AI Infrastructure Filing Cross-Section: 2026 Q2 Evidence Note

This dataset preserves a frozen, source-dated cross-sectional comparison of six U.S.-listed AI-infrastructure issuers grouped into three role-matched pairs: NVIDIA and AMD in compute accelerators; Arista Networks and Marvell Technology in networking and interconnect; and Super Micro Computer and Dell Technologies in server systems. The research question is: within these three issuer pairs, which differences remain after applying one frozen quarterly filing contract? The dataset contains 12 exact pairwise comparisons covering quarterly revenue growth, gross margin, research-and-development intensity and diluted-share change. Every observation preserves the issuer, fiscal period, period end, SEC acceptance timestamp, accession number, filing URL, comparability tier and evidence-object SHA-256. Descriptive findings: 1. Arista and Marvell revenue growth differed by 1.14 percentage points, while gross margin, R&D intensity and diluted-share change differed by 9.79, 15.58 and 5.46 points. 2. Super Micro and Dell revenue growth differed by 5.62 percentage points and gross margin by 0.28 points, while diluted-share change differed by 19.41 points. The Super Micro diluted-share observation is Tier B. 3. NVIDIA and AMD differed by 55.74 percentage points in revenue growth, 21.20 in gross margin, 14.58 in R&D intensity and 2.79 in diluted-share change. This is a six-issuer descriptive comparison. Fiscal-quarter end dates differ, the paired companies do not have identical product or revenue mixes, and company-wide filings do not isolate AI-only financial results. The dataset is not a statistical factor test, composite score, synchronized calendar-quarter ranking, valuation conclusion or return forecast. The evidence note is frozen from source release ai-infrastructure-cross-section-2026-09-01-v2. The Evidence Note JSON SHA-256 is 490838cbfb3d6187f454009427184d5bfdec9f72efda68f97d60bccee5950546. The source-release SHA-256 is 430406a8dbea1c21e7ca0ce09f829580616f33b0552ce0089e4b977d01cbc809. Canonical research note: https://ahasignals.com/research/ai-infrastructure-filing-cross-section-2026-q2/ AhaSignals is an independent research publisher. Company names and ticker symbols identify the relevant SEC filers only. AhaSignals is not affiliated with, endorsed by or sponsored by these issuers or the SEC. This dataset is provided for research and informational purposes only and is not investment, trading, legal, accounting or tax advice.