ride in the second half of 2025, but there might be no slowing them down as 2026 kicks off and a new slate of AI trends looks to reignite enthusiasm for many of the stocks that have sold off in recent weeks. Undoubtedly, the latest Santa slump in tech stocks certainly isn’t helping soothe investors’ AI bubble fears, especially as the bears look to take aim at some of the bigger winners in the past two years.
While 2026 might be an even rockier year for the top semiconductor stocks, as new entrants look to get a piece of the economic profits that have been raked in by Nvidia (NASDAQ:NVDA), I do see the potential for up-and-coming AI chip innovators to get a jolt, perhaps at the expense of Jensen Huang’s GPU empire.
Nvidia poised to continue leading the pack higher in the new year
Of course, the big Rubin GPU launch could tilt the tables right back in Nvidia’s favor, especially if agentic AI and enterprise adoption of AI accelerates, paving the way for even hotter AI demand. So, just as rivals, like Advanced Micro Devices (NASDAQ:AMD), look to close the gap with the GPU leader, Nvidia might pull off another big sprint ahead as it looks to continue to show the world that it’s the leader that cannot be caught.
Regardless of what happens for the semis in the new year, as some of the top players pull the curtain on a new slate of innovators at the 2026 Consumer Electronics Show (CES), I do see the tide as being able to lift all boats across the industry, provided that AI demand strengthens further.
If it stays as strong as last year or weakens a bit, however, perhaps it’s Dr. Michael Burry and the bears who will be cashing in as the falling tide drags down all (or perhaps most) boats in the waters. As the great Warren Buffett, who’s now officially retired, put it, “only when the tide goes out do you discover who’s been swimming naked.”
Nvidia’s Groq deal is a huge win as ASICs continue to be a major theme
With Nvidia making a year-end deal, licensing Groq’s impressive AI tech, I do think Jensen Huang and company are already focused on what’s up next with the AI chips.
Not only is Nvidia beefing up its low-latency inference capabilities with the deal, but it’s also bringing some of the brightest minds in the semiconductor world aboard. Undoubtedly, it’s products like Groq’s inference platform and its custom LPUs (Language Processing Units), which pose a potential risk to the GPU kings. With LPUs and GPUs under the same umbrella, it seems like Nvidia is ready to thrive, even in a world where the puck moves towards ASICs (Application Specific Integrated Circuits) in an environment where inference could really begin to boom.
Though it’s too soon to tell if the Groq deal should put the ASIC threat to rest, I do think Nvidia is well-equipped to make up for a relatively lacking year with another potential leg higher. Of course, a 35% gain in a year would be worthy of applause for firms that aren’t named Nvidia.
Either way, the path to a $5 trillion market cap seems set for the new year, as the ASICs and custom silicon (think NPUs, TPUs, and LPUs) look to capture a bigger slice of the unprecedented growth to be had, especially in inference. At the same time, frontier trainers could continue to stockpile Rubin GPUs as the race to AGI becomes even more fierce now that OpenAI has a serious rival on its hands.
The bottom line
In short, odds are the semi plays will sail through AI bubble concerns, high-performance memory bottlenecks, and other industry challenges. As long as the demand is there, the semis still seem to be worth sticking with, especially as some of the major players look to become more active on M&A.