Jensen Huang explains why Nvidia will grow an astounding 70% next year

September 10, 2026 · DigitalEdu

Nvidia CEO Jensen Huang projects roughly 70% revenue growth for the coming fiscal year, signaling the company expects its dominance in AI infrastructure to deepen rather than plateau. The forecast matters because Nvidia’s chip supply, software ecosystem, and capital-allocation choices now set the pace for every enterprise building or buying AI capabilities.

What Happened

In recent public remarks, Huang stated that Nvidia anticipates an “astounding” 70% increase in revenue next year, reinforcing a trajectory that has already made the company the primary hardware backbone for generative AI. He emphasized that Nvidia’s expanding web of partnerships — spanning cloud providers, enterprise software vendors, and sovereign AI initiatives — reflects genuine demand rather than circular deal-making designed to inflate numbers. The comments arrive as Nvidia’s newest Blackwell architecture begins shipping and as competitors, cloud hyperscalers, and governments scramble to secure allocation.

Why It Matters for Businesses

For business leaders, the growth projection is less about Nvidia’s stock price and more about what the signal reveals on the ground. A 70% jump implies sustained, massive capital expenditure by the largest cloud platforms and a widening pipeline of enterprise AI deployments that require Nvidia-grade compute. That has three practical consequences. First, lead times for high-end GPUs and DGX systems will remain tight, so procurement cycles need to start months before a project’s launch date. Second, the software stack — CUDA, cuDNN, TensorRT, and the growing suite of Nvidia Inference Microservices — will continue to be the default target for AI model optimization, meaning engineering teams that standardize on it avoid costly rework later. Third, the ecosystem of certified partners, OEM server builders, and managed-service providers will expand, giving mid-market companies more turnkey options but also requiring sharper vendor evaluation to avoid lock-in.

The excerpt also notes Huang’s insistence that Nvidia’s deals are not circular. In practice, this means the revenue surge is being driven by end-user consumption — enterprises training models, running inference at scale, and embedding AI into products — rather than by channel stuffing or round-tripping between friendly parties. For a business owner, that distinction matters: it suggests the underlying demand curve is real, so capacity constraints and pricing pressure are likely to persist.

Competitive and Ecosystem Ripples

The same news cycle highlights adjacent moves that reinforce the broader trend. Nscale, a GPU-cloud specialist, appointed former OpenAI and Meta executive Fidji Simo to its board ahead of a potential IPO, signaling that alternative cloud providers are professionalizing to capture overflow demand that the hyperscalers cannot absorb. Meanwhile, reports of AI agents flooding public services with automated requests illustrate how quickly inference workloads are scaling beyond training, creating a new, persistent baseline of compute consumption. Maven Robotics’ push to displace incumbent robot-deployment contracts shows that the automation wave is moving from pure software into physical workflows, another vector that ultimately feeds GPU demand for simulation, vision, and control models.

These developments are not isolated. They reflect a market where AI automation is already removing 40–60% of repetitive manual task time across industries, automated follow-up sequences are recovering up to 30% of otherwise-lost leads, and AI chatbots are resolving 60–80% of routine customer-support questions without human handoff. Each of those outcomes — drawn from consulting benchmarks and sales-automation studies — runs on infrastructure that traces back to Nvidia’s roadmap.

What To Watch

Business planners should monitor three signals over the next two quarters. First, Blackwell shipment volumes and any guidance updates from Nvidia’s earnings calls will indicate whether the 70% target is on track or if supply-chain bottlenecks are easing. Second, the pace of sovereign AI deals — national governments building domestic compute capacity — will affect global allocation and pricing for commercial buyers. Third, the maturation of alternative cloud providers like Nscale, CoreWeave, and others will determine whether a viable second tier of GPU capacity emerges, giving enterprises negotiating leverage and geographic redundancy.

The Bottom Line

Nvidia’s 70% growth forecast is a leading indicator that AI compute demand will stay white-hot, keeping high-end GPUs scarce and the CUDA ecosystem the de facto standard. Business owners should lock in capacity early, build on Nvidia’s software stack to minimize switching costs, and evaluate emerging GPU-cloud alternatives now — before the next procurement cycle forces a rushed decision.

Source: Original Article

For more on this, see Nscale adds former OpenAI exec Fidji Simo to its board ahead of potential IPO.