Elon Musk's xAI built what may be the world's largest AI supercomputer, called Colossus, in just 19 days by linking together 100,000 graphics processing units (GPUs) in Memphis, Tennessee. The speed of construction shocked the tech industry and has triggered a wave of investment interest in the companies supplying the specialized networking equipment that makes such massive data centers possible .

What Makes Colossus Such a Big Deal?

Colossus represents a fundamental shift in how companies approach AI infrastructure. Rather than building data centers gradually over years, xAI demonstrated that with the right technology and execution, a supercomputer of unprecedented scale could be assembled in weeks. The facility connects an enormous number of GPUs, the specialized chips that power AI model training and inference, into a unified computing system capable of processing massive amounts of data simultaneously .

The project didn't stop with Colossus 1. xAI followed up with Colossus 2, which opened in mid-January and was called the first "gigawatt-scale" AI data center, meaning it consumes enough electricity to power a small city. Both facilities are primarily focused on building Grok, xAI's AI system, though the infrastructure may support other Musk-related AI projects as well .

"This was a superhuman feat," said Jensen Huang, CEO of NVIDIA, the company that manufactures the GPUs powering Colossus.

Jensen Huang, CEO at NVIDIA

Which Companies Are Profiting From This AI Infrastructure Boom?

The real investment opportunity lies not in the AI companies themselves, but in what industry analysts call "picks and shovels" plays, the companies that supply the essential equipment and services needed to build these massive data centers. When Colossus linked together 100,000 GPUs in less than three weeks, it required specialized networking technology that's now becoming the new standard in AI infrastructure .

The big tech companies are spending enormous sums on AI infrastructure. In 2025 alone, major firms including Microsoft, Meta, Amazon, and Alphabet collectively spent $337 billion on AI infrastructure, with JPMorgan expecting even higher spending in 2026 . This spending spree is creating a significant opportunity for companies that provide the networking switches, fiber optics, and other equipment that connects thousands of GPUs into functioning supercomputers.

• Celestica (CLS): A provider of high-speed networking switches and equipment for data center hyperscalers, with roughly 75% of revenue coming from their Communications and Cloud business, which is growing at approximately 60% per quarter.
• Astera Labs (ALAB): An optical networking company that has become a high-growth play in the fiber optics and data center connectivity space.
• Arista Networks (ANET): A networking equipment leader focused on data center infrastructure and cloud connectivity solutions.
• Coherent (COHR) and Corning (GLW): Companies specializing in fiber optics technology essential for connecting distributed computing systems.

Celestica, in particular, has emerged as a likely candidate for investors seeking exposure to this trend. The company is expected to grow earnings at close to a 40% annual rate while trading at only about 30 times forward adjusted earnings, a relatively reasonable valuation for a company benefiting from the AI infrastructure spending spree . The company has maintained steady gross margins on its high-end switching equipment, suggesting it has pricing power in a competitive market.

How to Identify AI Infrastructure Investment Opportunities

• Look for "picks and shovels" companies: Focus on firms that supply essential equipment and services to data center operators rather than the AI companies themselves, as these tend to have more stable, predictable revenue streams.
• Track major tech spending patterns: Monitor announcements from Microsoft, Meta, Amazon, and Alphabet about AI infrastructure investments, as these signal which suppliers will see increased demand.
• Evaluate earnings growth and valuation: Compare a company's expected earnings growth rate to its current stock price multiple; companies growing earnings at 40% annually while trading at 30 times earnings may offer better value than higher-priced competitors.
• Assess gross margin stability: Companies maintaining steady or improving gross margins on their products demonstrate pricing power and competitive advantages in their markets.

The broader narrative here is that Elon Musk's ability to build Colossus so quickly has created a moment of inflection in the AI infrastructure market. The technology that made this possible is becoming the new standard, and the companies that supply it are seeing explosive demand. For investors, this represents a more tangible opportunity than betting on whether Grok will eventually surpass ChatGPT or whether xAI will achieve some theoretical "superintelligence breakthrough" .

The real story isn't about Musk's AI ambitions alone; it's about the infrastructure race that his success has accelerated. As companies worldwide rush to build their own AI supercomputers, the demand for the specialized networking equipment, fiber optics, and data center services will only grow. That's where the most immediate and measurable returns for investors may lie.