Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Erik Bernhardsson considers on-premise deployments and private cloud VPC peering to be legacy security patterns grounded in outdated perimeter defenses. Controlling the entire cloud stack centrally allows infrastructure teams to deploy new databases, monitor hosts, and ship rapid updates without navigating delicate customer enterprise release cycles.

Related Insights

The future is multi-cloud. Instead of creating proprietary lock-in, the winning strategy is to embrace open standards and build tooling that can run anywhere. The "lock-in" comes from making your own managed infrastructure so performant and easy to use that customers choose to stay out of preference, not necessity.

Many products can't be delivered as a service if their architecture wasn't designed for scalable, multi-tenant management. Cloud-native platforms offer baked-in efficiency, whereas retrofitted on-prem products create significant overhead for vendors and MSPs, hindering their ability to scale effectively.

ENT allows Global 2000 customers to deploy its software within their own cloud environment. This is a strategic decision to address enterprise concerns over data sovereignty, third-party risk, and intellectual property protection. The customer owns their data completely, a key differentiator in the sensitive security space.

Specialized AI clouds (NeoClouds) like CoreWeave emerged because hyperscalers' strengths—such as custom networking and security for multi-tenancy—were detrimental to the performance of large-scale, single-tenant AI workloads. This performance gap created a significant market opening.

Providers like Lightning AI (NeoClouds) must build for unpredictable, diverse customer workloads. This is harder than building for a single, known purpose like OpenAI does for its own engineers. NeoClouds require more performance headroom and robust multi-tenancy architecture to handle any task a customer might run.

Don't try to compete with hyperscalers like AWS or GCP on their home turf. Instead, differentiate by focusing on areas they inherently neglect, such as multi-cloud management and hybrid on-premise integration. The winning strategy is to fit into and augment a customer's existing cloud strategy, not attempt to replace it.

Nikesh Arora observes that enterprise clients cannot reasonably stitch together 40 separate security products when they barely comprehend one tool. Just as 23 disparate sales systems merged into platforms like Salesforce, cybersecurity point tools will inevitably consolidate into unified platforms to eliminate manual glue code.

The initial assumption of a centralized AI model (large hub, large spoke) is wrong. The new model will involve large foundational hubs, enterprise-specific training hubs, and distributed "spokes" of on-premise hardware for inference. This shift is driven by the need for data control and cost efficiency.

While the market rushed to pure-cloud solutions, Egnyte offered a hybrid model. This wasn't a compromise but a strategic advantage for enterprises where physics, like network latency on a construction site, made pure-cloud impractical. The control plane remained in the cloud, while the data plane could be local.

The rise of public cloud was driven by a business model innovation as much as a technological one. The core battle was between owning infrastructure (capex) and renting it (opex) with fractional consumption. This shift in how customers consume and pay for services was the key disruption.