Shares of semiconductor equipment leader Lam Research (LRCX -1.22%) rallied in September, rising 33.7% on the month, according to data from S&P Global Market Intelligence.

Lam didn't report earnings or make any big company-specific announcements during the month, but other prominent semiconductor and memory companies certainly did.

As a leader in complex etch-and-deposit equipment needed to produce leading-edge artificial intelligence (AI) chips and memory, this picks-and-shovels player rallied along with increasingly bullish forecasts for AI infrastructure investments.

Oracle, Micron, OpenAI, Nvidia, oh my!

The month started off well when sell-side analysts at Morgan Stanley increased their estimates for WFE (wafer front-end equipment) spending in 2026. This was somewhat of a big deal, since this year's spending is already supposed to reach $117 billion, up 14%, which would be a record high.

Earlier in the summer, lithography leader ASML Holdings had cast some doubt as to its growth path next year amid tariff and geopolitical uncertainty. However, the Morgan Stanley analysts predicted an additional 5% increase for 2026, on the back of increased DRAM memory spending and better-than-feared logic chip investments.

However, even Morgan Stanley's increased outlook may turn out to be too modest. Later in the month, truly eye-popping growth forecasts were put out by other chip and AI companies that could lead to an even greater capital expenditure boom.

This included Oracle, which reported a 359% growth in its cloud unit's remaining performance obligations, with much of the increase attributed to a $300 billion contract with OpenAI.

Subsequently, OpenAI announced another blockbuster deal with Nvidia, which agreed to invest up to $100 billion in OpenAI to fund 10 GW of data centers. That's a massive amount of new capacity, and appeared to increase the long-term growth outlook for AI chips, and by extension, the tools that make those chips.

Of course, AI infrastructure will also need tons of memory, especially high-bandwidth memory (HBM). To that end, memory leader Micron Technology reported earnings later in September that came in well above expectations, while management also forecast a significant increase in Micron's capital expenditure plans for next year.

Hand holds a square semiconductor chip.

Image source: Getty Images.

Lam is an extremely high-quality play on chip growth

After September's rally, Lam now trades near all-time highs, and at 34 times earnings, near its highest valuation in a decade.

However, we have never seen a chip boom quite like this AI boom. Additionally, Lam also isn't nearly as expensive as some other AI beneficiaries that may carry more risk.

For instance, while foundries and memory giants have to spend fortunes to expand capacity and grow their earnings, Lam and the rest of the semiconductor equipment stock are very capital-light, enabling them to grow in a highly cash-generative way, affording ample share repurchases and rising dividends.

As such, Lam's stock remains a hold for now, but a top-of-list name to buy on any dips -- as long as the long-term AI trajectory remains intact.