After a stretch of underperformance, NVIDIA’s stock could come alive after earnings, upcoming industry events and the expected launch of new AI models by companies using its Blackwell chips.

The biggest chip companies have seen their stocks underperform this year, and that dynamic doesn't make sense to one Cantor Fitzgerald analyst.

Cantor's C.J. Muse wrote that while makers of networking, memory, storage and semiconductor capital equipment have witnessed surging stock performances this year due to tight supply-and-demand imbalances, Nvidia and Broadcom shares have largely missed out on the party, trading in the red so far this year and underperforming the broader PHLX Semiconductor Index SOX.

"To put it simply, you can't have it both ways," Muse said in a Tuesday note.

What he means is that either Nvidia and Broadcom should be getting the same love as other AI infrastructure stocks, or the market is wrong about a supercycle. If the latter is true, highflying stocks like Micron Technology and Lam Research would be overvalued.

But Muse leans toward the former explanation, arguing that shares of the AI chip giants are undervalued - especially Nvidia. He sees potential catalysts ahead that could lift the stock 61.6% to his $300 price target.

A spark could come from Nvidia's earnings report later this month, assuming the company can post a meaningful beat relative to consensus estimates, Muse said. That said, Nvidia's stock has seen a generally tame reaction to recent earnings beats. After its latest report in November, the stock initially climbed 5.1% before closing down 3.2%.

Or perhaps the jolt could come from the first Blackwell-powered large language models, expected to become available in March. For one, the development would demonstrate that scaling laws are still intact, meaning that AI models can improve with more data and compute power.

And if Blackwell-trained AI models from OpenAI and other model makers perform beyond expectations, that would validate Nvidia's product roadmap, Muse said, and show that its full-stack solution will be difficult to beat.

He noted recent fears in the market over Google's custom tensor processing units, which the tech giants uses to train its Gemini AI models. Google doesn't sell its TPUs to third-party customers, but some worry that the company might start to do so, which would eat into Nvidia's massive market share. That's why it's so important for Nvidia to show that its technology is critical to the development of cutting-edge models.

In addition, Nvidia's annual GPU Technology Conference in March and the industrywide Computex trade show in June could feature developments that power the stock higher, Muse said. Nvidia has a chance at both events to show that the cost of training and running AI models is continuing to go down, Muse said. This in turn would enable Nvidia's technology to have a wider impact across the economy as the use of AI increases.

There's also a chart-oriented catalyst to watch. If Nvidia's stock can rise to between $195 and $200 over a multiday period, Muse said he thinks that would be enough to "set the stage for a change in momentum" and fuel buying linked to the fear of missing out. Nvidia's stock down 2% on Wednesday.