CarahCast: Podcasts on Technology in the Public Sector · Carahsoft

Networking Reinvented: A Conversation with Nokia

·39 min·2 clips

Nokia’s EDA can tune an AI fabric on demand and raise inferencing tokens per second.

1. CarahCast: Podcasts on Technology in the Public Sector frames this conversation around Networking Reinvented and Nokia’s role in public-sector data centers. 2. Scott Robach hosts Steve Vogelsang, the CTO of Nokia Federal Solutions, whose background includes networking work since the 1990s and a startup sale in the early 2000s. 3. The episode asks what Nokia can do for federal data centers as AI, automation, certifications, and security requirements reshape the market. 4. Vogelsang says Nokia is better known for wireless, but the company has shifted toward infrastructure after selling off its smartphone business. 5. He says the Alcatel-Lucent merger brought IP routing and optical transport strength into Nokia, and that data center became a newer portfolio area about six or seven years ago. 6. He describes AI data centers as different from traditional facilities because the workloads run on GPUs and require a network that cannot drop packets. 7. He says those AI fabrics must adapt to learning, tuning, and inferencing workloads to maximize utilization and improve performance. 8. Robach says many network automation projects are bespoke, and Vogelsang answers with Nokia’s event driven automation, or EDA. 9. Vogelsang says EDA uses a native Kubernetes-style infrastructure so the network can be managed like an extension of server workloads. 10. He says that approach can increase tokens per second for inferencing by tuning the fabric to the specific data set or problem space. 11. Vogelsang says Nokia Federal Solutions exists because Nokia is a Finnish company and needs a U.S.-based proxy entity to hold security clearances. 12. He says Nokia now has routing and optical transport products certified and is working on certifying its data center switching portfolio. 13. He also says the company needs FIPS certifications and other approvals to sell into DoD networks, and that the process requires test labs and long backlog waits. 14. Robach applauds the commitment behind FIPS 140-3 and DODIN APL work, and Vogelsang says Nokia is seeing positive DoD engagement. 15. When Robach raises public cloud and NeoCloud models, Vogelsang says DoD is more open than before and that GovCloud already offers FedRAMP-compliant infrastructure. 16. He says the next major growth area is inferencing data centers, which will likely matter more than model training as AI usage expands. 17. On edge computing, Vogelsang says Nokia supports edge cloud out to the tactical edge with 5G deployments and edge AI workloads for battlefield use. 18. He says Nokia’s portfolio includes high-end IP routing, optical transport, subsea capabilities, and data center systems that need to work as one end-to-end network. 19. The tone stays technical and collaborative, with Scott Robach pressing for specifics and Vogelsang answering in a direct, explanation-heavy interview style. 20. People working on federal AI, network automation, encryption, or data-center design would likely benefit most, while listeners seeking personal storytelling or light discussion may skip it.

As heard by us

A practical look at federal networking, cloud constraints, and certification demands.

This episode treats networking less like a buzzword and more like infrastructure work, using Steve Vogelsang's path from 90s-era networking to Nokia Federal Solutions to show how public-sector data centers are being pushed by AI, automation, and federal certification rules.

Read the full review in PlayNext →

Why you'd press play

You want the federal networking forecast behind AI-ready data centers.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from CarahCast: Podcasts on Technology in the Public Sector