Goldman Sachs Predicts AI Infrastructure Spending Could Hit More Than $1 Trillion in 2027. 3 AI Stocks to Buy.

Spending on artificial intelligence (AI) infrastructure is booming, with related capital expenditure (capex) expected to surpass $700 billion this year. While that is a staggering amount, Goldman Sachs said it doesn't see spending slowing next year, projecting it could reach between $920 billion and $1.4 trillion.

At $1.25 trillion, that would be about 3% of gross domestic product (GDP), which would still be below some past technology booms, including the U.S. and U.K. railroad build-outs in the late 1800s.

Let's look at three AI stocks to own as infrastructure capex continues to soar.

Image source: Getty Images.

1. Nvidia

Nvidia (NASDAQ: NVDA) has been the biggest beneficiary of early AI infrastructure spending, and it remains well-positioned to be one of the biggest winners. The stock is also attractively valued, trading at a forward P/E of just 16 times fiscal 2028 (ending January 2028) analyst estimates.

The company has been seeing extraordinary revenue growth, including 85% last quarter, as its graphics processing units (GPUs) remain the primary chip used to train AI models. It still has a wide moat in this area, as most early foundational AI code was written on its CUDA software platform.

Nvidia is also looking to be an important player in inference and agentic AI. Its "acquisition" of Groq gave it language processing units (LPUs), which are ideal for the decode phase of inference, while it has also introduced ARM-based CPUs, which can play a pivotal role in agentic AI.

Overall, Nvidia is set to remain one of the biggest winners from this AI infrastructure spending.

2. AMD

While Advanced Micro Devices (NASDAQ: AMD) lost to Nvidia in AI model training, AMD is much better positioned for the emerging trends of inference and agentic AI. Inference is much less technical than AI model training, and AMD's ROCm software has improved significantly over the past few years, loosening CUDA's stranglehold in this area.

The company's chiplet GPU design can also pack in more memory, which is important because inference constraints are generally memory-related rather than compute power. AMD already has two $100 billion GPU commitments related to inference, which should be a huge growth driver for the company in the coming year.

In addition to its opportunity in inference, AMD is also a leader in data center CPUs. This market is set to see explosive growth with the rise of agentic AI, which needs more CPUs to help manage the sequential reasoning required to power AI agents. As a result, the GPU-to-CPU ratio is expected to shift from 8:1 for training to 1:1 with agentic AI, with AMD predicting this could be a $120 billion market in a few years.

With the company riding two of the biggest trends in AI infrastructure, it looks like a big winner from this increasing spending.

3. Micron

The memory market is booming, especially DRAM (dynamic random access memory), as GPUs and other AI chips need to be packaged with a special type of DRAM called high-bandwidth memory (HBM) to achieve optimal performance. The shift toward inference, which tends to be more memory-bound than compute-bound, only escalates the need for more memory.

As one of the big three DRAM makers, along with Korean companies SK Hynix and Samsung, Micron (NASDAQ: MU) has been riding higher DRAM prices to record revenue, gross margins, and profits.

The entire DRAM market remains undersupplied, and with data center capex expected to continue surging, that dynamic is unlikely to change anytime soon, even as new capacity comes on.

Micron stock is inexpensive, trading at a forward P/E of 9 times fiscal 2027 (ending July 2027) estimates, and with the company starting to sign some longer-term contracts, it should help lessen some of the cyclicality in its business. That makes it a nice stock to own as AI capex continues to soar.

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