The boom in generative artificial intelligence (AI) has ushered in a golden age for chip designers NVIDIA (NVDA 1.40%) and its investors. Software and services veteran IBM (IBM -0.09%) also benefits from the same trend, but in a completely different way.

Which of these tech titans is the better AI stock in summer 2024? Let’s take a look.

Nvidia’s generative AI bonanza

Before the rise of generative AI, Nvidia wasn’t a particularly exciting story. Two years ago, the company was trying its best to distance itself from cryptocurrency mining, and lockdown-fueled consumer interest in video game gear was waning. Behind the scenes, Nvidia was already shipping tons of AI accelerator chips, though without much fanfare since no one had seen ChatGPT yet. If anything, Nvidia’s involvement in self-driving vehicle systems looked like a promising growth catalyst at the time.

Oh dear, how times have changed.

It is well known that ChatGPT’s artificial brain is based on Nvidia chips, and the company quickly became the preferred supplier of AI accelerator hardware. An Nvidia A100, H100, or L40S accelerator card costs between $8,000 and $30,000. It takes tens of thousands of these products to train a modern Large Language Model (LLM). These chip sales come in large, lucrative batches.

For example, Nvidia’s first-quarter revenues rose 262% year over year in May. Data center revenues, including the AI ​​accelerators mentioned above, accounted for 87% of those revenues, up from 60% in the same period last year and 45% the year before.

As a result, Nvidia’s profits and cash flows are exploding. And so is its share price. Nvidia investors have seen their shares increase tenfold since the two-year low in October 2022.

The stock has obviously risen for good reason, but the enthusiasm for AI seems a bit overheated. Whether you look at the price-to-sales ratio, the price-to-earnings ratio, or the price-to-earnings ratio, Nvidia is trading well above its long-term valuation average.

In short, Nvidia seems overvalued despite its rising financials. I would buy again if the stock were to drop dramatically, but I cashed in some of my Nvidia profits in the spring and right now it’s nothing more than a “hold” idea.

How IBM’s drastic change in strategy is paying off

Unlike Nvidia, Big Blue is no longer a hardware vendor. While its unique mainframe systems still offer some insight into the hardware side of the technology world, the company as a whole has shifted its focus to software and services.

Because IBM focuses exclusively on business-class customers, the market has been slow to recognize its AI prospects. Large customers must thoroughly evaluate new software tools while also obtaining budget approvals from multiple levels of management. This takes time, but the resulting contracts tend to be robust and long-lasting.

This is where the company stands today. Many testing processes and approvals have been completed and IBM is starting to generate serious revenue with its AI solutions under the name Watson.

In July’s second-quarter report, IBM reported $2 billion in generative AI orders, up from zero a year ago as the underlying Watsonx service launched in summer 2023. Management raised its full-year software growth forecast from mid-single digits to high-single digits – an impressive increase given that more than 80% of IBM’s software sales are tied to multi-year contracts. And CEO Arvind Krishna expects that growth rate to climb into double digits in 2025 and beyond.

Big Blue’s AI growth is more moderate than Nvidia’s, but I also expect it to last longer. Sudden bursts in chip sales are one thing. A rising tide of long-term software contracts is another. I’d much rather invest in the slower growth with longer-lasting consequences.

And then there’s the valuation aspect. Compared to Nvidia’s high multiples, IBM’s valuation metrics look like a fire sale:

NVDA PS ratio data by YCharts

Why IBM is a better AI stock than Nvidia today

I still hold some of my old Nvidia shares, but I don’t plan to buy more any time soon. I’ll probably put new money into my IBM position, as this tech giant still looks undervalued.

Wall Street is slow to forgive IBM for a decades-old strategy shift, although I would argue it was exactly the right move at the right time. The next few years should show investors how well the new focus on software, services, cloud computing and AI will pay off in the long run. This is a great AI stock to buy now and hold for decades.

And at the same time, Nvidia seems to be heading for a sharp price correction. Call me again when the stock has fallen at least 30%, or better yet, more.

While Nvidia’s rapid growth is impressive, the company’s valuation is too high. IBM, on the other hand, offers a more stable growth story as its radical strategy shift begins to pay off, making the company an attractive option for new investment.

You heard it here first.