Cloud Wars Live with Bob Evans

In today’s Cloud Wars AI Minute, I break down Nscale’s rapid shift from Bitcoin mining to AI data centers and why its enormous contracted backlog raises important questions about the risks and realities of scaling AI infrastructure.
Highlights
0:05 — So, one of the things that I wanted to bring up, because as I was doing my research, I ran across Nscale. Now, Nscale is probably the biggest company that none of you even know who they are, and it's basically a data center hyperscaler.
1:48 — So, this company that literally has $103.4 billion in the ability for them to go and contract is only making $140.6 million as of H1 2026 revenue. So, that means they're way off of their revenue targets against what they're actually having to invest. If you take like 85% of what this company is funded by, it's funded by Anthropic, which is about $44.6 billion, and Microsoft at $43.88 billion, and that's primarily the workloads that they have, and that's why their H1 net loss is about $1.02 billion.
3:03 — And they've got these hardware contracts to be able to go do it. But those contracts need hardware, and so it's really interesting to see all the movement in this space. Because is this company overleveraged? We're not quite sure.
3:21 — In the last 24 hours, we saw Microsoft walk away from a contract, and then Anthropic doubling down on a contract with this particular vendor. So, it's going to become very interesting to see what happens as we watch this market expand, as the demand for data center capacity continues to grow, and companies like Nscale have the ability to sign these really big deals and these big contracts.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I discuss why Oracle must continue delivering innovation and financial growth while reassuring investors about its debt and Larry Ellison's leadership presence.
Highlights
00:15 — On the one hand, Oracle has got this incredible momentum in its cloud and AI business, and a staggering volume of RPO backlog, but at the same time, there's lingering and ongoing concern about the level of debt that Oracle has had to take on in order to fund the buildout of these AI data centers across the country and around the world.
00:49 — It's got to continue its innovation blitz across all the things it's doing in OCI, with its AI database, its Fusion Agentic applications, and across the full Oracle product lineup. And we'll see that this week at AI World. Now, also, the momentum it's got from its Q1, which ended August 31st: it had OCI revenue up 121% to $7.4 billion, and the guidance for that is very strong going forward.
01:51 — That's not revenue yet, so that's one of the largest backlogs any company's ever had. Microsoft's right now is $678 billion, so just slightly larger than Oracle's. But very few companies have ever had to face a backlog like this and the responsibility to build that out.
02:42 — What Oracle has to do is continue to show product innovation and financial growth. It'll be having a financial analyst meeting later this week at AI World, so we'll see what comes out of that. Now, the other big point here with AI World is: Will Larry Ellison be there? And if so, in what capacity, and what level of visibility will he have?
03:09 —In early September, for the first time ever, Ellison was not present at all on the [earnings] call. He didn't make any statement. He didn't answer any questions. Oracle went public in 1986, and until recently, Ellison was not only on, but he was the heart of those earnings calls, those engagements with the public. He used those to give Oracle's vision of the future and where Oracle was headed, where the technology was headed.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I examine how SAP's AI-first strategy helped drive 26% cloud revenue growth and why Christian Klein believes the momentum is accelerating.
Highlights
00:01 — We had some interesting numbers from SAP last week that I talked about: Q3 cloud revenue up 26% to seven and a half billion, and its cloud backlog up 28% to $27 billion. I wanted to chat a little bit today about Christian Klein, the SAP CEO. He said there are three reasons why AI has been so important in helping to generate these sorts of numbers.
00:37 — Here's what he said: There's certainly better technology capable now with AI to help accelerate these migrations, but the big thing, he said, is getting better business outcomes and being able to focus on those. The second thing he said: AI and the enhancements there make the suite, the SAP Business Suite, relative to best-of-breed. He said it makes it much more attractive.
01:12 — And he said the Business AI platform overall has been a huge driver for customers, and especially in this area now around the Business Data Cloud. He said that is the thorny issue that's going to unleash the potential for AI to come in and really help drive these better business outcomes.
02:28 — One of the reasons, I think, is because SAP has been not only going all in on AI for the last few years, but in addition to that, I think Klein has always kept uppermost in his mind: What are the outcomes that are most important for our customers? The business outcomes most important for them, and are we putting the right amount of emphasis on helping deliver those?
03:04 — The important thing here about that is that the Business Data Cloud is unifying all this stuff that SAP has been putting together over the last five decades. How companies work, where their data is, what their processes are like, the industry-specific nature of what they have, and I think this Business Data Cloud, with the various ideas that Klein put forward, I think that's going to be the most important.

Visit Cloud Wars for more.

In today's Cloud Wars AI Minute, I examine how GitHub Copilot squads and Claude Code agent teams are changing the way AI agents communicate, collaborate, and solve complex problems.
Highlights
00:12— And today, we want to talk about AI coding squads. We're seeing this change about how we're seeing agents being able to go and work on different, really complex projects. And we used to have this one assistant that would go out and everything else, but now we're starting to see teams of agents going and doing things, and we see different configurations of this.
00:54 — So you could have a front-end developer, a back-end developer, an auditor, all of these different things, and they're just basically different personas. And then the beauty of this is where we talk about something like Claude Code subagents, which has been out for a while. This is more the concept of a task agent, where you have a parent that is going to spin off subagents underneath it.
01:34 — Well, when we get into GitHub Copilot squads, we actually see this changing, where what we're having is that each of these different personas are created as members of a team, and the teams work together to be able to come with, like, a common shared team memory and allow them to be able to communicate back and forth, to be able to assign tasks and collaborate and communicate with one another.
02:02 — And this makes it where you end up with a more robust way of actually solving problems, where you're not having to orchestrate it all through a single, a single agent that's supposed to make the decisions on everything. Now you have different personas that can weigh in their opinions and play their role. You can kind of think of it just the same way you would have a human team.
03:15 — But we're going to start seeing more and more of the different AI creators or the different providers of agents starting to take this tack, where they're going to have more of a squad or a team of agents versus having a parent-child relationship, like we've seen in the past. So, something to keep your eye on and pay attention to as you continue to move forward in your AI journey itself.

Visit Cloud Wars for more.

In this episode of Cloud Wars Live, Bob Evans speaks with G2 Krishnamoorthy, Vice President of AWS Databases, about how agentic AI is reshaping enterprise data strategies and accelerating business innovation. Krishnamoorthy explains why AI-ready data foundations, knowledge graphs, and purpose-built databases are critical to unlocking the potential of intelligent agents. The conversation explores how AWS customers are dramatically reducing processing times, why forward deployed engineering is transforming technology partnerships, and how security, interoperability, and infrastructure economics will influence the next generation of enterprise AI applications and database innovation.
AWS's AI Data Strategy
The Big Themes:
AI-Ready Data Foundations Drive Innovation: As enterprises embrace agentic AI, establishing a strong data foundation is becoming essential to unlocking sustainable business value. G2 Krishnamoorthy compares agentic AI to a powerful car and the data foundation to the road it needs to travel. Regardless of how sophisticated AI becomes, organizations must ensure their data is accessible, well-structured, and ready to support innovation. Importantly, these investments extend beyond AI applications.
Forward Deployed Engineers Accelerate Transformation: AWS is investing in forward deployed engineering (FDE) to help customers translate emerging technologies into production-ready business solutions. Rather than simply providing database products and technical guidance, AWS engineers work directly within customer environments to understand specific challenges, demonstrate solutions, and support implementation. Krishnamoorthy explains that this approach can span cloud migration, data integration, cataloging, contextual knowledge development, and AI-enabled process redesign.
Agentic AI Requires Stronger Security Governance: As AI agents gain the ability to perform actions on behalf of users, security and governance become increasingly critical. Krishnamoorthy identifies two major priorities for AWS: ensuring trusted agentic operations and strengthening defenses against emerging vulnerabilities. AI agents must operate within appropriate authorization, authentication, and auditing controls while maintaining the identity and permissions of the users they represent. AWS is working with Amazon Bedrock and AgentCore capabilities to establish guardrails around agentic experiences.
The Big Quote: “You need to have the elasticity. You really need to actually be able be truly effortless in terms of managing capacity. You really want to reduce the cost of operations dramatically."
More from G2 Krishnamoorthy:
Follow G2 Krishnamoorthy on LinkedIn or learn about knowledge graphs and AgentCore Identity.

Visit Cloud Wars for more.