AMD Has Promised OpenAI and Meta Up to 320 Million Shares for a Penny Each. Here's What That Does to the Share Count.

AMD’s OpenAI and Meta agreements cover up to 12 gigawatts of GPUs and include warrants for up to 160 million shares apiece at $0.01; no warrant shares had vested as of June 27. CEO Lisa Su described revenue per gigawatt as in the “double-digit billions,” implying at least $120 billion in revenue if all 12 gigawatts are delivered, while AMD’s data-center revenue more than doubled year over year to $6.7 billion in Q2 2026. Full warrant exercise could increase the share count by about 16%, and the article cautions that AMD trades at roughly 40 times expected 2027 earnings—or about 47 times if all warrant shares were outstanding.

Analysis

The warrants are best viewed as a contingent customer-acquisition rebate, not immediate dilution: AMD issues shares only if large deployments occur, so the same execution that creates dilution should also create revenue. The missing variable is incremental profit per gigawatt. Management’s EPS-accretion claim is not enough without contract-level margins, delivery costs, and the warrant tranche schedule; revenue growth alone does not establish that the equity giveaway is attractive.

Near term, shipments and customer readiness matter more than the $600 threshold, which the stock has already crossed. Over 1–3 months, monitor shipment commencement, quarterly data-center growth, and any evidence that the commitments are converting into firm orders. Over 6–18 months, execution could validate a credible second source to Nvidia and increase AMD’s bargaining power with AI customers. Conversely, delays in accelerator availability, high-bandwidth memory or advanced packaging, or customer infrastructure readiness would defer both revenue and warrant vesting.

The contrarian point: headline dilution overstates the immediate risk because vesting is conditional, but the market may underprice the valuation risk of paying today for multi-year delivery and earnings. Full vesting would raise the share base materially; buybacks are unlikely to neutralize it at the cited share price. No short solely on dilution: successful delivery could support earnings enough to offset dilution. The thesis weakens if shipments slip or data-center growth decelerates; it strengthens if margins and EPS rise alongside deployment.

Key Decisions for Investors

• Avoid adding AMD solely on the announced customer capacity. For existing positions, consider trimming or using a defined-risk collar rather than an outright short: upside remains substantial if deployments execute, while elevated valuation leaves the stock exposed to delivery delays or earnings disappointment.
• Set an execution alert for AMD’s next reports: verify shipment timing, recognized revenue versus deployment commitments, data-center gross-profit contribution, and warrant vesting by tranche. Treat management’s revenue-per-gigawatt estimate as unvalidated until these metrics appear.
• For a relative-value watch, compare AMD with Nvidia as AMD proves delivery reliability; do not initiate a pair trade on this article alone. A sustained acceleration in AMD data-center revenue with stable or improving profitability would support the challenger thesis; shipment slippage or margin deterioration would falsify it.