Banking giant Goldman Sachs believes that the hundreds of billions of dollars being allocated to the AI buildout will not slow down anytime soon.
In a new Bloomberg interview, Goldman Sachs Asset Management’s co-head of public tech investing, Brook Dane, says the bank’s baseline forecast points to acceleration in AI spending for a longer period of time.
According to Dane, the durability of the multi-year AI CapEx indicates investors should have exposure to companies directly benefiting from the spending spree, primarily picks-and-shovels plays.
“And so our baseline is you want to still be involved in the semiconductor names, the networking names, things like that that are the direct beneficiaries of the spending… Investors need exposure to the picks and shovels, and this build-up because it’s still early and still happening. The margin structures right now are high relative to history.
But we think durable because of the fact that you know that this relationship between your ability to put compute in the ground and generate revenue for the model companies and for the cloud vendors is still very tight. So you know there isn’t a big discontinuity event happening from either the supplier or demand side that would upset the margin structures.”
Dane specifically mentions one stock that he thinks could outperform the broader market even though it is already close to trading at record highs.
“One of them that we think right now the market hasn’t fully gotten its arms around is what’s happening with Marvell. And last year, there was a lot of controversy around Marvell and their ability to hold share at Amazon and some of the other big vendors. What we think the market’s missing is around all of the ASIC business. Marvell has this great opportunity to attach XPU, attach products and grow substantially. So we think that’s a specific company where there’s a real opportunity around the upside.”
Marvell (MRVL) is a semiconductor company that provides high-performance silicon solutions for computing, networking, security, and storage. Its XPU is designed to handle intense AI workloads, such as model training and inference, more efficiently than standard CPUs.
As of Thursday’s close, MRVL is trading at $119.93, down about 6% from its all-time high of $127.48, which it hit in January of 2025.
Featured image: Getty Images via Unsplash
Disclaimer: Opinions expressed at CapitalAI Daily are not investment advice. Investors should do their own due diligence before making any decisions involving securities, cryptocurrencies, or digital assets. Your transfers and trades are at your own risk, and any losses you may incur are your responsibility. CapitalAI Daily does not recommend the buying or selling of any assets, nor is CapitalAI Daily an investment advisor. See our Editorial Standards and Terms of Use.