Resilient Digital Infrastructure

We should have always been building resilient infrastructure, but we often haven't been. Now that we're ever more dependent on our digital foundations, new focus is being given to ensuring that they're robust enough to ensure that they can truly keep our enterprises humming. Analysts Jean Atelsek, Melanie Posey and host Eric Hanselman have all been at infrastructure conferences and dig into how expectations and options are changing. There are now regulatory requirements for resilience and the increasing complexity of modern infrastructure can make achieving it more challenging. Enterprises have become more hybrid, linking together cloud and on-premises capabilities. The mad dash to AI has pushed the interconnection of multiple providers to gain access to the latest models. Is what we're building now "really darn intractable"? Software vendors like Broadcom are shifting to integrated platforms to better achieve automated operations and colocation providers like Equinix are increasing the ease with which they build interconnection. Enterprises are racing to raise their operational sophistication as they increasingly face existential consequences from infrastructure failures.   More S&P Global Content: Compute sovereignty: The strategic importance of digital infrastructure Next in Tech | Ep. 222: FinOps – Managing Cloud and AI Costs AI in action: unleashing agentic potential Hyperscaler earnings quarterly: What price inference? For S&P Global subscribers: An agentic architectural approach to enterprise energy resiliency FinOps in the age of agentic AI Security-related outages drive resiliency evolution Service providers race to meet surging enterprise demand for AI infrastructure Host/Author: Eric Hanselman Guests: Jean Atelsek and Melanie Posey Producer/Editor: Dylan Scheible 

The headlines around the levels of capital spending on AI have talked about financial market reactions and the impacts on balance sheets, but there are other effects this spending is having. Research director Brenon Daly returns to the podcast to talk about how reduced cash flow and dwindling reserves are hitting the M&A markets with host Eric Hanselman. Hyperscalers had been steady acquirers as they look to grow organically. They've gone from an active footing to taking a big step back from acquisitions. With so much capital being directed at AI infrastructure, there's very little left for acquisitions.  With the leaders in strategic acquisition pulling back, the range of exit options for early stage companies has narrowed. Hyperscalers have not only been acquirers, but they also set the tone for acquisition trends. SaaS deals have picked up some of the slack, but this is a market that has been significantly curtailed and it will take some time to come back.   More S&P Global Content: Next in Tech | Ep. 259: The RSAC Conference – Agents on The Loose For S&P Global subscribers: Building AI doesn't leave much money for buying Hyperscaler earnings quarterly: Cash flows running low as AI portfolios evolve Going the wrong way on the 'K' Bruises and bargains in the tech M&A market Host/Author: Eric Hanselman Guest: Brenon Daly Producer/Editor: Dylan Scheible 

We've talked extensively about the demand side of data centers and AI and it's important to consider what's taking place in sustainably addressing it. Analyst Tony Lenior returns to the podcast to discuss his recent research with host Eric Hanselman. Booming demand has raised concerns that hyperscalers and data center operators would abandon sustainability goals, but, while some timelines have been pushed back, they continue to be some of the largest buyers of renewable Power Purchase Agreements (PPA) and are driving carbon-free energy pursuits. That's despite three major energy shocks in the last six years. Battery energy storage systems (BESS) are seeing wider deployment and rapid technology advances and are a means to addressing wind and solar's intermittency.  Water is another key resource associated with the data center operation and an area where progress is being made. It's part of a more complex set of interconnections, though as direct water consumption is typically what's measured, but secondary water consumption can be just as significant. Nuclear power is carbon-free, but consumes large quantities of water for cooling. Hydropower has even more complicated impacts on water use, despite its fundamentally sustainable nature. Sustainability concerns are critical to understand and very challenging to fully resolve.   More S&P Global Content: Renewable energy curtailment offers a surplus opportunity for Texas data centers Next in Tech | Ep. 266: Managing data center risk Nearly two-thirds of planned US BESS projects eligible for 40% tax credits For S&P Global Subscribers: Microsoft opens new windows into renewable investment, doubling down on decarbonization Amazon delivers on green commitments, but emissions rise amid the AI boom Google's search for power sustainability is yielding results, but AI slows progress Powering data centers: Leveraging energy community incentives in a changing regulatory landscape Host/Author: Eric Hanselman  Guest: Tony Lenoir Producer/Editor: Dylan Scheible 

Labeled Security Summer Camp by some, the Black Hat and DEF CON security conferences have been the premier events for security researchers and practitioners. Analysts Dan Kennedy and Justin Lam returned from the heat of Las Vegas to discuss their take on what went down in the desert with host Eric Hanselman. More than 23,000 attendees were reported to be part of the sprawling Black Hat start to the week and DEF CON has grown to be its equal in size. There was more discussion about the realities of what AI can do for and against security operations. The implications around Mythos have raised fears of a cascade of vulnerability disclosures that will overwhelm our ability to remediate them, but a growing risk comes from AI tooling being able to create more complex environments and processes to make complex vulnerabilities more easily exploitable. On the security defense side, AI's ability to interpret complex interactions and relationships is showing potential to reduce the risk of rolling out new controls by understanding what policies could break. It's an area where there have historically been delays in improving protections where concerns that new controls might break existing applications or access held security teams back. One of the greatest benefits of the conferences, though, may be the ability of security professionals to gather together and have the informal interactions that are needed to support each other in the complex world that information security has become.   More S&P Global Content: Next in Tech | Ep. 267: Mythos and Security Generative AI Digest: OpenClaw and Moltbook make a splash Generative AI funding: A sober retrospective and the trends shaping 2026 RSAC Conference 2026: Confronting both the promise and threat of agentic AI For S&P Global subscribers: Information security budgets rise, but must address cloud security, AI governance – Highlights from VotE: Information Security AI's impact in security and its application are not always aligned – Highlights from VotE: Information Security Fulcrum accountability, aligning personas and progress in data security Host/Author: Eric Hanselman  Guests: Dan Kennedy and Justin Lam Producer/Editor: Dylan Scheible 

Quantum computing might be seen by some as a distant future, but quantum computers are doing useful work today and having impacts on a range of areas in technology. Analyst Ellie Brown and host Eric Hanselman were at IBM's Quantum Day and explore not only what they saw there, but also the broader implications that advances in quantum technology are having. Behind all of the shiny stainless steel cabinets are the exotic superconducting "chandeliers" that have become the face of quantum computing and also racks of classical computing systems that are translating computational results and managing operations and interconnection. While the quantum systems are advancing rapidly, the pipeline of skilled talent that can put it to work is still under development. The ecosystem of coding tools, simulators and algorithms is moving forward, but it has to be able to scale. Quantum impacts on information security and trust are just beginning to be understood in the wider technology community. With the risk of quantum computing breaking some types of encryption looming around the end of the decade, organizations of all sizes should be scrambling to assess their risks and starting to implement quantum safe protections. 451 Research data shows that enterprises are starting on the journey, but they need to be much further ahead to meet timelines that keep getting closer with each new quantum advancement.   More S&P Global Content: Quantum Quarterly — Q2 2026: Momentum meets innovation Disrupting the future: Quantum in data centers? Next in Tech | Ep. 276: Quantum CEO Series: Andy Ory Next in Tech | Ep. 248:SC25 Supercomputing conference For S&P Global subscribers: Quantum Computing Market Monitor & Forecast Quantum road maps signal steady gains and a shift from promise to performance Quantum's advances come with major cybersecurity implications Host/Author: Eric Hanselman Guest: Ellie Brown Producer/Editor: Dylan Scheible 

Adoption of AI has been dramatic, but the boost that the release of the OpenClaw agent technology delivered in November of last year was even more so. The mad rush to put agents to work has had wide ranging impacts, including market impacts on the supply of the venerable Mac Mini computer. That intensity and the need to manage the shockwaves agents are creating are what analysts Jean Atelsek and Marissa Incera joined host Eric Hanselman to discuss. There are now a proliferation of agents and they are creating demands on enterprise infrastructure that are outstripping those of initial AI adoption. The quandary enterprises are wrestling with is how to harness agentic power without limiting the innovations they can provide and still bounding the activities they can undertake. Effective agent use requires more than just an agent framework running on a single system. There has to be mediated access to data, management of context and memory in the agentic flow and security around agent action. The headlines are already filled with stories of the havoc that can be wreaked by misguided attempts. It requires a lot of diligence to get this right.   More S&P Global Content: Generative AI Digest: OpenClaw and Moltbook make a splash Generative AI funding: A sober retrospective and the trends shaping 2026 Next in Tech | Ep. 273: FinOps and AI Hyperscaler earnings quarterly: What price inference? For S&P Global Subscribers: Optimizing agentic AI systems Pricing the agentic enterprise: Megawatts, outcomes and intelligence in industrial environments Infrastructure and data challenges persist, agentic ones await – Highlights from VotE: AI & Machine Learning Host/Author: Eric Hanselman Guests: Jean Atelsek and Melissa Incera Producer/Editor: Dylan Scheible Published With Assistance From: Feranmi Adeoshun and Sophie Carr

The world of IT is filled with technical qualifications in all manner of disciplines. They've become an expected baseline for knowledge in broad fields like networking and security, as well as narrower, vendor-specific knowledge and skills. But what happens as IT leaders move up to more senior roles? Steve Clarke, cofounder and director of Freeman Clarke joins host Eric Hanselman to talk about the challenges that senior IT professionals have in establishing their bona fides as technical business people. The role of a CIO, CTO, CISO or CDO requires technical depth, but the more important part is the ability to integrate that knowledge with business operations. It's not an easy transition to undertake and even more complicated to identify.  Discussions that are taking place in the 451 Research 451 Alliance community brought up the idea of a business level certification process and how it would have to characterize that unique knowledge set. The Standard is an approach that Steve and his colleagues began to address this issue. The 451 Alliance membership is free and open to all qualified IT professionals and business leaders who want to contribute to industry-leading research. Members take anonymous online surveys in exchange for access to research results that power their IT strategy.   More S&P Global Content: Join the 451 Alliance The Standard Certification The Problem with the Status Quo For S&P Global Subscribers: US tech spending intent dips in Q2, but projects to grow in Q3 – Tech Demand Indicator Highlights from Q2 2026 Beyond connectivity: A CIO's road map from network APIs to business impact Enterprises cautiously optimistic for AI amid data management challenges Host/Author: Eric Hanselman Guest: Steve Clarke, cofounder and director of Freeman Clarke and Tech Leaders Connect Producer/Editor: Dylan Scheible Published With Assistance From: Feranmi Adeoshun and Sophie Carr

As enterprises integrate greater levels of AI into their operational workflows, many are concerned about the effectiveness and accuracy of the decisions with which it's being entrusted. Mike Fratto joins host Eric Hanselman to talk about his recent research into the ways in which observability can provide greater context to AI systems and administrative teams. Data is the raw material from which AI value is created and a lack of solid operational data will mean that management systems won't have a complete perspective. Observability platforms have been expanding to collect and correlate a broader range of data and they can be particularly useful, if they're integrated well. Observability data can also assist in building confidence in automation and support automation efforts by creating scripts and validating control decisions. It can be a huge help to staff to digest and make sense of existing run books and processes and guide decisions in the event of failures. When novel incidents occur, being able to rule out certain areas can be just as powerful as finding the root cause. No, the network isn't always the cause of the outage… More S&P Global Content: Next in Tech | Ep. 271: AI Networking Next in Tech | Ep. 247: Security and Observability For S&P Global subscribers: Updating processes and building trust are key to full operational autonomy – Highlights from VotE: DevOps Observability Market Monitor & Forecast Agent orchestration and observability go hand in hand: The case for governance in motion Host/Author: Eric Hanselman Guest: Mike Fratto Producer/Editor: Dylan Scheible Published With Assistance From: Feranmi Adeoshun and Sophie Carr

There is a dramatic push of technology into space. Low Earth Orbit (LEO) satellites are changing the broadband communications market and there's speculation about putting data centers in orbit. Analysts Ellie Brown and Johan Vermij join host Eric Hanselman to look at the practical aspects of this push into a new frontier as a preview to their upcoming webinar. The cost of access to space has been dropping significantly, but there are many considerations that impact the viability of some of these plans. LEO satellite lifetimes are shorter than geosynchronous ones and the sustainability impacts of the launch volumes necessary to maintain multiple, massive constellations are daunting. Space is an attractive vantage point for observation with imaging to support geopolitical needs and ecological ends, like methane emissions tracking. Synthetic Aperture Radar (SAR) is piercing cloud and vegetation layers to reveal landscape details. Space-based data centers are seeing a lot of interest, but popular misconceptions about cooling capabilities may be adding more encouragement than is warranted. While space is "cold", radiating heat is more complicated in its vacuum. The International Space Station, with a massive radiation system, rejects only 70 KW of heat, less than a few racks of equipment and much less than a typical GPU cluster. And the need to service equipment on orbit hasn't been meaningfully addressed. Smaller satellites with distributed processing could pre-process data before it's sent to earth, but that will require greater levels of coordination to manage the massive scale on which this would have to operate.  More S&P Global Content: Join the webinar:  Orbit as the Next Data Frontier: Capital Flows, Risk, and Realities For S&P Global Subscribers: NVIDIA gives 'space cred' to orbital data center race with Vera Rubin Space module announcement STMicroelectronics eyes $3 billion in annual space revenue as demand for LEO satellites surges As the space sector enters the infrastructure era, M&A could become a force to be reckoned with Host/Author: Eric Hanselman  Guests: Ellie Brown and Johan Vermij Producer/Editor: Dylan Scheible Published With Assistance From: Feranmi Adeoshun and Sophie Carr

The popular image of quantum computers is the sparkling "chandelier" of the superconducting approach, but this next episode in our Quantum CEO series looks neutral atom quantum computers. Quantum analyst Ellie Brown returns to the podcast to join Andy Ory, the CEO of QuEra, to discuss their approach, the benefits of practical fault tolerant qubits and the advantages of quantum hubs that may have a dearth of beer with host Eric Hanselman. Neutral atom systems don't require the ultra-low operating temperatures used for superconduction. They stabilize the atoms they're using with lasers that still their motion. That free space setting allows simpler interaction between larger numbers of qubits that improves scaling of quantum computations. Stability reduces error rates and can improve the effectiveness of quantum error correction. Quantum computing has become a very buzzy topic, with IPO's and major funding announcements. There is growth of the supporting ecosystems around quantum application development, with software tools and cloud-based services offerings that are expanding access. The greatest challenge is the supply of a quantum workforce that can put quantum capabilities to work. Systems doing optimization approaches, like quantum annealing, are here today. The date for the availability of general purpose, practical quantum computers keeps getting closer. Some have said 2030 might be a target date and QuEra is talking about delivering their next-generation system in 2028. It's time to be getting quantum-literate today! More S&P Global Content: Long-term advantage is triggering a quantum wave, but cost and immaturity remain barriers Cybersecurity, talent needs and ongoing adoption were key themes at Quantum.Tech USA 2025 Access to quantum hardware remains cloudy, but more options are starting to appear For S&P Global subscribers: Quantum road maps signal steady gains and a shift from promise to performance Long-term advantage is triggering a quantum wave, but cost and immaturity remain barriers – Highlights from VotE: Digital Pulse Quantum Computing Market Monitor & Forecast AI, quantum and high-performance computing join forces at SC25 Host/Author: Eric Hanselman Guests: Andy Ory and Ellie Brown Producer/Editor: Dylan Scheible Published With Assistance From: Feranmi Adeoshun and Sophie Carr