「NVIDIA proxy stock」CoreWeave co-founder interview: AI demand seems to be intensifying every day

Original Title: An Interview with CoreWeave Executives: AI Demand Seems to 'Intensify' Every Day

Original Author: Tae Kim

Translation & Compilation: Peggy, BlockBeats

Editor's Note: This interview offers a window into observing the AI compute cycle: demand has not cooled down following the previous wave of GPU hoarding; instead, it is being further propelled by agents, reasoning, and enterprise-level AI applications.

This article features interviews with Brannin McBee, Co-founder and Chief Development Officer at CoreWeave, and Nick Robbins, Vice President of Corporate Development and Investor Relations, discussing the current state of AI demand and the neocloud market. The core takeaway from CoreWeave executives is straightforward — AI demand seems to intensify in new ways every day, and the real bottleneck is shifting from simply "having GPUs or not" to more complex infrastructure issues: powered data center shells, CPUs, storage, electricians, supply chain execution capabilities, and how much customers are willing to pay for next-generation compute.

CoreWeave's unique position lies in the middle of the AI infrastructure chain. It serves leading clients like OpenAI, Anthropic, Meta, Google, Microsoft, and Nvidia, while also directly sensing demand shifts from research labs, enterprise clients, and hyperscale cloud providers. Therefore, it sees not just "whether GPUs are in short supply," but a structural transformation in AI workloads themselves. With the rise of agentic AI and reasoning models, compute demand is no longer centered solely on GPUs; the importance of CPUs and storage is also growing. Next-generation data center designs must reserve space for Vera CPUs, Vera Rubin servers, and more storage.

This also explains why the competition in AI infrastructure is shifting from simple chip procurement to more comprehensive engineering delivery capabilities. Whoever can secure powered data centers faster, deploy servers, streamline supply chains, and optimize cost per token will be closer to the core of this AI capital expenditure cycle. CoreWeave repeatedly emphasizes being "customer-driven," but behind this is a larger judgment: AI cloud providers are no longer just selling compute power; they are proactively rebuilding the next generation of AI factories based on the roadmaps of their most advanced clients.

For investors and industry observers, the most noteworthy aspect of this interview is not any single data point, but the direction of changing AI infrastructure demand: GPUs remain important, but bottlenecks are spreading; Nvidia remains central, but CPUs, HBM, storage, and data center power capabilities are becoming new variables; AI demand is still growing, but future success may depend on who can deliver complex infrastructure consistently, stably, and at scale.

The following is the original text:

CoreWeave is considered an innovative early market leader in the neocloud space.

It is the only cloud provider to achieve the highest "Platinum Rating" from AI research firm SemiAnalysis. Founded in 2017, CoreWeave provides large-scale GPU compute power to startups and large enterprises.

Key Context recently interviewed CoreWeave Co-founder and Chief Development Officer Brannin McBee, and Vice President of Corporate Development and Investor Relations Nick Robbins, to discuss the state of AI demand and the neocloud market.

Below are edited highlights from the conversation:

AI Demand Continues to Intensify

Tae: When did the wave of demand for agentic AI begin to explode?

Brannin: We saw the real beginning in the fourth quarter of last year. At that time, we were having engineering-level discussions with customers about products they expected to bring to market in the first quarter of this year.

This has always been a very important perspective for us in viewing customer demand. We have a deeply interconnected engineering relationship with our customers. It is this very relationship that allows us to see trends ahead of time, rather than reacting passively after changes occur.

If I look at it from the product perspective of the AI market, I would say the first quarter was the moment of a huge inflection point for inference and AI consumption, and this acceleration continues even now.

Tae: What is the current state of AI demand? Compared to a few months ago, has there been absolutely no sign of slowdown in recent weeks?

Nick: It seems to intensify in new ways every day.

Tae: Please talk about the rising demand for CPUs relative to GPUs in the wave of agentic AI. Will we see rows of Vera CPU racks deployed alongside Nvidia GPU servers?

Brannin: CoreWeave has been running CPUs since 2023. We have always had a full cloud product. So the question is not whether we are just starting to add CPUs, but what do customers actually need? Is this demand increasing in a relative sense? The answer is, very clearly, yes.

As agents and reasoning capabilities truly emerge in models, storage demand is also rising compared to previous generations. I believe this trend will continue.

Nick: The answer to your question is yes. You will absolutely see a large number of Vera CPUs deployed alongside a large number of Vera Rubin servers. Last year, we actually fundamentally redesigned our base data center plans to make room for more storage and more CPUs to be deployed right next to the GPUs.

We did this because we are in a very unique position within the entire ecosystem. We are the only independent cloud provider serving all the users of the most advanced technology. No other independent AI cloud provider can say that Anthropic, OpenAI, Meta, Google, Microsoft, Nvidia, etc., are all their customers.

This creates a beneficial flywheel, or positive feedback loop, for our business: we can understand where customers are taking the technology and plan accordingly.

Bottleneck is No Longer Just GPUs

Tae: In the future, will you primarily use Nvidia Vera CPUs?

Nick: It depends on the specific workload. Our actions are driven by customer demand. We do expect to be an early and significant adopter of Vera CPUs, as we have disclosed. Currently, our fleet is still primarily AMD, but over time, this could change based on customer needs. Customer interest in Vera CPUs is very strong.

Brannin: This is also a good reminder of how our contracts work. As you know, over 98% of our revenue is contract-driven. We are not guessing what infrastructure customers want. Customers tell us very specifically what configurations they need. Everything is customer-driven. The customer defines what we build.

Tae: Talk about the competitive landscape. How do you go to market and compete against neoclouds like SpaceX, Nebius, Oracle, and hyperscalers like Azure, AWS, Google?

Brannin: On differentiation, I prefer to look at it from the perspective of third-party validation. Excluding China, nine out of the top ten global AI labs use our platform. SemiAnalysis has consistently rated our performance as being in a class of its own at the highest level. I don't think we get this GPU allocation because of personal connections with Jensen.

This shows that suppliers have deep confidence in our execution track record and engineering capabilities, believing we can best represent their product capabilities on a global scale.

Nick: We are able to win hyperscaler customers because we are very good at execution. We can build these systems incredibly fast, and they run extremely well. We win research lab customers because we provide the highest-performing version of the technology and the best performance per token.

We win enterprise customers because the infrastructure genuinely works well, and we have built an excellent, best-in-class orchestration layer, which is also a source of recognition like the Platinum rating.

But increasingly importantly, among AI cloud providers, we have built the most mature layer of capabilities covering inference and development tools, helping enterprises put AI into production.

This means we are building and delivering products that ultimately help enterprises with lower technological maturity convert data into models, and then into agents that can run internally, and we can cross-sell CoreWeave cloud services in the process.

Tae: What is the current bottleneck? Is it the powered data center shell? The GPU? The electrician?

Brannin: It's powered shells. More specifically, the components inside these shells. You mentioning electricians specifically is exactly right. It's a complex area.

But importantly, we already have 49 such sites online and running. We are not pinning our hopes on just one or two sites. We have done this 49 times.

This is a very deep execution track record.

It also means we have accumulated a great deal of knowledge, knowing how to handle supply chain issues, and which suppliers are suitable or unsuitable to work with in this supply chain.

Editor's Note: Powered shells refer to the data center building itself, excluding the actual computational server hardware.

Tae: What can you share about the cost and shortage of HBM memory? How are you dealing with it? Do customers bear the cost of price increases?

Nick: The answer is yes. Our business model is designed to lock in the price we charge customers for GPUs simultaneously as we sign the GPU purchase order and determine the cost we will pay. More broadly, the server price, which obviously includes the HBM cost.

This is how we insulate ourselves from the impact of daily price fluctuations.

If our component costs rise in the next transaction, we will reflect that cost in the price we believe we can charge customers, thereby protecting our margins. We are well-protected in passing these costs on to customers. This is something we are watching very closely.

Currently, procuring components is not the biggest bottleneck. The biggest bottleneck is the powered shell. But at some point in the future, this answer could swing back and forth.

Tae: How do you expect the deployment ramp for Vera Rubin to unfold? What will the second half of this year look like?

Nick: We were clearly the first company globally to power up and fully validate the VR, specifically the Vera Rubin cabinet. We did the same with GB200 and GB300 last year. I expect VR to start appearing later this year.

I expect the truly large-scale, very strong deployment ramp to span the entire year of 2027. The cadence will be similar to GB: GB started appearing in 2025, but the truly massive ramp-up actually spanned the whole of 2026. Meaning, quite a bit was deployed at the end of last year, but this year is the year for the truly large-scale deployment of GB.

I expect a very similar cadence for VR over the next 12 to 18 months.