Artificial intelligence (AI) continues to be the biggest driving theme in the market today, and there is little reason to think that this won't continue. Demand for both AI infrastructure and services appears insatiable, and it still looks as if we're still in the very early innings of this trend.
Against this backdrop, let's look at the top two AI stocks to buy right now.
Image source: Getty Images.
1. Nvidia
NASDAQ: NVDA
Key Data Points
Nvidia (NVDA 1.21%) is the king of AI infrastructure, and the company's recent acquisitions have made it even stronger. It's best known for its graphics processing units (GPUs), which provide the processing power for the majority of AI workloads. GPUs are particularly dominant in large language model (LLM) training, where the company's CUDA software platform adds to its wide moat. Nearly all foundational AI code was written on CUDA, and that code only works natively with Nvidia's chips.
With its recent acquisition of SchedMD, Nvidia has only expanded its software moat. SchedMD is the developer of Slurm, an open-source software platform that helps manage GPUs by determining which tasks they perform and when. With this acquisition, Nvidia now controls the primary orchestration platform for AI chips. While it says it will keep Slurm open-source, its control over the platform will allow it to more tightly integrate it with CUDA to offer an even more seamless experience.
Then, on Christmas Eve, the company acquired top talent from Groq and signed a licensing agreement with the company for its technology. The deal essentially gives Nvidia access to Groq's language processing units (LPUs), which are specialized chips designed specifically for AI inference. Demand for AI inference processing is eventually expected to become larger than demand for training, so this deal can be viewed as Nvidia playing both offense and defense to get ahead of that shift.
Overall, Nvidia remains the company best positioned to profit from the continued AI buildout, and its recent acquisitions only strengthen its position. The stock is also reasonably valued, trading at a forward price-to-earnings (P/E) ratio of about 25, based on analysts' estimates for its fiscal 2027, which will begin in late January 2026.
2. Alphabet
NASDAQ: GOOGL
Key Data Points
Alphabet (GOOGL 1.67%) (GOOG 1.63%) is arguably the best positioned AI company because it is the only one not reliant on Nvidia.
While other companies are working on designing their own custom AI accelerator chips, Alphabet's Tensor Processing Units (TPUs) are now in their seventh generation and have been battle-tested by running Google's workloads for more than a decade. Those years of experience aren't something that its competitors can easily emulate.
As such, the company enjoys a big structural cost advantage in both AI training (having trained its world-class AI model Gemini) and inference relative to companies that rely largely on Nvidia for chips. Its TPUs have proven so good that Anthropic signed a large deal with Alphabet to deploy TPUs to power its AI workloads. Morgan Stanley estimates that for every 500,000 TPUs that Alphabet rents out, it generates around $13 billion in revenue.
Alphabet's other advantage over its cloud computing competitors is that it owns a world-class LLM that rivals OpenAI's ChatGPT. First, this lets it capture more cloud computing revenue by offering its own model. Second, it can monetize its AI model more readily by integrating it into its products, including Google Search, its Android operating systems, YouTube, Google Maps, Gmail, and its workplace productivity tools. With lower costs for training and inference, as well as more platforms upon which it can deploy and monetize its models, Alphabet holds significant advantages over OpenAI and other LLM developers.
As the most vertically integrated AI company, Alphabet is in a strong position, and its advantages should only widen in the coming years. Meanwhile, the stock is attractively valued, trading at a forward P/E of 28.
