ANET vs IBM: AI Revenue Race Has a Clear Winner

Arista Networks (ANET) and IBM (IBM) are both trying to cash in on the artificial intelligence boom, but their recent quarterly revenue reports tell two very different stories. Arista, which sells high-speed networking gear for data centers, has posted eight straight quarters of revenue growth, climbing from $1.8 billion in Q3 2024 to $3.0 billion in Q2 2026. That's a steady upward march, and the company expects even more—forecasting $3.3 billion for Q3 2026.

IBM, on the other hand, has seen its revenue bounce around like a yo-yo. From $15.0 billion in Q3 2024, it dipped to $14.5 billion in Q1 2025, then jumped to $17.0 billion in Q2 2025, only to fall again. In Q2 2026, it brought in $17.2 billion, but that's still below its Q4 2025 peak of $19.7 billion. This volatility is a hallmark of IBM's business, which mixes software, hardware, and consulting.

The reason for the difference? Arista is riding the AI wave directly. Companies are rushing to build out data centers to handle AI workloads, and they need Arista's high-performance switches. IBM, meanwhile, is dealing with a slowdown in its mainframe business—sales dropped 7% year-over-year in Q2—and its consulting division was flat, as customers pause spending.

IBM also cut its full-year 2026 sales forecast, admitting that customers are shifting their budgets toward AI infrastructure providers like Arista. To counter this, IBM is betting on quantum computing, recently acquiring HRL Laboratories to boost its efforts in that field.

For investors, this comparison highlights how different business models can lead to vastly different outcomes in the AI era. Arista's focus on networking for AI data centers is paying off, while IBM's broader, more traditional tech portfolio is struggling to keep pace.

For investors, this revenue comparison is more than just numbers—it's a window into which company is better positioned to profit from the AI boom. Arista's consistent growth shows it's capturing demand from hyperscale data centers and cloud providers, which are spending heavily on AI infrastructure. Its 45% operating margin in Q2 2026 is a testament to its pricing power and operational efficiency, making it a profitable way to play AI.

IBM's volatile revenue and lowered guidance suggest it's losing ground in the AI race, at least for now. Its mainframe business is cyclical, and the current downturn means customers who upgraded recently are done for a while. Consulting, another key segment, is unpredictable as clients delay big projects. This has forced IBM to look to the future—quantum computing—but that's a long-term bet that may not pay off for years.

The market is taking notice. Arista's stock has been on a tear, while IBM's has been relatively flat. If you're looking for a pure-play AI infrastructure stock, Arista looks like the stronger choice. But IBM isn't a lost cause; its quantum computing research and hybrid cloud offerings could provide a boost down the road.

For now, though, the revenue trends suggest that Arista is the better bet for investors who want to ride the AI wave. IBM's challenges highlight the risks of investing in companies with diverse, cyclical businesses that may not fully benefit from the current AI spending spree.