Cloud Wars Live with Bob Evans

In today's Cloud Wars Minute, I break down a major move by the U.S. government and AWS.Highlights00:02 —The U.S. government is being given access to $1 billion in AWS credits for use until the close of 2028 to help bolster federal cloud adoption and IT modernization efforts.00:38 — We're seeing companies in the Cloud Wars Top 10 and beyond offering the U.S. federal government massive discounts to secure contracts under the administration's AI Action Plan. Companies like DocuSign, Adobe, and Oracle are reportedly providing discounts of up to 70 to 75%. OpenAI is offering ChatGPT at a discounted rate to federal users.01:27 — Secondly, there's a shift we're witnessing. Instead of clashing with the government, these companies are offering significant discounts because they're valuing the opportunity upfront. AI that aligns with the government and private industry will undoubtedly be safer and more reliable.  Visit Cloud Wars for more.

Wei Hu is the Senior Vice President, High Availability Technologies, at Oracle. In today's Cloud Wars Live, Hu sits down with Bob Evans for a wide-ranging discussion on Oracle’s globally distributed database, AI-native workloads, and how Oracle is helping businesses meet data sovereignty requirements while delivering high performance, elasticity, and always-on availability across regions.Where AI Meets Your DataThe Big Themes:Globally Distributed Exadata Database on Exascale: Oracle’s Globally Distributed Exadata Database on Exascale Infrastructure delivers something few cloud providers can: high performance, high availability, and full compliance. Built on Oracle’s powerful Exadata platform, this architecture removes the traditional need to purchase or manage hardware. Organizations can start small and elastically scale compute and memory across multiple regions.Agentic AI and Vector Search at Enterprise Scale: Oracle’s database innovation is designed for real-world AI demands, especially agentic AI. AI agents need massive compute, consistent availability, and extremely fast access to live business data. Oracle’s globally distributed architecture supports in-memory vector indexes for lightning-fast retrieval augmented generation (RAG), making AI more responsive and effective. Additionally, Oracle keeps AI close to the data — eliminating stale data issues and ensuring compliance.Built for a Sovereign Cloud World: Data residency and sovereignty are no longer optional, they’re legal imperatives. Countries around the world are implementing strict rules on where data must be stored, how it can be accessed, and who can process it. Oracle addresses these challenges with policy-driven data distribution, allowing customers to define how and where data lives. Whether it’s for compliance with India’s payment data regulations or Europe’s GDPR, Oracle enables precise control without requiring app changes or replication of the full stack.The Big Quote: “The other thing that's interesting about agentic AI is that it's very dynamic. The work comes in, the demands comes in, like a tidal wave. Then it goes away, right, then a little, then when it comes again, there's another tidal wave. So, what you really want to do is have an infrastructure that is elastic, that can scale up and down depending on the demand."More from Wei Hu and Oracle:Connect with Wei Hu on LinkedIn and learn more about Globally Distributed Exadata Database on Exascale Infrastructure.* Sponsored Podcast * Visit Cloud Wars for more.

In today's Cloud Wars Minute, I explored why the groundbreaking Oracle–Google Cloud partnership around Gemini AI marks a bold new era for multi-cloud collaboration.Highlights00:14 — It's a brilliant move by Oracle and Google Cloud to form a partnership where Gemini AI is now available on the Oracle Cloud. Gemini on OCI is pretty wild. Both Oracle and Google compete viciously against each other. At the same time, they're both customer-centric enough to say: " ... in this other area, we're going to create new and significant value for customers."01:52 — Too often, I think tech vendors get into this mindset of: "Well, they're a competitor, so I'm not going to do anything with them — even if that might make life better for customers."I think those days are over. That's why I feel this is truly transformative.02:08 — Both executives from Google Cloud and Oracle spoke in great detail about how this partnership will make life easier for customers. They also talked about how this reflects their new philosophy: "We've got to be open. We've got to give choice — regardless of where it comes from."03:05 — Beyond Oracle and Google Cloud, I'd love to see AWS do some things with competitors. Microsoft has done some of this, but I think there's still room for all these companies to do more. It's the best of both worlds for customers: being able to get combinations of technologies and vendor capabilities rolled into single packages.03:41 — You're going to have both Google Cloud and Oracle salespeople now able to sell Gemini on Oracle Cloud. Just like about a year ago with the multi-cloud agreements — where Oracle Database became available on Azure, AWS, and Google Cloud — those hard and fast walls that used to separate vendors from ever cooperating are coming down. And they're coming down quickly.04:12 — What we see here is that tech vendors have to be not only world-class in the technology they're developing, but also in how they're willing to go to market in unprecedented ways to drive new and significant value for customers. That's going to be one of the primary yardsticks by which vendors are measured — not just by the power of their technology. Visit Cloud Wars for more.

In today's Cloud Wars Minute, I take a closer look at AWS’s market positioning and the narrative shared during its recent earnings call.Highlights00:36 — I was puzzled by some of the math and the numbers that Amazon CEO Andy Jassy, speaking on behalf of AWS on the Amazon earnings call, used in trying to describe the competitive position that AWS finds itself in.01:28 — We've got AWS, now on a $123 billion annualized run rate, and which grew 17.5% in Q2, which is extremely good for a company that size. The challenge comes in when you compare it with the growth rates, both in recent-quarter revenue, but also RPO, or backlog. AWS is far, far behind the other three hyperscalers on those metrics, or the rate of acceleration as they each grew from Q2 over Q1.02:12 — On the earnings call, Jassy said the second player has revenue that's only 65% as big as AWS. So that has to be Microsoft. So he's saying that Microsoft's cloud revenue is more in the range of $75 to $80 billion, but it's more than twice that — $187 billion. I don't get that. Maybe he's trying to say it's only about infrastructure, but he doesn't say that.03:46 — The issue is that the AWS growth rate is lower than that of Microsoft, Google Cloud, and Oracle, and it has been lower for at least the last eight quarters. Yet on the earnings call, Jassy said, "Sometimes we grow faster than they do, sometimes they grow faster than we do."04:07 — So, perhaps Andy Jassy, who has been an extraordinary executive, I'm not questioning his overall capability, his record stands for itself. But Jassy is choosing to try to play a little bit of a shell game here, trying to say that Microsoft's whole cloud revenue isn't all there, or some is illegitimate — something like that.05:26 — It's something that AWS has to address, own up to, and figure out what to do about it. And I think AWS is a great company. I don't think it helps AWS's cause for Andy Jassy to be using some numbers and representations of the market that seem to clash with reality. Visit Cloud Wars for more.

In today's Cloud Wars Minute, I explore AWS's bold new approach to eliminating AI hallucinations using automated reasoning and formal logic.Highlights00:04 — AWS has announced that automated reasoning checks, a new Amazon Bedrock guardrails policy, are now generally available. In a blog post, AWS's Chief Evangelist (EMEA), Danilo Poccia said that: "Automated reasoning checks help you validate the accuracy of content generated by foundation models against domain knowledge. This can help prevent factual errors due to AI hallucinations."00:38 —The policy uses mathematical logic and formal verification techniques to validate accuracy. The biggest takeaway from this news is AWS's approach differs dramatically from probabilistic reasoning methods. Instead, automated reasoning checks provide 99% verification accuracy.01:10 — This means that the new policy is significantly more reliable in ensuring factual accuracy than traditional methods. The issue of hallucinations was a significant concern when generative AI first emerged. The problems associated with non-factual content are becoming increasingly damaging. This new approach represents an important leap forward. Visit Cloud Wars for more.

 In today's Cloud Wars Minute, I unpack the strategy behind Microsoft’s $368B in contracted cloud business and Q4 surge.Highlights00:39— CEO Satya Nadella explained what was behind this huge growth in Microsoft's Azure revenue in Q4, which ended June 30. Overall cloud revenue for the quarter was up 27% to $46.7 billion. This is the first time it's ever released any revenue figures for Azure.01:14 — Azure was up — I think it was 34% — to more than $75 billion. So, I divided four into 75, got almost $19 billion. It said it was “more than,” so I'm going to go with Q4 Azure revenue of $19 billion. And in Q4, it grew 39%. So 34% Azure growth for the year, spiking in Q4 to 39%. And its RPO (remaining performance obligation) was up an astonishing 37% to $368 billion.02:09 — Nadella pointed to, first, classic migrations from on-prem to the cloud. He cited an enormous initiative that Microsoft undertook with SAP to move Nestlé's huge SAP estate from on-prem to the cloud. He talked about cloud-native applications scaling very rapidly.03:16 — And third, he talked about AI workloads: the investments that Microsoft is making and the progress it is making on building out its infrastructure for Azure to be able to handle all of this new and rising demand. And he bristled a little bit at the notion that some other hyperscalers are doing more in the way of data centers and regions and gigawatt capacity and data center capacity04:22 — So again, an extraordinary quarter there from Microsoft Azure, sort of at the heart of so much of this. We'll have a lot more detail on this in an article that we'll be posting later this morning on Cloud Wars. Visit Cloud Wars for more.

In today’s Cloud Wars Minute, I cover Snowflake’s investment in Hightouch to advance agentic AI for personalized marketing.Highlights00:03 — Snowflake Ventures, the investment arm of Snowflake, alongside Capital One Ventures, has announced an investment in Hightouch, a data and AI platform focused on personalized, targeted marketing. In 2024, Hightouch launched an agentic AI solution for lifecycle marketing called AI Decisioning.00:35 — Adam Kaufman, Vice President, Global Head of Industry Go To Market at Snowflake, shared insights about the investment: "By building directly on top of Snowflake's unified platform for data and AI, Hightouch enables marketers to execute faster with less complexity and more control. We're excited to support their continued growth and innovation in the AI Data Cloud."01:08 — This investment represents the confidence that major players like Snowflake have in agentic AI for marketing use cases. Interestingly, agentic AI for marketing was one of the first use cases to be discussed when the technology emerged not so long ago, and has obvious applications, particularly in the area of personalization.01:31 — However, this level of AI-driven personalization comes with its challenges, like data privacy regulations — especially regarding consent, transparency in decision making, and data security. This investment by Snowflake Ventures is encouraging because it should help shape the direction of agentic AI-driven marketing. Visit Cloud Wars for more.

In today's Cloud Wars Minute, I take a closer look at what may be the biggest B2B deal in tech history, Larry Ellison’s $100B+ agreement with OpenAI. Highlights00:15 — We've seen some extraordinary things happen with OpenAI and the alliances it has had. For the first several years of its existence, it was tightly paired with Microsoft. That has changed dramatically now, and OpenAI is now tightly entwined with Oracle.02:01 — What Oracle said in a document it filed was that it has a cloud services contract that will pay it more than $30 billion annually, starting in fiscal 2028. I doubt that Sam Altman at OpenAI wants to go out every year and shop around for new cloud infrastructure providers. So, I think it's reasonable to expect this is going to last two years, three years, perhaps more.02:39 — So now, OpenAI, aligned with Oracle, could have picked anybody. What happened to the Microsoft and OpenAI bromance? Going back to 2019, Microsoft invested $1 billion into OpenAI. In 2021, it followed up with more, and in 2023, at the beginning of the year, the companies each put out announcements saying: We have entered into a deeper, longer-term, incredibly strategic, two-way technology transfer.03:38 — And suddenly we see this change. In just two short years after this big announcement by both of them, we see, in the White House, the first day of President Trump's second term, who shows up but Larry Ellison, Masayoshi Son, the CEO of SoftBank, and Sam Altman, the CEO of OpenAI, to announce a $500 billion initiative called Project Stargate.04:22 — You could have looked at this and said: Microsoft is the biggest tech company in the world. It's the biggest cloud company. It has the most money. It has all these advantages. They'll never let OpenAI slip away. But they did, or was it not so much that Microsoft let it slip away as Larry Ellison said, Hey, I have a better approach. I have a better way here.05:02 — But these shifts we're seeing — where new types of business are done in different ways, with different combinations and partnerships — are challenging the status quo. All these companies have to continue to do new and better things in new and better ways that drive more value and better business outcomes for customers. Visit Cloud Wars for more.

In today's special edition of Cloud Wars Live, Bob Evans talks with Kris Rice, Senior Vice President, Software Development, Oracle Database, about the rapid rise of AI, the strategic impact of the new MCP Server, Oracle’s unique approach to integrating LLMs with private data, and what to expect at this year’s CloudWorld. Rice shares how Oracle is meeting developers where they are, enabling seamless AI-powered interactions with enterprise data.Inside Oracle’s MCP StrategyThe Big Themes:Oracle’s Strategic Integration of MCP: Oracle’s implementation of the Model Context Protocol (MCP) represents more than a technical upgrade. Rice explains that MCP, though technically simple, becomes powerful because of how it’s supported across AI agents, Copilots, cloud desktops, and IDEs. Oracle’s distinctive approach lies in integrating MCP directly into its existing developer environments rather than layering on new tools.Oracle Marries AI with Data Sovereignty: Oracle has the unique ability to merge large language models (LLMs) with enterprise-grade data privacy and sovereignty. While many companies hesitate to use public AI platforms for sensitive queries, Oracle’s approach — running LLMs on private GPUs in Oracle Cloud Infrastructure (OCI) — lets customers leverage cutting-edge AI while keeping data entirely within their own virtual cloud networks.Oracle CloudWorld 2025: Looking ahead to CloudWorld, Rice promises more than vision — he promises practical value. Attendees can expect hands-on labs, demos, and real-world use cases showing exactly how Oracle’s MCP server and GenAI services can be applied immediately. From natural language-driven infrastructure provisioning using Terraform MCP to business logic queries run autonomously by AI agents, the event will be centered on tangible outcomes, not speculative futures.The Big Quote: “I saw the other day that the Google AI Studio supports MCP. So you could literally go to the Google AI studio, drop in your Oracle Database MCP support, and talk to your Oracle database that's on GCP, and never leave the boundaries of Google."More from Kris Rice and Oracle:Connect with Kris Rice on LinkedIn and learn more about Oracle and MCP* Sponsored podcast * Visit Cloud Wars for more.

In today’s Cloud Wars Minute, I discuss Satya Nadella’s bold Q4 claim that Microsoft is the #1 cloud provider, leaving rivals to fight for second place.Highlights00:12 — Microsoft laid out a very clear, unmistakable story in Q4 that it is the biggest and right now certainly the most powerful, most influential cloud provider. CEO Satya Nadella used the opportunity of these unprecedented financial results to spread the word that, “We’re number one, and all of the other hyperscalers — Google, Oracle, and AWS — can fight for the number two spot.”01:06 — Q4 cloud revenue was up 27% to 46.7 billion. Full year cloud revenue was $168 billion, up 23%. One that really blows me away, its RPO for Q4 was $368 billion up 37%, so that is remaining performance,obligation, contracted business not yet recognized as revenue. Then for Azure, for the full year, $75 billion and a 39% growth rate in Q4. So pretty powerful stuff.02:04 — Nadella said, "We're the leaders in breadth of AI products and AI infrastructure." He said, "In our overall tech stack, nobody can touch us, and that's how we're able to deliver at this scale." He said, "We have more data centers than anybody else does," cutting against some claims that have been made by Oracle and others. "We are outperforming everybody else."03:16 — I thought it was interesting that, without identifying those specific companies, he referred to our competitors and that we're doing more than them. Now, I think this was a smart move by Nadella, right? Because we're not talking about some technical features — these are the technological foundations that business leaders will be betting the future of their companies on.03:55 — So, the hyperscalers with the best overall capabilities are going to be the ones that those business leaders tend to go with. I think it's a smart move by Nadella here to say, in my view, here's where we stand. We're out ahead of the others on all this. Now, I'm sure Google Cloud, Oracle, and AWS will have their own points of view on this, and I would encourage all of them to speak up.04:20 — I'm not talking about name-calling, but a very clear and well-articulated, reasoned, customer-oriented set of here's where I stand versus competitors here. I was glad to see this from Microsoft. I hope everybody does it, and later this morning on Cloud Wars, I'll have a detailed article offering the verbatim comments that Nadella made about these different subjects. Visit Cloud Wars for more.