The Edge Computing Podcast with Fexingo: Local Compute, CDNs, and Distributed Infrastructure

Lucas and Luna dig into one concrete edge computing use case that's saving companies serious money: local content caching and compute at the network edge. They walk through a real mid-sized e-commerce company that shifted 40 percent of its CDN traffic to local edge nodes and cut its monthly bandwidth bill by 37 percent — approximately $14,000 a month. The hosts debate the upfront cost of deploying edge servers versus the long-term savings, and discuss which industries benefit most (streaming, IoT sensor networks, online gaming). They also touch on the trade-offs: increased hardware maintenance overhead, the need for local IT skills, and the risk of vendor lock-in with proprietary edge platforms. By the end, listeners will know one specific number to quote at their next team meeting and a framework for evaluating whether edge compute makes financial sense for their own business. Plus, a brief moment on listener support that keeps the podcast ad-free.

#EdgeComputing #Technology #Bandwidth #CDN #ContentDelivery #LocalCompute #DistributedInfrastructure #CloudCosts #Latency #IoT #Streaming #OnlineGaming #ITInfrastructure #CostOptimization #EdgeServers #NetworkEdge #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Edge Computing Podcast, Lucas and Luna explore the overlooked bottleneck in distributed infrastructure: heat. As compute moves closer to the user—inside warehouses, retail stores, and telecom huts—traditional data center cooling methods fail. Lucas digs into the numbers: a typical edge node generates 2 to 10 kilowatts of heat in a space designed for zero HVAC. They discuss direct-to-chip liquid cooling and immersion cooling as emerging solutions, with real examples from startups like Iceotope and industry adoption by Nokia and Intel. Luna challenges whether the cost premium makes sense for low-power edge nodes, while Lucas argues that density in AI inference at the edge is driving the shift. They also touch on how the CDN industry is quietly retrofitting Points of Presence. A concrete, numbers-driven look at the thermal side of edge computing—and why it matters for reliability and total cost of ownership.

#EdgeComputing #DataCenterCooling #LiquidCooling #ImmersionCooling #ThermalManagement #DistributedInfrastructure #EdgeAI #CDN #Iceotope #Intel #Nokia #DirectToChip #PowerUsageEffectiveness #EdgeNode #Technology #FexingoBusiness #BusinessPodcast #Infrastructure

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Edge Computing Podcast, Lucas and Luna explore the shift from training AI models in the cloud to running inference at the network edge. They examine how companies like Tesla and Apple are deploying on-device AI for real-time tasks, the role of CDNs like Cloudflare's Workers AI in enabling low-latency inference, and the economics of edge inference versus cloud-based processing. The hosts discuss the trade-offs in model size, latency, and cost, and highlight a case where a manufacturer reduced defect detection time by 83% using edge inference. They also address the challenges of model updates and security in distributed environments. By the end, listeners will understand why edge inference is becoming a critical component of modern distributed infrastructure.

#EdgeComputing #AIInference #OnDeviceAI #RealTimeAI #Tesla #Apple #CloudflareWorkersAI #LatencyReduction #DistributedInfrastructure #LocalCompute #ModelOptimization #EdgeAI #Technology #FexingoBusiness #BusinessPodcast #EdgeInference #CDN #InferenceAtEdge

Keep every episode free: buymeacoffee.com/fexingo

Edge computing is often framed as a cloud-vs-local battle, but the real story is geography. Lucas and Luna unpack why latency-sensitive applications — from autonomous vehicle fleets to real-time financial trading — are pushing compute back toward the physical world. They walk through a concrete example: a delivery drone fleet that shaves 40 milliseconds off decision time by processing directly at a local node instead of routing through a regional data center. The hosts also examine the hidden economics of edge — how bandwidth costs, data sovereignty laws, and AI inference are quietly reshaping the CDN landscape. If you think edge is just a buzzword, this episode will make you think again.

#EdgeComputing #LocalCompute #CDN #DistributedInfrastructure #Latency #RealTimeData #AutonomousVehicles #DroneDelivery #BandwidthCosts #DataSovereignty #AIInference #FogComputing #Technology #TechPodcast #FexingoBusiness #BusinessPodcast #LucasAndLuna #EdgePodcast

Keep every episode free: buymeacoffee.com/fexingo

Content delivery networks were built for a world where data lived in centralized data centers and users requested static assets. But the rise of edge computing is flipping that model on its head. In this episode, Lucas and Luna explore how CDNs like Cloudflare and Fastly are pivoting from caching images to running serverless compute at the edge. They break down the economics: why a shift from bandwidth pricing to compute pricing matters, how latency requirements for real-time applications like autonomous vehicles or live video processing make centralized cloud impractical, and what the typical latency drop looks like when you move from a regional data center to a local edge node. The conversation also touches on the tensions between CDN providers and cloud hyperscalers, and whether we're heading toward a world where every cell tower has a server. If you work in infrastructure, product, or just want to understand where the internet is headed, this episode gives you a concrete framework.

#EdgeComputing #CDN #Cloudflare #Fastly #ContentDeliveryNetwork #Latency #Serverless #DistributedInfrastructure #CloudComputing #Technology #TechPodcast #NetworkInfrastructure #ComputeAtTheEdge #AutonomousVehicles #LiveStreaming #DataCenter #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna unpack the security paradox at the heart of edge computing: the more distributed your infrastructure, the larger your attack surface. They walk through the 2023 Okta breach that compromised 134 customers via a stolen service account credential, and explain why legacy VPNs and perimeter-based security models break down when your compute nodes live in a factory in Ohio, a warehouse in Amsterdam, and a cellular tower in rural Japan. The core argument: zero trust network architecture (ZTNA) isn't optional anymore — it's the only way to authenticate machine-to-machine traffic at scale. Lucas breaks down the three principles of zero trust for edge nodes — never trust, always verify; microsegmentation; and least-privilege access — using the 2021 Microsoft Exchange hack as a counterexample of what happens when you skip them. Luna brings data from a Gartner report predicting that 60% of enterprises will phase out VPNs for edge access by 2027. They close on whether zero trust adds too much latency for real-time edge workloads, and what chip-level security like Intel SGX and AMD SEV might offer.

#EdgeComputing #ZeroTrust #Cybersecurity #OktaBreach #ZTNA #Microsegmentation #VPNPhaseOut #MicrosoftExchangeHack #Gartner #IntelSGX #AMDSEV #MachineToMachine #NetworkSecurity #Latency #TechPodcast #Infrastructure #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Edge computing promises to reduce latency, but how do you actually process data at the edge without a full data center in your living room? In this episode of The Edge Computing Podcast, Lucas and Luna dive into the emerging world of local edge servers—compact, fanless devices that sit in your home or office and handle AI inference, video analytics, and IoT data on the spot. They explore the specific case of the Nvidia Jetson Orin Nano, a $500 module that can run 40 video streams simultaneously, and discuss how companies like Vapor IO and Fastly are building hybrid edge networks that process 80% of requests locally. Lucas breaks down the economics: at $0.0003 per inference, local compute can be 10x cheaper than sending data to the cloud. Luna questions whether consumers will tolerate another box in their home, and they both examine the security trade-offs of distributed infrastructure. By the end, you'll understand why the edge isn't just about CDNs anymore—it's about putting real compute power where the data lives.

#EdgeComputing #LocalCompute #NvidiaJetson #VaporIO #Fastly #LatencyReduction #AIInference #IoT #CDN #DistributedInfrastructure #HybridEdge #TechPodcast #Technology #FexingoBusiness #BusinessPodcast #EdgeServers #RealTimeAnalytics #DataProcessing

Keep every episode free: buymeacoffee.com/fexingo

In the premiere episode of The Edge Computing Podcast with Fexingo, Lucas and Luna explore why the cloud is moving closer to you. They anchor on a specific number: by 2026, more than 50% of enterprise-generated data will be created and processed outside traditional centralized data centers. Using Akamai's 2024 acquisition of a 5G edge platform as a concrete case, they break down what edge computing actually means — from autonomous vehicles needing millisecond reaction times to retailers processing inventory data inside local stores. Lucas explains how content delivery networks like Cloudflare are evolving into full compute platforms, while Luna questions whether this is just a marketing rebrand of old distributed systems. They establish the show's ethos: no hype, just the real infrastructure shifts. The conversation closes with a forward look at how edge computing could reshape privacy, latency, and even the geography of the internet.

#EdgeComputing #CloudComputing #Akamai #Cloudflare #5G #DistributedInfrastructure #ContentDeliveryNetworks #AutonomousVehicles #Latency #DataPrivacy #EnterpriseTech #Infrastructure #Technology #FexingoBusiness #BusinessPodcast #PodcastLaunch #LocalCompute #EdgeAI

Keep every episode free: buymeacoffee.com/fexingo