Newshot

What we learned from Nvidia’s first earnings call without Jensen Huang

Summary

Nvidia Corp. reported its quarterly results this week, and contrary to the prevailing narrative, the sky didn’t fall. Revenue more than doubled year-over-year. The Vera Rubin chip platform achieved the fastest product ramp in the company’s history, and management raised its fiscal 2028 growth outlook to roughly 75%, up from 70%. By all accounts, it […] The post What we learned from Nvidia’s first earnings call without Jensen Huang appeared first on SiliconANGLE .

Original text(via SiliconANGLE)

Nvidia Corp. reported its quarterly results this week, and contrary to the prevailing narrative, the sky didn’t fall.

Revenue more than doubled year-over-year. The Vera Rubin chip platform achieved the fastest product ramp in the company’s history, and management raised its fiscal 2028 growth outlook to roughly 75%, up from 70%. By all accounts, it was a solid beat and raise.

This was Nvidia’s first earnings call without co-founder Jensen Huang, whose resignation as president and chief executive was announced on Nov. 12. Board Chair Harvey Jones told investors that the resignation followed the “disclosure of personal behavior inconsistent with the company’s code of conduct, standards, and cultural norms.” Hours after Huang publicly said he would be willing to return, Jones said the board “does not intend to reinstate Jensen as CEO” and would not take questions about the circumstances or the board’s ongoing investigation.

Colette Kress, the longtime chief financial officer who is now also serving as interim CEO, set the tone in her first minute: “Our ambition will not change. Our passion for innovation will not change, and our commitment to customers, partners, and shareholders will not change.”

After listening to the call, I believe Nvidia answered the short-term question and left the long-term one open. The company made a convincing case that it can execute without its founder. It has yet to show who decides where it goes next. Here are five things we learned and what they mean for those who buy, budget for or operate artificial intelligence infrastructure.

1. The machine will run without the founder

The fear going into the call was disruption. The quarter closed before Huang’s exit, so the results can’t settle that. The guidance can, and it went up. Hyperscalers brought Vera Rubin into production in about a month, half the time Grace Blackwell required. The roadmap is set through Vera Rubin Ultra next year and Feynman in 2028, and Kress said the supply agreements extend through the end of the decade.

She explained that the past month has been “Business as usual.” Kress added, “We’ve engaged with our customers, suppliers and partners. They’re all 100% behind us.” She also noted that TSMC leadership was at Nvidia’s offices this week.

I have long held that Nvidia’s moat was never the silicon. It’s the accumulated layers of systems, software and supply chain built around it. The past week was the first real test of that view, and so far, it’s holding. A yearly product cadence this complex runs on thousands of engineers and a supply chain planned years in advance.

Asked about talent retention, Kress acknowledged that “there are diverse opinions” among employees while insisting they are “committed to doing their life’s work” at Nvidia. The company’s talent can’t be overstated, and although Huang was the face of the company, the employees are what made it tick.

Earlier this week, Cisco Systems Inc. CEO Chuck Robbins appeared with Brian Sozzi. When asked whether he still trusts Nvidia, Robbins replied immediately, “Absolutely.” He added, “The industry’s bigger than one person. I don’t care how good you are. There are a lot of smart people. There are 42,000 people at Nvidia. So I’m not all that concerned.” Huang may have been the most visible Nvidia employee, but the company is full of smart, driven people. Leadership understands this, and I expect turnover to have minimal impact.

Employee retention is a top-of-mind issue for Nvidia executives. Business Insider reported this week that the company held its first all-hands meeting after Huang. The article noted that Kress announced plans to accelerate pay raises and equity grants, Jones discussed the CEO search and Buck outlined the roadmap. It’s reasonable to believe that Huang’s departure was a shock to employees, but the business goes on, and I expect things to settle down in short order.

For information technology buyers, nothing changes about what you’ve ordered or planned over the next 18 to 24 months. Don’t pause a deployment because of a leadership headline. The advice I gave in August still applies: Welcome Nvidia’s simplification and hedge the dependency.

2. The board wants a diplomat as much as a technologist

One of the more interesting comments on the call came from Jones, not Kress. In describing the CEO search, which will consider both internal and external candidates, he said the board “appreciates that we’re living in a world where advanced silicon technologies and national security dynamics are driving major changes in our marketplace. Our new CEO will need to navigate those waters carefully.”

That was paired with immediate action. On the morning of earnings, the board created a standing national security and export control oversight committee of three directors that will retain outside experts. Kress disclosed that Nvidia is no longer shipping licensed Hopper data center products to China and noted that the U.S. Commerce Department issued new export controls this week, which the company “still has some time to study.”

Huang was a product visionary who also served as the industry’s chief diplomat. The board is now signaling that fluency in Washington is a core requirement, on par with technical vision.

For IT leaders at multinationals, export policy is now a deployment variable. The new Commerce rule is intended to block remote access to advanced chips via data centers in third countries, so the location of your capacity and who can access it are now compliance questions. Bring your compliance team into AI infrastructure planning early.

3. Allocation has become a readiness test

With demand still far ahead of supply, who gets Nvidia systems matters more than what they cost. Kress laid out the criteria more clearly than I’ve heard before. First are land, power and shell, meaning whether a customer’s facility is ready to receive and use the gear in the quarter it ships. Second is capital. Third, for customers that aren’t investment-grade, is proof of off-take, or evidence that someone has committed to use the capacity.

“All are getting the same pricing,” Kress said. “That is super important for us, and we just need to assure that they are ready on their side in terms of being able to accept the compute.”

Those same factors were incorporated into the guidance. Kress said the outlook “considers data center readiness, the strength and durability of end customer demand, and our customers’ access to capital.” The reality is that the AI bottleneck has shifted from the chip to the building, the grid and the balance sheet. Nvidia is serving as the industry’s underwriter, determining who is ready.

The lesson for enterprise buyers is that readiness is the currency. If you want on-premises systems, secure power, cooling and funding before requesting allocation. If you can’t, buy capacity from a provider that has already cleared this hurdle. In my view, this also helps explain why sovereign AI is growing so quickly for Nvidia. Those projects often arrive with land and power already in place.

4. Nvidia’s checkbook is getting pickier

Over the past year, Nvidia has used its balance sheet to seed the ecosystem through guarantees, investments, and commitments to neoclouds and AI labs. Though neoclouds remain important, Kress brought a different tone to this audience. She said Nvidia will “continue to be disciplined and selective” in supporting neoclouds and that it is “rigorously evaluating future investments, commitments, and guarantees” as other sources of capital become available. Asked about vendor financing, she acknowledged that “a lot of things have changed over this last quarter.”

Two notable details emerged during the call. When Bernstein Research analyst Stacy Rasgon pressed on what he read as new disclosure language about neocloud commitments, Kress explained that Nvidia’s guarantees take effect only after a partner has raised its own debt and equity, built the facility and installed the compute. If the project stalls, the guarantee lapses.

“Is it new? No. Did we change anything? No,” she said. Even so, spelling it out clarifies Nvidia’s participation in neoclouds. When asked about uses of capital beyond the share buyback, which she said Nvidia still plans to complete, she put suppliers first: “The top priority is going to be our suppliers.” Prepayments to expand capacity come before customer financing, which will be reserved for “very strong customers.”

This was a critical piece of information for investors. I’ve previously compared vendor-financed neoclouds to the CLECs of the internet era, where the financing organization was left holding the bag. A CFO-led Nvidia appears to see the same risk. The result is a shakeout in which well-capitalized neoclouds with contracted demand grow stronger, while thinly funded ones struggle to secure systems. This should offer some comfort to investors that circular financing has been de-risked.

The flip side is that someone still must fund the buildout outside the hyperscalers. Kress pointed to infrastructure capital partners establishing financing platforms. Whether that money arrives as fast as Nvidia steps back is what to watch for.

My advice to IT buyers is to ask, before signing a multiyear deal with a neocloud, how it’s financed, how much capacity is committed and where it stands in Nvidia’s allocation queue. Those were niche questions a year ago. They should be standard due diligence now.

5. Agents are pulling demand well beyond the GPU

The demand story has shifted from training models to running agents, broadening what customers need from Nvidia. Citing OpenRouter data, Kress said agentic workloads now account for roughly 80% of tokens. Falling token prices are driving more usage, keeping GPU rental rates climbing.

The CPU growth story shouldn’t surprise anyone. In May, I wrote that agentic AI opens a massive new spigot for Nvidia on the CPU side. This quarter, demand outpaced supply. “Agentic AI is driving extraordinary demand for CPUs, far outstripping supply,” Kress said, noting that Nvidia’s Vera CPU is now adopted by CoreWeave, Meta, Nscale and Open AI. Agents spend much of their time calling tools, running code and moving data, which is general-purpose work. A CPU shortage amid a GPU boom shows how quickly the workload is changing.

Nvidia is also filling in the layers above and around the hardware. The agent safety platform pairs OpenShell, which limits what agents can access, with Sentry, which runs on BlueField data processing units and monitors agent behavior in silicon. In my view, the biggest barrier to deploying agents in production is control, not model quality.

The network is part of this, too. Kress said co-packaged optics is in volume production, and Lambda’s deployment, announced at this week’s SC26 conference, is the first large-scale proof point. Nvidia is a formidable networking vendor, though buyers mixing accelerators will still want to weigh a network supplier that isn’t also selling a competing chip.

For IT buyers, plan CPU capacity for agent workloads now, and insist that agent controls reside outside the model, in infrastructure you can audit.

Dark-horse CEO candidates

The question of who the new CEO will be might not be just about Nvidia but about the entire industry. If the board goes internal, Ian Buck seems to be the frontrunner. Externally, the names most bandied about are Advanced Micro Devices Inc. CEO Lisa Su, Arm Ltd. CEO Rene Haas and Google Cloud CEO Thomas Kurian, or we could see Manuvir Das make a comeback and return to the company.

Most industry watchers describe Nvidia as a chip company, a view that is about a decade out of date. Nvidia’s value comes from the full systems it has built around its silicon. Each is a stack tailored to an industry: hardware, libraries, frameworks, pretrained models, simulation and developer tools, all tuned to work together.

The next CEO must treat silicon as a key layer of an open, secure, agentic enterprise stack. Enterprises won’t buy GPUs. They’ll buy outcomes: agents that run safely on governed data across clouds, on-premises and at the edge. Whoever runs Nvidia next must sell that outcome, play well within an ecosystem Nvidia doesn’t fully control, and make security a core part of the platform. Here are three dark-horse candidates to consider:

Matt Garman. The Amazon Web Services Inc. CEO has led the world’s largest cloud through the shift to AI. He knows what it takes to turn infrastructure into consumable services at enormous scale. Garman also lives the open-stack approach: AWS offers model choice, runs its own Trainium chips alongside Nvidia’s, and sells to enterprises that want flexibility rather than lock-in. He would bring operational discipline and a deep feel for the customer. The question is whether Nvidia’s biggest customer-turned-competitor can credibly lead the supplier.

Jeetu Patel, Cisco Systems Inc.’s president and chief product officer, is the best storyteller in enterprise tech after Huang himself. He has spent the past few years rebuilding Cisco’s portfolio around secure AI infrastructure, integrating networking, security and observability into a single platform. That is exactly the “open, secure, agentic” stack Nvidia must deliver. Patel thinks in platforms, moves quickly,and has been personally using AI agents to shape his product views. He also has deep enterprise credibility, which Nvidia needs more than additional credibility with hyperscalers.

Charlie Giancarlo, CEO of Everpure (formerly Pure Storage), may be the most underrated name here. He built networking businesses at Cisco, spent time in private equity, and has repositioned Pure as a data management company. Storage revenue is growing thanks to hyperscaler wins. Giancarlo understands that agentic AI depends on data and that the data layer and the network are where enterprise AI succeeds or fails. He is a systems thinker who has led a platform transition before.

Two final points on the CEO hire: For some reason, there is a tremendous amount of industry chatter that Jensen Huang was the only person at Nvidia with key relationships with suppliers, partners, and other ecosystem providers. I know for a fact that is not true. Nvidia’s leadership, which includes Buck, Deb Shoquist (executive vice president of operations), Nico Caprez (vice president of global AI development), Gilad Shainer and Kevin Deierling (networking leadership), Kress and others, all have deep relationships with the ecosystem.

Also, I wouldn’t expect the board to make a snap decision about its next CEO. The roadmap is baked beyond Feynman, and the company can continue for a couple of years, if needed. This is an important hire; the board should take its time and ensure it’s the right person.

Final thoughts

This call confirmed what many industry watchers, including me, have long believed: The moat was built in layers, and those layers held. Supply is constrained, the roadmap is set and customers are still asking for more than Nvidia can build.

Kress closed the call by saying, “We’ve got this.” On execution, I believe her. Kress is a steady hand right now, and the discipline she brings to capital allocation should make the AI buildout healthier. The bets that created Nvidia’s position, though, were placed years before the market arrived. Post-Jensen, Nvidia will continue to perform as it did; as Robbins put it, no one is bigger than the industry.

My advice to IT buyers is to stick with current plans and closely monitor the CEO search. Since Nvidia’s roadmap is baked through Feynman and beyond, it can ride momentum for the next two years or longer.

Zeus Kerravala is a principal analyst at ZK Research, a division of Kerravala Consulting. He wrote this article for SiliconANGLE.

A message from John Furrier, co-founder of SiliconANGLE:

Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.

  • 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more
  • 11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network

Are you an AWS customer? Support SiliconANGLE financially by buying your AWS services from our Marketplace portal page and links: https://siliconangle.com/aws-marketplace/

Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.

The text above was extracted automatically from the source page and may miss images, tables or code blocks. The original is authoritative.

Published by SiliconANGLE
This site only summarises and aggregates; copyright stays with the original author. Headlines and summaries follow the source — where they differ, the original wins.
Read the original ↗