The biggest driving force in the market continues to be artificial intelligence (AI), which is why now can be a great time to pick up shares of some of the top companies in the space. Let's look at three AI stocks to buy now.
Nvidia
The AI infrastructure boom shows no signs of slowing down, as evidenced by Taiwan Semiconductor Manufacturing making the decision to greatly ramp up its capital expenditures this year. Fabs (chip manufacturing facilities) need to be running near full capacity to be profitable, so this was not a step taken lightly. The foundry (known as TSMC for short) needed evidence that AI chip demand was not a passing fad, and it was convinced that this demand is here to stay.
NASDAQ: NVDA
Key Data Points
That's great news for Nvidia (NVDA +1.53%), which is a large TSMC customer. Its graphics processing units (GPUs) are the primary chips used to run AI workloads, and its main manufacturing partner just signaled that AI infrastructure demand is set to be a long-term secular trend. With about a 90% market share in the GPU space, Nvidia continues to be one of the companies best positioned to benefit from the continued AI data center construction.
Image source: Getty Images
Broadcom
While Nvidia's dominance in AI infrastructure is unquestioned, Broadcom (AVGO 1.61%) is also making great strides by helping hyperscalers (owners of large data centers) make custom AI chips to handle some of their AI workloads. The company is a leader in ASICs (application-specific integrated circuits), providing the building blocks to help customers' designs become physical chips that can be manufactured at scale.
ASICs are preprogrammed chips that lack the adaptability of GPUs, but they tend to deliver strong performance for the tasks they're designed for and are more energy-efficient. This becomes increasingly important with AI inference, which is an ongoing cost.
NASDAQ: AVGO
Key Data Points
Broadcom helped Alphabet develop its highly successful tensor processing units (TPUs), and it's set to benefit from increased TPU deployments as Alphabet is starting to let large customers order the custom chips to be deployed through Google Cloud. Anthropic has already placed a $21 billion order with Broadcom for TPUs to be delivered this year.
Several other companies are working with Broadcom to design their own custom AI chips, including OpenAI. Citigroup analysts have projected that Broadcom's AI revenue could increase fivefold in the next two years from just over $20 billion this past fiscal year to $100 billion. That's huge given that the company just recorded a bit under $64 billion in revenue for fiscal 2025.
Micron Technology
For GPUs and AI ASICs to perform at their best, they need high bandwidth memory (HBM), a specialized form of dynamic random access memory (DRAM) that stores data and helps AI chips quickly retrieve and transmit it. Given that AI model training and inference are all about speed, and HBM helps improve processing speeds, demand is through the roof. The shift to agentic AI, where AI agents need to make real-time decisions, only makes HBM more important.
NASDAQ: MU
Key Data Points
This is a huge boon to Micron Technology (MU +0.52%), which is one of the big three DRAM makers along with Samsung and SK Hynix. HBM is much more complex to manufacture than regular DRAM and requires three to four times the wafer capacity. However, the unit economics are much better, which is leading these companies to focus production on HBM, causing a supply shortage throughout the entire DRAM market.
This is leading to skyrocketing memory prices and supplies being sold out in advance. While Micron is working to increase HBM capacity, the supply-demand situation is likely to remain extremely favorable for quite a long time, making it a top AI stock to own over the coming years.
