Vultr’s $1B Ohio AI Cluster Uses AMD Chips — Useful Traction, Not a Valuation Fix

What happened

Cloud provider Vultr announced plans to spend more than $1 billion to build an Ohio AI cluster that will use AMD chips, marking another commercial deployment for AMD’s data-center hardware. This is a tangible design-win that complements AMD’s recent Q3/2025 momentum in Data Center revenue and the company’s push with Instinct accelerators and EPYC CPUs. However, the Vultr program is a single large customer and does not resolve the core execution questions flagged in our DeepValue report: CoWoS/HBM packaging constraints, export-control exposure (MI308-related charges), and software/ecosystem gaps relative to Nvidia. At today’s share price, which embeds aggressive AI success, one more design win only modestly reduces execution risk and does not meaningfully increase the margin of safety implied by our DCF anchor. Investors should treat the Vultr announcement as confirmation of go-to-market traction but not as proof that AMD will deliver the sustained hyperscaler scale or margin expansion required to justify current multiples.

Implication

The Vultr announcement validates that customers are willing to deploy AMD-based AI infrastructure and diversifies AMD’s customer set beyond headline hyperscalers. In the near term this may help revenue visibility and support incremental MI300/MI350 shipments, but it does not erase scarcity risks around HBM and CoWoS packaging or sovereign export restrictions that previously produced an ~$800m MI308 charge. Given AMD’s current P/E and FMP DCF gap, investors should not view this single win as a catalyst that meaningfully reduces downside risk. Monitor whether Vultr-scale deployments translate into sustained order flow, improved allocation from foundries/memory suppliers, and customer TCO benchmarks versus Nvidia; those would be the real evidence that justifies re-rating. Absent clear, repeatable hyperscaler-level ramps and resolution of supply/regulatory constraints, the stock remains richly priced and speculative on AI execution.

Thesis delta

The Vultr win nudges the execution ledger slightly more positive by adding another sizeable customer deployment, marginally lowering the probability of zero hyperscaler traction. That said, it does not alter our central valuation concern: the market already prices aggressive AI share gains and margin expansion, so the judgment remains WAIT until durable, multi-hyperscaler ramps and supply/regulatory clarity emerge.