Get your free personalized podcast brief

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

Evan Conrad says SF Compute shipped the ability for customers to subdivide their InfiniBand fabric without asking the operator. That lets platform-as-a-service companies redraw cluster boundaries on the fly and scale up or down into the market. He calls it virtualizing the data center and making it programmable.

Related Insights

Because autonomous agents enable a surge of personalized micro-applications used intermittently, hosting software on persistent infrastructure is economically untenable. Next-generation serverless architectures must automatically spin down unused instances and dynamically scale resources like liquid across diverse workloads to support sustainable compute economics.

The evolution from physical servers to virtualization and containers adds layers of abstraction. These layers don't make the lower levels obsolete; they create a richer stack with more places to innovate and add value. Whether it's developer tools at the top or kernel optimization at the bottom, each layer presents a distinct business opportunity.

Evan Conrad says SF Compute performs cloud-like work—data-center site visits, bill-of-materials design, cluster architecture, rack rollout, and backplane management of PDUs, switches, DPUs and BMCs—but refuses to own hardware. Ownership would let it manipulate the forwards market, so it stays independent like Switzerland.

Unlike general-purpose cloud resources, AI training infrastructure with specialized networking (e.g., InfiniBand) and storage cannot be added fungibly. It requires significant pre-planning and deep integration, breaking the standard cloud deployment model of simply adding more commoditized compute or storage as needed.

Traditional network fabrics route erratic, destination-agnostic packets, but AI training and inference produce massive, predictable data flows. This architectural shift enables optical circuit switching directly into the compute complex, blurring the traditional operational boundary between scale-up and scale-out datacenter topologies.

Inspired by Google, Cloudflare made an early decision to build its global network using inexpensive, commodity hardware instead of specialized equipment. This software-centric approach allows them to scale their infrastructure rapidly and cost-effectively, a key structural advantage over competitors.

Evan Conrad says two VMs from one operator can differ by 40% in kernel-launch performance, and storage, WAN transit, preemption and chip generation also change value. Rather than wait for standardized contracts, SF Compute lets buyers and sellers attach multiple properties to orders, accepting wider spreads and arbitrage for flexible buyers.

While AI training requires massive, centralized data centers, the growth of inference workloads is creating a need for a new architecture. This involves smaller (e.g., 5 megawatt), decentralized clusters located closer to users to reduce latency. This shift impacts everything from data center design to the software required to manage these distributed fleets.

SF Compute argues that for a true compute market to exist, it needs an independent operator that handles physical settlement. This means running the data centers to standardize quality and earn Wall Street's trust, much like an electrical grid operator.

Historically, data centers were designed and built like unique architectural projects. Now, the need for rapid, global scale is forcing the industry to adopt a manufacturing mindset, treating data centers like cars or planes produced on an assembly line. This shift creates a new market for production orchestration software beyond traditional factories.