Aviz Networks Podcast Series

In this episode of the Aviz Podcast, host Ilona Gabinsky sits down with Louis Toth, Co-Founder & Managing Partner at Alter Venture Partners and Aviz board member, for a candid conversation on where AI is headed next, and what it means for founders, investors, and the infrastructure powering it all.
With over 30 years of venture capital experience, Louis shares his investor's lens on the forces shaping AI's next wave: the remaking of the edge, the rise of AI factories, the hidden opportunity in networking, and the rapidly shifting rules for what makes a startup defensible in 2026.
What you'll learn:
1.) Why AI is a secular trend, not a bubble
2.) How the edge tech stack is being rebuilt from the ground up
3.) What makes Aviz uniquely positioned in the AI networking space
4.) The 3 filters Louis uses to evaluate every early-stage startup
5.) Why networking is the most underrated layer of the AI boom
6.) How founders can build companies that survive a faster-moving market

The AI era is rewriting the networking playbook, and the software opportunity is bigger than anyone expected. In this episode of the Aviz Networks Podcast, Ilona Gabinsky sits down with Alan Weckel, Co-Founder and Analyst at 650Group, to unpack why AI networking is on track to exceed $200B by the end of the decade, with Ethernet leading the charge and software making up nearly half the value.
Key takeaways:
1.) AI networking is no longer just a hyperscaler story. Neoclouds and enterprises are driving unprecedented growth
2.) Open networking is accelerating as operators connect frontend, backend, scale-up, and DCI fabrics
3.) Observability, telemetry, and AI for NetOps are becoming non-negotiable for scaling operations
4.) The path ahead points to self-driving networks powered by deep telemetry and private AI At Aviz Networks, we are building the open, end-to-end software stack for the AI era, spanning Open Networking, Deep Network Observability, and Private AI for NetOps.
Watch the full conversation to see why networking has never been more exciting.
#AINetworking #OpenNetworking #NetOps #AIforNetworks #DataCenter #Ethernet #SONiC #NetworkObservability #AIInfrastructure #AvizNetworks #650Group

In this episode of the Aviz Podcast Series, host Ilona Gabinsky sits down with Kevin Wollenweber, Senior Vice President and General Manager of Data Center and Internet Infrastructure at Cisco, to discuss how AI is reshaping networking requirements and why SONiC has become critical infrastructure for hyperscalers, neo clouds, and enterprises.
Kevin shares his perspective on what he calls the "networking super cycle," drawing on nearly 30 years at Cisco to explain why the current pace of AI adoption is unlike anything he has witnessed before. The conversation covers the growing role of SONiC as an open source network operating system, why Cisco has become a major contributor and investor in the SONiC ecosystem, and how the Cisco and Aviz Networks partnership is helping bring SONiC to customers who lack the in-house engineering resources that hyperscalers have.
Topics discussed:
1.) Why AI is accelerating hardware and software adoption cycles faster than the internet or cloud transitions
2.) The shift from large language models to agentic workflows and physical AI
3.) Why SONiC has reached mainstream adoption in data center and AI infrastructure
4.) How Cisco and Aviz Networks are partnering to simplify SONiC deployment for enterprises and service providers
5.) The rise of agentic workflows and their impact on network utilization What skills and mindset will matter most for networking professionals in the AI era
This conversation offers a candid look at how networking, often overlooked in AI discussions dominated by GPUs, is becoming the critical enabler of AI infrastructure at scale.
Aviz Networks builds end-to-end AI software networking solutions and partners with organizations like Cisco to accelerate open networking adoption.
#SONiC #AInetworking #Cisco #OpenNetworking #DataCenter #AIInfrastructure #NetworkingPodcast #AvizNetworks

Ilona Gabinsky talks with Mark Longwell, Director of Telco and Edge Alliances at Red Hat, about how AI factories are reshaping infrastructure and why the Red Hat and Aviz Networks partnership matters.
Highlights:
1.) It is not your father's data center anymore. AI shifts the mindset from software first to infrastructure first.
2.) GPU racks now draw 50 to 100 kilowatts, roughly 10x a CPU rack.
3.) The biggest misconception about AI infrastructure is that it is easy.
4.) When networks and applications run in silos, the customer feels it first. Aviz Networks unifies the network with deeper observability and telemetry on the Red Hat platform.
Chapters:
00:00 Introduction
03:38 Why AI factories change infrastructure
05:54 The biggest operational challenges
08:53 The biggest misconception
10:05 Breaking down silos
12:25 Where the partnership goes next
14:14 Advice for starting your AI factory journey Subscribe for more on AI infrastructure and open networking.
#AIFactory #AIInfrastructure #RedHat #AvizNetworks #OpenSource #Networking #DataCenter

What Is an AI Factory? Networking, Orchestration, and the Future of GPU Infrastructure
In this episode of the Aviz Podcast Series, host Ilona Gabinsky, VP of Marketing at Aviz Networks, sits down with Mohan Atreya, Chief Product Officer at Rafay, and Thomas Scheibe, Chief Product Officer at Aviz Networks, to unpack what it really takes to build and operate an AI factory.
The conversation covers why an AI factory is fundamentally different from a traditional data center, and why networking, not just GPUs, is the piece that ties everything together. Mohan and Thomas explain how multiple networks (compute, tenant access, and storage) must be coordinated to keep GPU clusters running efficiently, and why multi-tenancy, automation, and orchestration are essential to operating at scale.
They also break down the Rafay and Aviz partnership in detail, showing how a programmatic API integration allows GPU provisioning and network configuration to happen automatically, without manual tickets or human bottlenecks. Along the way, they introduce a new way of thinking about the cloud concept of a VPC, reframed as a Virtual Private Factory for the AI era.
The discussion closes with practical advice for enterprises and neo clouds just starting their AI factory journey, along with predictions for how infrastructure design will evolve as deployments scale toward gigawatt-level capacity.
Topics discussed:
1.) What defines an AI factory versus a traditional data center
2.) Why networking is central to GPU infrastructure performance
3.) Multi-tenancy and secure resource isolation
4.) The Rafay and Aviz API driven automation model
5.) Rethinking VPCs as Virtual Private Factories
6.) Monitoring GPU and LLM level metrics alongside infrastructure
7.) What AI factories will look like in the next two years
Subscribe for more conversations on AI infrastructure, networking, and the future of AI native data centers.
Learn more about Aviz Networks: https://www.avizNetworks.com
#AIFactory #Networking #AIInfrastructure #GPU #CloudComputing #Aviz #Rafay #AINative #DataCenter #Podcast