Nvidia (NVDA +1.72%) was the gold standard of AI investing up until a few weeks ago. Now, that notion is being challenged by an unlikely source: Alphabet (GOOG 2.31%) (GOOGL 2.51%). Alphabet isn't what investors think of when they think about AI computing hardware, but recent developments may be pushing Alphabet in that direction.
Furthermore, Alphabet has other AI endeavors that could push it over the edge to make it a better stock pick than Nvidia. However, Nvidia still expects monster growth that Alphabet can only dream of. Which company makes for the better investment in 2026? I think it depends on your appetite for risk.
Image source: Getty Images.
Alphabet and Nvidia are starting to compete
Traditionally, Alphabet and Nvidia haven't been competitors. Nvidia supplied its graphics processing units (GPUs) that can be used to train artificial intelligence models and run other computing workloads in a cloud computing environment, and Alphabet bought them. However, Alphabet started developing an alternative to GPUs a while back: tensor processing units (TPUs).
In a head-to-head match where these computing units have to run multiple workload types, Nvidia's GPUs would win hands down. However, if you had a very specific workload and that's the only workload these computing units saw during their lifespan, TPUs would win and have the benefit of being lower-priced.

NASDAQ: NVDA
Key Data Points
This may make TPUs attractive for some use cases, but not all of them. That's why Alphabet only used TPUs for internal use or made them available to rent through Google Cloud. But that may be changing. Recently, Alphabet and Meta Platforms were in the headlines on reports that Alphabet was considering selling TPUs to Meta. This would mark the first time that TPUs were made available for outside purchase, and investors started to wonder if this could be a major crack in Nvidia's business model.
However, I don't think Nvidia investors have anything to worry about. Nvidia's GPUs are still the top choice for many companies, even if they're expensive. In Nvidia's results for its fiscal 2026 third quarter, CEO Jensen Huang noted that the company is "sold out" of cloud GPUs. That quote provides some important context when comparing Alphabet and Nvidia, as it may indicate that Meta had to turn to alternative computing sources because it couldn't obtain the computing power it desired from Nvidia.

NASDAQ: GOOGL
Key Data Points
If that's the case, then Nvidia investors have nothing to worry about. But if Alphabet starts selling its TPUs openly (there haven't been any confirmed reports that it's actually going to do this), and Nvidia starts losing business to it, then Nvidia investments could be in trouble.
I think Nvidia will be fine over the long term. After all, there is a massive computing market available.
The data center buildout will reach extreme levels by 2030
By 2030, Nvidia projects that global data center capital expenditures will reach $3 trillion to $4 trillion annually. That's a monster number, and if it pans out, there will be plenty of room for companies like Nvidia and Alphabet to thrive selling computing hardware. While that may seem like an outlandish projection, investors must realize that Nvidia knows more than the average investor because it can see orders already placed years in advance. Nvidia could be wrong on the actual dollar figure, but it's likely right in the trend.
This indicates monster growth for Nvidia, and Wall Street analyst projections back this up. For fiscal year 2027 (ending January 2027), they expect Nvidia to deliver 48% revenue growth to $316 billion. Contrast that with Alphabet's projection of 14% growth to $454 billion, and it's clear that Nvidia is expected to have a far more successful 2026 than Alphabet is.
Furthermore, Alphabet's stock is actually more expensive than Nvidia's when you value the stocks using next year's projected earnings.
GOOG PE Ratio (Forward 1y) data by YCharts. PE = price-to-earnings.
I think this makes Nvidia the better stock pick for 2026, but if Nvidia starts to lose significant market share to another competitor or Alphabet, don't be surprised if Alphabet is the better performer. Furthermore, if the AI hyperscalers don't announce record-setting data center capital expenditures for 2027 after setting records each of the previous years, Nvidia's stock may struggle.
It all boils down to whether the AI data center buildout continues. If it does, Nvidia will be a top pick. If it doesn't, Alphabet is a far better investment option.






