AI & Cloud Trends for August 2025

Todd Robinson, Founder and President of OpenMetal.io, discusses the resurgence of bare metal infrastructure driven by AI workloads, digital sovereignty requirements, and companies reassessing public cloud economics. The conversation explores how organizations are finding cost and control advantages in bare metal solutions, particularly for long-running applications. It examines OpenMetal's open-source approach using technologies like Ceph and OpenStack to deliver flexible infrastructure alternatives.SHOW: 959SHOW TRANSCRIPT: The Cloudcast #959 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwNEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS" SPONSORS:[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcastSHOW NOTES:OpenMetal websiteTopic 1 - Welcome to the show. Tell us about your background, and you developed a passion around bare metal and cloud services?Topic 2 - Between AI (GPU rental), digital sovereignty initiatives, and even virtualization alternatives, it feels like bare metal is having a resurgence. Give some a sense of what the demands for bare metal solutions look like today?Topic 3 - As companies understand the economics of having used public cloud services, are there certain use-cases that become immediately obvious where more private, hosted, bare metal services just make more sense? Topic 4 - OpenMetal is based on open source technologies like Ceph and OpenStack. How important to customers about the technologies under their applications, or do the economics and control aspects play a bigger role in their decisions? Topic 5 - I’m often asked if there is a model about when it makes more sense to use on-demand service vs. more fixed services. Is there a rule of thumb (e.g. longevity of an application, amount of change, etc.) that you’ve found drives the most success at picking the right environment for applications?Topic 6 - OpenMetal could be described as a public or private cloud service. Do you find that there is still the stigma over “private cloud” that we saw when the hyperscalers were initially growing so quickly?FEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

How did Oracle Cloud (with OpenAI) suddenly command the news cycle, with a chance to potentially overtake much larger clouds in a few years? Let’s explore how Oracle’s focus on bare-metal and bandwidth might have ripple effects on Cloud and AI.SHOW: 958SHOW TRANSCRIPT: The Cloudcast #958 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcast[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search “Interconnected by Equinix”.SHOW NOTES:WHAT ARE THE ALTERNATIVE PATHS FOR AI? The Cloudcast #330 - Oracle’s Next-Generation Cloud IaaSOracle Chosen as TikTok’s Secure Cloud Provider- [Oracle Press Release]Oracle Cloud and OpenAI sign $300B deal - [WSJ]Oracle Cloud - [Platformnomics]Oracle and OpenAI - [Ed Zitron]FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

Ole Lensmar (@olensmar, Founder/CTO at @TestKube_io) discusses how Kubernetes-native testing platforms are designed to address limitations in traditional CI/CD testing workflows. The conversation covers how TestKube differs from existing testing environments, expands test coverage opportunities for development and QA teams, and provides best practices for testing in Kubernetes environments.SHOW: 957SHOW TRANSCRIPT: The Cloudcast #957 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwNEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS" SPONSORS:[Interconnected] Interconnected is a new series from Equinix diving into the infrastructure that keeps our digital world running. With expert guests and real-world insights, we explore the systems driving AI, automation, quantum, and more. Just search Interconnected by Equinix.[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcast[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.SHOW NOTES:TestKube - A Kubernetes-native platform that powers Continuous Testing for today’s AI-accelerated developmentTestKube (open source)Why did we start TestKube (Ole Lensmar)Topic 1 - Welcome to the show. Tell us about your background and what led you to start TestKube.Topic 2 - Let’s talk about the origins of TestKube. What were some areas where you saw people having frustrations or limitations that were holding back their ability to do proper testing to get things into production?Topic 3 - Let’s talk about the basics of TestKube. Can you talk about how it’s different from existing testing environments, or how people use CI/CD todayTopic 4 - Does TestKube expand what a typical Dev-team, or QA-team would test, or does it create new opportunities for test coverage that were very difficult before? Topic 5 - What are some of the results or feedback you’ve heard from people using TestKube?Topic 6 - What are some best practices you’re seeing as people begin to evolve how they test for their Kubernetes environments?Topic 7 - What’s the best way for people to get started with TestKubeFEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

Apple’s iPhone had the Android ecosystem, but where is the Android of AI going to come from to challenge the OpenAI/NVIDIA duopoly? Let’s explore the possibilities. SHOW: 956SHOW TRANSCRIPT: The Cloudcast #956 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcastSHOW NOTES:WHAT ARE THE ALTERNATIVE PATHS FOR AI? Consumer: Ad-supportedConsumer: Understandable tiersEnterprise: ELAs for Co-Pilot, GeminiEnterprise: Model building tools (e.g. AWS Sagemaker, Google Vertex, etc.)FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

Brian Gracely (@bgracely) Aaron Delp (@aarondelp) and Brandon Whichard (@bwhichard, @SoftwareDefTalk) discuss the top stories in Cloud and AI from August 2025.SHOW: 955SHOW TRANSCRIPT: The Cloudcast #955 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST:  "CLOUDCAST BASICS"SPONSORS:[TestKube] TestKube is Kubernetes-native testing platform, orchestrating all your test tools, environments, and pipelines into scalable workflows empowering Continuous Testing. Check it out at TestKube.io/cloudcast[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.SHOW NOTES:Link to August 2025 News and ArticlesFEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

From extreme ups to startling downs, every week can feel like the peak of expectations and the trough of disillusionment for AI.  SHOW: 954SHOW TRANSCRIPT: The Cloudcast #954 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.[VASION] Vasion Print eliminates the need for print servers by enabling secure, cloud-based printing from any device, anywhere. Get a custom demo to see the difference for yourself.SHOW NOTES:THE UPS AND DOWNS OF AI - THE CONSTANT HYPE CYCLEHealthy Competition [YES]Consumer and Enterprise Markets [YES]Market leader(s) [YES, sort of]Well-Defined, profitable business model [NO]Open, lower-cost alternative emerged [YES/NO]Usage patterns emerging [YES/NO][ups] Constant high-profile VC, Sovereign wealth, hyperscaler funding of AI startups[ups] Constant high-profile CAPEX spending by hyperscalers, model builders, data center builders[ups] Rapidly growing user-bases[ups] Growing revenues at some companies (NVIDIA, OpenAI, Google, Microsoft, Anthropic)[downs] Growing losses at high-profile companies  [???] Frequent, high-profile movement by key people at high-profile companies (engineers, leadership)[downs] Frequent, partial acquires of founders and key engineers, but not entire companies[downs] Secondary market scales of startup shares, bypassing traditional secondary and public markets[downs] No “NetFlix of AI” company[downs] No “AI Agent” success stories[???] Consumer “winner-take-all” mindset from AI companies[downs] Enterprise companies struggling to create ROI+ projects (in early days)[???] Enterprise “bundles” raising prices (CoPilot, Gemini, etc.)[downs] Unclear if new frontier models are getting better than previous versions (e.g. GPT-5)[???] Are inference prices coming down? [ups] Consumers have many excellent AI choicesIt’s unclear if AI companies have created any moats yet; it’s unclear if LLMs can be differentiatedChatbots, developer-assistants and document management are use-cases. What else?Are agents ready to be mainstream yet? Pick-axe providers are making the money right now (NVIDIA, Broadcom, etc.), but is there moat entirely on super-premium HW?User-experiences are still being understoodWill AI + Ads (business model) be a big bang event, or happen gradually? FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

Amit Kinha - Field CTO @DoITInt | FinOps Foundation Board Member discusses how the relationship between technology and financial accountability has evolved, and how mainstream FinOps is shifting IT focus around innovation. SHOW: 953SHOW TRANSCRIPT: The Cloudcast #953 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwNEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS" SPONSORS:[VASION] Vasion Print eliminates the need for print servers by enabling secure, cloud-based printing from any device, anywhere. Get a custom demo to see the difference for yourself.[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.SHOW NOTES:DoIT - Intent Aware FinOpsDoIT hires Citigroup FinOps Star as ExecutiveFinOps FoundationFinOps X (event)Topic 1 - Welcome to the show. Tell us about your background and your involvement in FinOps.Topic 2 - We’ve been through the early days of cloud, where we were told that cloud was cheaper, and then after COVID, everyone seemed to realize that cloud was actually more expensive. Where are we with FinOps and companies understanding how to think about cloud spending?Topic 3 - You work with the FinOps Foundation. What types of roles do you see focused on FinOps, and how is that evolving as there is greater cloud cost visibility across an entire organization? Topic 4 - How has the technology around FinOps evolved? How much is still manual? How much is piecing together different costs from different systems? And how much is evolving to have an AI component? Topic 5 - How do the tools and platforms from DoIT help to make FinOps easier for companies? What are some of the key areas of focus, and some insights you’re hearing from the companies that you work with directly? Topic 6 - If you had to give a CIO or CFO guidance on how to best think about FinOps, what are the top things you would focus on? FEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

As we begin to transition from the Cloud era to the AI era, what types of changes can we expect to see happen in the market, within our businesses, and for individuals? SHOW: 952SHOW TRANSCRIPT: The Cloudcast #952 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.[VASION] Vasion Print eliminates the need for print servers by enabling secure, cloud-based printing from any device, anywhere. Get a custom demo to see the difference for yourself.SHOW NOTES:WHEN TWO ERAS OVERLAP - CLOUD AND AIThe early stages of a new era tends to be separate (cloud-native, AI-native), but eventually tends to build upon the previous era’s foundationsIt usually takes a decade before an era stops being important or prioritized It usually takes 3-5 years before best practices for new applications emergeFinding new era talent is difficult (and expensive), but current era technology can be optimized (reduce costs, modernize)We’re going to see “Shadow AI” and “Bi-Modal Cloud/AI”The will be a lot of “X is Dead” during the first few years of the transition, until people (again) remember that nothing in IT dies - especially in the enterpriseThere will be a lot of numbers/reports coming out that seem completely contradictory - $B investment, but $0 ROI from usersFEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

After a very long roadtrip, let’s explore the lessons we can learn from some of the greatest business models in tech and how or if they apply to the AI era companies. SHOW: 951SHOW TRANSCRIPT: The Cloudcast #951 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotwCHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"SHOW SPONSORS:[VASION] Vasion Print eliminates the need for print servers by enabling secure, cloud-based printing from any device, anywhere. Get a custom demo to see the difference for yourself.[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.SHOW NOTES:7 Powers of Competitive DynamicsMicrosoft (Vol 1, Vol 2)Google Meta AWSNVIDIA (Vol 1, Vol 2, Vol 3) LESSONS FROM THE GREATEST BUSINESS MODELSFounder control, nearly immediately profitable, growth strategies, How long Enterprise businesses take to matureIs there an AI business model that’s profitable? Is there a business model that isn’t Enterprise or Ad-supported?Microsoft - commoditize your compliments, understand market expansions, Google - Customer acquisition, scaling economics, AdWords, AdSenseAWS - Meta - Adapting the offeringsNVIDIA - Vertical integration (HW + SW)FEEDBACK?Email: show at the cloudcast dot netTwitter/X: @cloudcastpodBlueSky: @cloudcastpod.bsky.socialInstagram: @cloudcastpodTikTok: @cloudcastpod

Haseeb Budhani (@haseebbudhani, CEO @rafaysystemsinc) discusses the evolution from traditional DevOps to platform engineering and what "Enterprise Ready" Kubernetes looks like in 2025. We explore AI workloads running on Kubernetes and how modern orchestration solutions can transform teams from bottlenecks into enablers. We also cover the security considerations for GPU-enabled AI workloads and balancing developer self-service capabilities with proper governance and control.SHOW: 950SHOW TRANSCRIPT: The Cloudcast #950 TranscriptSHOW VIDEO: https://youtube.com/@TheCloudcastNET NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST: "CLOUDCAST BASICS"SPONSORS:[DoIT] Visit doit.com (that’s d-o-i-t.com) to unlock intent-aware FinOps at scale with DoiT Cloud Intelligence.[VASION] Vasion Print eliminates the need for print servers by enabling secure, cloud-based printing from any device, anywhere. Get a custom demo to see the difference for yourself.SHOW NOTES:Rafay websiteTopic 1 - Welcome to the show, Haseeb. Give everyone a quick introduction.Topic 2 - Let’s start by talking about the evolution of Kubernetes as a platform. You’ve said and we’ve talked about on this show for some time how Kubernetes is more of a platform to run platforms. We’ve also seen trends in the industry and shifts in what it means to be DevOps or Platform Engineering in recent years. You've positioned Rafay as a Kubernetes Operations Platform that's now evolved into a Cloud Automation Platform. How do you define the difference between Kubernetes management and true platform engineering?Topic 3 - What does “Enterprise Ready” Kubernetes look like in 2025?Topic 4 - Let’s flip over to AI/ML and GPUs with Kubernetes for a bit. Many developers and data scientists aren’t aware of the underlying platform they run on. I saw a stat recently that about 95% of AI runs on Kubernetes, either on-prem or in the cloud. Despite this, Platform teams are often stuck doing manual GPU provisioning, which doesn't scale with AI adoption. How do modern GPU orchestration solutions change the platform team's role?Topic 5 - With GPU workloads often handling sensitive data and AI models, security becomes even more critical. How should organizations approach security and compliance in their GPU-enabled Kubernetes operations?Topic 6 - "Most developers don't want to write YAML or manage clusters — they just want to ship software." How do you balance giving developers the self-service capabilities they want while maintaining the control and governance that platform teams need?FEEDBACK?Email: show at the cloudcast dot netBluesky: @cloudcastpod.bsky.socialTwitter/X: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod