ARM vs. Marvell Technology: What Revenue Trends for These Artificial Intelligence Companies Tell Investors

Marvell reported revenue of $2.7 billion for Q2 2026, compared with ARM's $1.3 billion, and showed a more consistent upward quarterly trend across the periods presented. Marvell raised its FY2028 revenue outlook to $20 billion from $18 billion and set a FY2031 sales target of $70 billion to $90 billion, citing demand for its custom AI-related chips. ARM revenue also grew year over year but followed a more uneven pattern as it shifts toward data-center CPU chip production.

Analysis

Revenue momentum favors MRVL, but the comparison is not an apples-to-apples measure of competitive strength: Arm’s licensing and royalty mix and Marvell’s product-heavy sales can produce very different revenue scales and economics. Seasonal sequential dips at Arm do not by themselves show weakening demand, while MRVL’s rising sales do not establish that AI-related growth converts to durable margins or cash flow.

The key second-order risk for MRVL is that custom silicon and optical demand may be concentrated among a small set of large customers and tied to their infrastructure spending. Expanded capacity raises the cost of an eventual demand mismatch. Broadcom and hyperscalers’ internal chip programs are relevant competitive checks. For Arm, selling its own data-center CPU designs could enlarge reported sales but also change its economics and put it closer to direct competition with established CPU suppliers; verify product revenue, customer adoption, and margins before treating it as a growth inflection.

Near term, MRVL’s raised FY2028 outlook may sustain momentum, but the FY2031 sales ambition is a long-dated target, not evidence of booked demand. Over 1–3 months, watch guidance conversion, customer concentration, and gross-margin and cash-flow trends; over 6–18 months, watch whether capacity expansion and Arm’s product strategy translate into profitable growth. A revenue-only relative trade risks paying for MRVL’s growth while overlooking valuation and business-model differences.

Key Decisions for Investors

• Do not initiate a blind long-MRVL/short-ARM pair on these revenue series alone. Relative valuation, comparable segment growth, margins, and cash conversion are missing; treat MRVL’s stronger sales trajectory as a catalyst watch rather than proof of superior risk-adjusted value.
• For MRVL, consider adding only on a pullback or after results confirm that raised guidance is translating into delivered revenue without material margin or cash-flow deterioration. Reassess if guidance is cut, major-customer demand weakens, or capacity expansion runs ahead of orders.
• For ARM, wait for evidence that data-center CPU offerings are gaining customers and contributing meaningful, profitable revenue. A rise in reported product sales without margin and adoption data would not validate the strategic shift.
• Verify the reporting-period alignment and underlying filings before modeling sequential growth. The thesis weakens if comparable-period growth narrows for MRVL, or if Arm’s data-center contribution accelerates while MRVL’s outlook or profitability disappoints.