Nvidia's China stance and CoreWeave deals: what sales teams should do today
Nvidia's CEO Jensen Huang said there are no talks about selling Blackwell AI chips to Chinese clients. If that sticks, Nvidia could see a quarterly hit of $2-$5 billion.
At the same time, CoreWeave signed multibillion-dollar contracts with Nvidia in Q3, potentially lifting remaining performance obligations from about $30B to $60B. Nvidia remains a key pillar of the "Magnificent Seven," with AI spending and forward guidance still steering market sentiment.
Key facts (fast)
- China freeze on Blackwell: No discussions to sell Blackwell chips to China; potential $2-$5B revenue impact per quarter.
- Market context: Nvidia sits inside the "Magnificent Seven," where AI capex and upbeat guidance continue to move markets.
- CoreWeave demand signal: Q3 multibillion-dollar contracts with Nvidia could push RPO from ~$30B to ~$60B.
Why this matters for sales
- Budget timing: AI budgets aren't slowing; they're shifting. Some buyers will accelerate orders to secure capacity.
- Lead times: Supply could tighten as non-China demand crowds the queue.
- Messaging: Buyers want certainty on access, timelines, and total cost - not hype.
- Risk management: Multi-sourcing (cloud + specialized GPU providers) becomes a standard ask.
Talk tracks you can use
- Capacity assurance: "We can secure capacity through multiple providers and lock timelines up front to reduce project risk."
- Time to value: "Let's match workloads to the fastest path: reserved GPU instances for training, and flexible burst for experiments."
- Financial clarity: "We'll structure phased commitments so finance sees ROI milestones before expansion."
Accounts to prioritize now
- AI-heavy builders: GenAI startups, model labs, and ISVs scaling inference.
- Enterprise AI adopters: Vertical apps in healthcare, finance, industrials, and retail with budgeted pilots moving to production.
- Infra buyers: Cloud service providers, MSPs, and data center operators bundling GPU capacity.
Qualification questions that reveal budget and urgency
- Which AI workloads (training, fine-tuning, inference) are priority this quarter?
- What's your target go-live and what blocks it - capacity, compliance, or cost?
- Do you require Blackwell specifically, or are H100/H200/alternative SKUs acceptable?
- How much flexibility do you want between on-prem, hyperscaler, and specialized GPU clouds?
- What's the approval path for a 12-24 month capacity reservation?
Forecast and pipeline moves
- Pull-ins: Advance deals with buyers who can commit now for guaranteed capacity.
- Stage contracts: Phase 1 (pilot capacity), Phase 2 (scale), with clear exit ramps.
- Offer optionality: Quote GPU-agnostic architectures and multi-cloud setups to cut supply risk.
- Protect margin: Use usage tiers, SLA-based pricing, and surge rates for burst workloads.
Common objections and crisp responses
- "We're worried about supply." "We'll reserve baseline capacity and add burst through a secondary provider to keep milestones intact."
- "Pricing looks volatile." "We can fix rates for reserved capacity and index burst to public spot benchmarks with caps."
- "We need certainty on chips." "We'll architect for portability across compatible GPUs and lock change windows so performance stays consistent."
Signals to watch
- Nvidia earnings and guidance on data center revenue and lead times.
- CoreWeave RPO updates and capacity expansions.
- Hyperscaler capex commentary (GPU commitments, regions coming online).
- Spot pricing and queue times for popular GPU SKUs.
Bottom line
China is a no-go for Blackwell right now, which tightens the funnel elsewhere. Pair capacity guarantees with flexible deployment options, and make your buyers feel safe signing sooner. The teams who promise access and deliver clear ROI milestones will win the budget conversations this quarter.
Resources: Learn more about Blackwell on the official Nvidia page here. If you want fast, practical AI upskilling for sales roles, browse our courses by job here.
Your membership also unlocks:
