3 AI Chip Stocks to Buy in September

September is historically a volatile month for stocks, especially ahead of midterm elections. However, the market also tends to perform very well once the elections are over. That combination is something investors should be able to take advantage of this month.

With semiconductor stocks being whipsawed around recently, let's look at three great stocks to pick up in September that look set to outperform over the next year and beyond.

Image source: Getty Images.

Nvidia

Nvidia (NASDAQ: NVDA) remains arguably one of the most attractive stocks in the market today. The chipmaker continues to grow like gangbusters, while its valuation remains cheap. The stock trades at a forward P/E of only 14.5 times analyst estimates for fiscal 2028 (ending January 2028), while it recently projected that revenue would grow by 70% that year.

The company's graphics processing units (GPUs) are the primary chips used for AI workloads, especially for AI model training, as most foundational AI code was written on its CUDA software platform and optimized for its chips. However, it also has strong offerings for inference and agentic AI through its language processing units (LPUs), which it got through its acquisition of Groq, and the Arm-based central processing units (CPUs) that it has developed.

Together with a world-class networking portfolio, Nvidia is now a complete AI infrastructure company that can offer out-of-the-box dedicated AI solutions to its customers. That should be its next big growth driver.

Advanced Micro Devices

While it can't compete with Nvidia in the large language model (LLM) training market, Advanced Micro Devices (NASDAQ: AMD) is much better positioned for inference. The chiplet design of its GPUs can be packaged with more memory, which is one of the main limiting factors for inference, while it also recently made its own inference chip acquisition in Talaas. Talaas designs chips with specific AI models embedded, helping reduce latency. However, AMD is not just attacking the market on one front; it also formed a partnership with Cerebras to use its chips for the memory-intensive decode phase of inference and its less expensive GPUs for the more compute-heavy pre-fill phase.

At the same time, AMD has a big opportunity with agentic AI as the leader in server CPUs. The company has long been taking share from Intel, and its newest high-core CPUs have been designed specifically to handle agentic AI. With the ratio of GPUs to CPUs going from 8-to-1 for AI model training to 1-to-1 for agentic AI, AMD sees this becoming a $220 billion market by 2030.

Trading down more than 15% off its highs, AMD looks like a solid stock to pick up this month, as it is riding two of the most powerful trends in AI infrastructure.

Broadcom

Broadcom (NASDAQ: AVGO) has quietly become one of the best bargain growth stories in the AI infrastructure space. The company is a leader in ASIC technology, helping customers turn their chip designs into custom AI accelerators. It is also one of the premier data center networking providers.

The company helped Alphabet develop its Tensor Processing Units (TPUs), and that business is set to be its biggest driver moving forward. Not only will it continue to collect billions of dollars from Alphabet, but selling TPUs to Anthropic will become its biggest revenue generator. At the same time, it also helped OpenAI develop its new, widely praised Jalapeño chip, and Meta Platforms is another large customer. As a result of these new chip designs, Broadcom expects its AI revenue to double to $115 billion in fiscal 2027 and then double again to $230 billion in fiscal 2028.

Despite its projected soaring revenue growth, the stock trades at a forward P/E of only 19 times fiscal 2027 analyst estimates and about 12 times the fiscal 2028 consensus, making it look like a huge bargain.

Geoffrey Seiler has positions in Advanced Micro Devices, Alphabet, and Broadcom. The Motley Fool has positions in and recommends Advanced Micro Devices, Alphabet, Arm Holdings, Broadcom, Intel, and Nvidia. The Motley Fool has a disclosure policy.