As businesses take their first steps towards adopting AI, one of the first advice they receive is to address the quality of their data. However, while often overlooked, another fundamental element that is critical to the success of AI deployments is network infrastructure.
At Cisco Live, ZDNET spoke with SVP and GM Anurag Dhingra of the Enterprise Connectivity and Collaboration Group to learn more about the role network infrastructure plays in AI Revolution. Dingra emphasized the importance of taking action as he does not see a future in which network infrastructure could be an afterthought.
Also: AI agents are around, but not autonomous – what does that mean for us
“The reason network infrastructure is a bottleneck for AI is that you can already see the span of AI agents,” Dinla said. “Can you imagine multiple agents acting like humans generating more traffic at machine speed and at machine scale?”
AI agents appear everywhere
The age of AI agents that can perform tasks on your behalf is beginning. AI Agent Orchestration works together to support humans by accessing the same resources that humans do, such as the web, YouTube, and more. It will inevitably test the organization's network infrastructure, as it adds a lot more individuals who want the same connection, whether people or agents.
result? A similar experience that you get when you go to a stadium where too many people compete for the same service. Organizations need to plan this shift by allowing their network infrastructure to support more users as well as more complex and constant load profiles.
Beyond the widespread use of AI agents in the workplace, there is also work to make models more specialized, smaller, cheaper and less computational requirements. These developments allow these models to run locally on the device.
“These two things come together and lead to agents appearing everywhere in the workplace. It's not just a data center,” Dhingra said.
How can businesses prepare?
Whether it's a router or a wired access point, companies often consider them as short-term investments when considering investing in network infrastructure. However, this is changing as AI development is growing rapidly. A Cisco survey found that 97% of businesses think that their network needs to be upgraded to make their AI and IoT initiatives successful.
“Organisations should consider network infrastructure as an enabler of AI capabilities,” Dhingra said. “I would like to invest now so that next year or the next year, not to make this a bottleneck.”
Also: How Cisco Stops Attacks in Network
Dhingra warns that companies may see bigger pictures and regret not expanding in a year or two to meet the AI-enabled productivity they are targeting. Rather, he said it would benefit businesses from purchasing infrastructure that could expand.
At the meeting, Cisco unveiled the latest generation of routers and switches aimed at keeping up with AI conversions and workloads in the workplace. These include newer versions of the Cisco family of routers, 8100, 8200, 8300, 8400, and 8500, and newer versions of the campus LAN switches in the Cisco Catalyst family, 9350 and 9610.
Get the top stories of the morning in your inbox every day Tech Today newsletter.
