Artificial intelligence (AI) chip stocks are having a torrid time on the market in 2025, and that's not surprising. The sector has been hit hard by multiple headwinds of late, such as the viability of the huge sums being spent by tech giants on beefing up AI hardware and the growing concerns about a possible economic downturn in the U.S.
This explains the 10% decline in the PHLX Semiconductor Sector index so far this year. Not surprisingly, two of the sector's key components -- Nvidia (NVDA 0.74%) and Advanced Micro Devices (AMD 0.17%) -- have also witnessed sharp declines in their share prices in 2025. What's worth noting is that both these semiconductor stocks are benefiting from the growing adoption of AI, which is evident from their latest financial results.
But if you have to choose one of these two AI stocks for your portfolio right now, which one should it be? Let's find out.
Nvidia is crushing AMD in the AI chip market
Nvidia dominates the market for AI chips and enjoys a huge lead over AMD. Nvidia sold $36 billion worth of data center graphics cards in the previous quarter alone, which was more than seven times the revenue that AMD generated from data center graphics cards in the entirety of 2024.
NASDAQ: NVDA
Key Data Points
Nvidia is on track to record sequential growth in the data center GPU business once again in the current quarter, as management said on last month's earnings conference call. Its revenue forecast of $43 billion for the current quarter would translate into a year-over-year increase of 65%. That's more than double AMD's estimated growth of 30% for the current quarter to $7.1 billion.
Nvidia enjoys such a big lead in the AI chip market that it is growing significantly faster than AMD despite having a much larger revenue base. Looking ahead, there is a solid chance that it will continue to crush AMD in the market for AI data center GPUs, since the former's latest-generation chips are already ramping up at a remarkable pace.
The company says that its Blackwell AI GPUs delivered $11 billion in revenue in the fourth quarter of fiscal 2025 (which ended on Jan. 26), making it "the fastest product ramp in our company's history, unprecedented in its speed and scale." It said that production of Blackwell processors is in full swing right now and that it's focused on expanding supply quickly to meet the strong customer demand.
AMD, uncharacteristically, didn't give a concrete full-year forecast for its AI GPU sales for this year. That's because the company says it is going through "a little bit of a product transition time frame in the first half of the year." CEO Lisa Su said that this transition could lead to a "flattish" performance in the data center business in the first half of 2025 as compared to the second half of 2024.
The company attributes the transition to its decision to speed up the launch of its MI350 series of AI GPUs to mid-2025, after originally aiming for the second half of the year. It looks like AMD has been forced to release its MI350 processors earlier than expected in response to the terrific start that the latest Blackwell processors have achieved.
That's not surprising since the company announced the MI350 last year to rival Blackwell. AMD's next-generation processors are better than Nvidia's Blackwell processors on paper. However, better specs won't necessarily translate into better sales for two reasons.
First, Nvidia sells entire systems that include software and other parts such as memory and networking components. The company locks customers into its software ecosystem -- called compute unified device architecture (CUDA) -- that a community of developers can use to create programs that eventually allow its customers to cut down on the development time of their generative AI applications.
Second, it has a strong grip over the AI chip supply chain. Both Nvidia and AMD are fabless chipmakers, which means that they simply design their chips. The actual fabrication is done by foundry giant Taiwan Semiconductor Manufacturing, popularly known as TSMC. Nvidia has cornered the majority of TSMC's production lines. According to one report, it has secured more than 70% of TSMC's advanced chip-packaging capacity.
So, even if AMD brings out a powerful chip, it may be constrained by capacity. That's why Nvidia is likely to remain the dominant player in this lucrative market, which is expected to have a compound annual growth rate of 31% through 2031.
The verdict
It should be clear by now that Nvidia is likely to remain the better AI stock of the two companies discussed in this article. And its valuation gives investors yet another reason to buy it over AMD, as the following chart shows.
NVDA PE Ratio data by YCharts; PE = price to earnings.
Nvidia is significantly cheaper than AMD based on trailing earnings multiples. Moreover, their forward earnings multiples are closely matched as well. Considering that Nvidia is growing much faster than AMD and it can sustain its dominance in the AI chip market, it makes sense for investors to consider buying it over AMD right now.
