Better Investment Option: SpaceX or Cerebras?

The IPO market is heating up, with SpaceX expected to become the largest IPO of all time. Meanwhile, chipmaker Cerebras Systems (NASDAQ: CBRS) was off to a hot start, opening at $350 on May 14 after pricing its IPO at $185.

Both SpaceX and Cerebras are working to position themselves to become eventual artificial intelligence (AI) leaders in their respective markets. Let's take a closer look to see which one has the better chance at succeeding.

SpaceX: Looking to put data centers in space

When you think of SpaceX, you probably think of the company's rocket-launching unit. After all, this is how SpaceX got its start, developing a reusable rocket that greatly reduced costs by being able to recover and refurbish a rocket's most expensive components instead of dumping them in the ocean. However, this is the company's smallest segment.

The company's largest segment and the only one that is profitable is its satellite internet service, Starlink. Its number of subscribers has climbed from 2.3 million at the end of 2023 to 10.3 million at the end of Q1 2026. Meanwhile, the segment's revenue jumped 32% to $3.3 billion in Q1, while its operating income rose 15% to $1.2 billion. This is a nice recurring-revenue business, though it is set to face increased competition soon from Amazon Leo, and that's not why the company is set to be valued at nearly $2 trillion.

The company's money-losing AI business is the big reason behind the excitement for the IPO. SpaceX sees this as being a $26.5 trillion opportunity. The company's big pitch is that there is a massive opportunity to create data centers in space, where near-constant sunlight will be able to power them with solar energy.

While that sounds something out of a sci-fi novel, others have said this is possible, including Amazon founder Jeff Bezos, who also started space company Blue Origin. However, there are some obvious constraints. Chips have a finite lifespan of only five to 10 years, and the physics of cooling them in the vacuum of space is a huge technical obstacle that needs to be overcome. There is also the issue of cosmic radiation. This can damage chips and other hardware, even causing bits to flip and corrupt data. In addition, there is the whole issue of assembling these data centers in space and fixing something if it breaks.

Image source: The Motley Fool.

Cerebras: Looking to upend the inference market

While SpaceX is trying to take AI data centers to space, Cerebras is trying to flip the inference market on its head with its wafer-sized engine chips. While training AI models is all about raw compute power, inference tends to be much more memory-bound than compute-bound. By incorporating static random-access memory (SRAM) directly onto its chips instead of packaging them with high-bandwidth memory, Cerebras' chips can perform inference 15 times faster than graphics processing units.

The catch is that SRAM is physically bulky, and Cerebras' chips are about the size of an iPad. They also have special cooling and power management requirements, which is why the company only sells or rents out complete end-to-end systems and not individual chips. This makes its systems pricey, which is why it has largely been a niche solution. However, OpenAI has reportedly made commitments for upward of $20 billion for its servers, marking a major step-change for the company.

The better IPO to bet on

While SpaceX hasn't yet hit the market, the valuation that the company is looking to get, and undoubtedly will, is pretty absurd. CEO Elon Musk is a master of hype and likes to make big promises, so I can see the stock doing well. There isn't much to justify the company's proposed valuation beyond hype, but Musk-backed Tesla has long commanded a premium valuation based more on ideas than fundamentals.

Cerebras' stock also isn't cheap, and it needs to prove it can move beyond being a niche player in the inference market. However, the opportunity in front of it is much more tangible and closer to happening than SpaceX's outer space data centers, which is why I'd prefer it of the two. That said, both are highly speculative stocks at this point.

Geoffrey Seiler has no position in any of the stocks mentioned. The Motley Fool has no position in any of the stocks mentioned. The Motley Fool has a disclosure policy.