“What does an AI data center look like? It does not look like that. You don't have random arrival. Links are consistently fully utilized. GPUs want to be fully utilized. Traffic patterns are actually much more predictable than they ever were before, particularly in training.”
a16z Podcast · Startups & Venture · November 2026
Context
Asked what changes in networking for AI, Volpi explains that internet networking was built around bursty, unpredictable traffic. AI clusters break that assumption, which he says opens room for cheaper network designs and threatens incumbent vendors.
Transcript
…out the workload, and you would under-provision the device to connect more bandwidth because not everybody all communicates at the same time through the thing. And the outputs of that inbound are highly variable. Okay, that's the internet. That's the workloads that we live with. That was almost
dogma, right? For a long time. Okay,
so fast forward today. What does an AI data center look like? It does not look like that. You don't have random arrival. Links are consistently fully utilized. GPUs want to be fully utilized. Traffic patterns are actually much more predictable than they ever were before, particularly in training. They're incredibly predictable. They're less predictable in inference, but you can see the traffic pattern, especially if you do good batching. I think of
it is still very predictable, right? Yeah, especially if you do big batching. Yeah, exactly. Like
you can definitely make it more predictable. So the fundamental tenet of the workload has changed. Yet, if you look at the dominant networking architecture today, which is the Broadcom, what's it not Trident? The Tomahawk. Tomahawk, which is the Dune acquisition that they did, is a StatMux packet…
So essentially, back to circuits, is what you're saying. Yeah.
More from a16z Podcast
