Arista Networks vs. IBM: Comparing Quarterly Revenue Trends Between These Artificial Intelligence Giants

Arista Networks shows consistent quarter-over-quarter revenue growth over the last eight periods, with revenue rising from $1.8B (Q3’24) to $3.0B (Q2’26), while IBM’s revenue is described as more cyclical (e.g., $15.0B in Q3’24 vs $14.5B in Q1’25, then $19.7B in Q4’25). For the quarter ended June 30, 2026, Arista posted a 45% operating margin versus IBM’s 15%. The article attributes Arista’s steady gains to AI data-center infrastructure demand and cites IBM cutting its 2026 full-year sales forecast amid customer spending shifts toward AI buildouts.

Analysis

The cleanest read-through is that AI infrastructure spending is still accruing to the picks-and-shovels layer with the highest operating leverage, and ANET sits closer to the bottleneck than the model vendors. That matters because networking is not discretionary once a cluster build starts, so relative revenue stability should command a premium as long as hyperscaler capex remains elevated. The risk is that this is still a consensus crowded trade: if one or two large cloud buyers pause orders for a single quarter, ANET’s multiple can de-rate faster than its earnings can reset.

IBM’s lumpier profile is less about business quality than about mix: mainframe refreshes, consulting budgets, and software renewals are all tied to different buying cycles, so the stock behaves like a bundle of separate bets rather than a compounder. The market often over-credits “turnaround” stories here; the more important near-term question is whether the company can stabilize organic growth without leaning on buybacks and acquisitions. The quantum angle is effectively out-of-the-money for the next 12-24 months; it can support narrative, but it is not a credible offset to ongoing revenue variability.

Contrarianly, the spread may be partially overdone. ANET’s stability is real, but its forward expectations are now tied to uninterrupted AI spending, while IBM’s volatility creates a lower bar for upside if consulting and infrastructure simply stop deteriorating. The key falsifier for the bullish ANET / bearish IBM relative view is not one quarter of noise; it is either ANET guiding below the current AI-capex cadence or IBM posting two consecutive quarters of organic stabilization with margin expansion.

Key Decisions for Investors

• Long ANET / short IBM, 1:1 dollar-neutral, for the next 1-3 earnings cycles; thesis is that AI-networking demand should continue to outgrow IBM’s enterprise spend mix, with downside if ANET revenue guide flattens or IBM posts two straight quarters of organic stabilization.
• If already long ANET, trail exposure rather than chase: add only on 3-5% pullbacks after hyperscaler capex commentary confirms spending durability; treat any guidance miss as a multiple-risk event, not just an earnings miss.
• Use IBM as a hedge against an AI infrastructure basket: short IBM against a long basket of ANET/NVDA-style AI picks-and-shovels names if you want lower net beta and cleaner exposure to cluster buildout versus legacy IT spend.
• Avoid paying up for IBM on the quantum narrative alone; wait for a visible operating inflection in consulting or infrastructure before considering a long, because the catalyst path is 6-18 months, not weeks.
• Set an alert on ANET’s next quarter revenue guide and hyperscaler capex commentary; if ANET’s guide comes in below the current growth cadence, trim exposure immediately because the market will likely compress the premium before fundamentals roll over.