Broadcom recorded a markedly higher operating margin than Marvell Technology in their most recent fiscal quarters, underscoring widening profitability gaps even as both firms deepen ties to large artificial intelligence customers.
Margins and recent deals
In filings covering the early August quarters, Marvell reported an operating margin of 17% for the period ended Aug. 1, 2026. Broadcom disclosed an operating margin of 54% for the quarter ended Aug. 2, 2026. Both companies have emphasized products and partnerships oriented to AI compute and data-center infrastructure.
Marvell has appointed a new chief financial officer and disclosed a custom hardware agreement with Alphabet-owned Google, positioning the company to supply specialized components and systems to a major cloud provider. Broadcom has rolled out new enterprise cloud programs and completed a hardware collaboration with OpenAI, moves that aim to tie its networking and infrastructure software to the needs of large-scale AI deployments.
| Company | Operating margin (latest quarter) | Notable recent actions |
|---|---|---|
| Marvell Technology | 17% (quarter ended Aug. 1, 2026) | New CFO; custom hardware agreement with Google |
| Broadcom | 54% (quarter ended Aug. 2, 2026) | New enterprise cloud programs; hardware collaboration with OpenAI |
What the numbers indicate
Operating margin — the share of revenue remaining after covering operating costs — is one metric that highlights differing business models and scale. Broadcom’s margin figure reflects a combination of high-margin software and diversified infrastructure businesses that can lift overall profitability. Marvell, a vendor focused on data infrastructure silicon and networking components, reported a lower margin in the most recent quarter even as it continues to pursue large customer agreements.
Investors and industry observers look at revenue trends together with margins to assess which companies are capturing the largest share of spending on AI and cloud infrastructure. The firms did not disclose comparable revenue figures in the summary data, but public filings and company commentary place both firms at the center of a broader technology shift toward specialized hardware for AI workloads.
Implications for the AI supply chain
Both companies’ moves underscore how hyperscalers and AI system builders are driving demand for customized chips and integrated hardware-software solutions. The Google custom hardware agreement with Marvell and Broadcom’s collaboration with OpenAI are representative of a strategy by cloud and AI leaders to co-develop components tuned to performance and efficiency requirements.
- Scale matters: Broadcom’s higher margin suggests greater leverage from software and services layered on top of hardware.
- Customer ties: Deals with Google and OpenAI signal access to high-volume, strategic buyers in AI infrastructure.
- Investor focus: Revenue growth and margin expansion will be closely watched as indicators of which vendors capture recurring demand.
For enterprises and cloud customers, the trend toward bespoke hardware arrangements may yield higher-performing AI systems, but it can also create supplier concentration risks and lengthen procurement cycles.
Outlook and context
Semiconductor and infrastructure suppliers are recalibrating offerings as large AI models drive new patterns of compute, networking and storage. While Marvell emphasizes contracts and organizational changes — including a new chief financial officer — to position itself for growth, Broadcom’s results highlight how combining hardware with higher-margin software and services can materially change the economics of supplying AI infrastructure.
Data cited here are current as of Sept. 8, 2026, from company filings and public disclosures. Market watchers will monitor upcoming quarterly reports and contract announcements to see whether revenue growth begins to narrow the profitability gap or whether Broadcom extends its margin advantage through expanded software and cloud program sales.
