Getty Images/iStockphoto

News

Cisco rides the AI boon

Cisco reported a 9% revenue increase, driven by AI demand. The company plans to launch new AI infrastructure products, focusing on enterprise data center trends.

Antone Gonsalves
By
Published: 13 Feb 2025

Cisco delivered strong quarterly results and expected even higher returns as businesses buy the company's gear and software to support AI in the data center.

On Wednesday, Cisco reported that revenue for the second quarter of the fiscal year rose 9% from a year ago to $14 billion, beating analyst estimates. For the current quarter ending in April, the company forecast revenue from $13.9 billion to $14.1 billion. The company raised its projections for the fiscal year by almost $1 billion, to between $56 billion and $56.5 billion.

Chuck Robbins, CEO, CiscoChuck Robbins

"We absolutely believe that a lot of this demand is being driven by customers preparing for AI," Cisco CEO Chuck Robbins told analysts during the company's earnings call.

The company reported roughly $700 million in AI infrastructure orders in the first half of the fiscal year, which ended in January. Orders came from enterprises and telecommunication companies refreshing their networks to deliver AI services, Robbins said.

Still, many enterprises are testing AI and trying to determine the best path toward the highest ROI, said Tom Nolle, principal analyst at consulting firm Andover Intel. Finding the return from AI will come before infrastructure purchases.

"While enterprises have been gaining insight into that issue, they're far from being there," Nolle said. "I think Cisco is simply trying to prepare an AI position, but they're in no position to drive AI deployments as a network player."

Robbins said he expects enterprise spending to increase as the company rolls out its AI infrastructure portfolio this year.

Upcoming products

Cisco plans to start shipping models from its recently introduced N9300 series of Smart Switches in the spring and summer. The hardware is powered by Cisco's Silicon One E100 network processor, which has 4.8 Tbps of capacity for handling traffic from data-intensive AI applications. For security, the switches have an attached AMD data processing unit (DPU) that runs Cisco's Hypershield, a distributed security architecture used to analyze data for threats.

By the end of March, Cisco plans to ship AI Pods, a preconfigured infrastructure stack that includes compute, networking and AI software. The scalable, prebuilt hardware for edge deployments includes Nvidia GPUs for inference, a process for fine-tuning AI models.

In May, Cisco plans to ship HyperFabric. The cloud-managed AI data center system includes Cisco servers, Nvidia GPUs and DPUs, Cisco 6000 series switches with 800G Ethernet, Vast Data storage and Nvidia AI Enterprise software.

The products built with Nvidia will primarily target inference in the enterprise data center, an emerging tech market that Robbins said he believes is Cisco's sweet spot. That market will be much bigger for Cisco than the training of huge generative AI models underway in cloud providers.

"We believe that opportunity is an order of magnitude higher than what we've seen in training to date," Robbins said.

Innovation from startups like DeepSeek, a Chinese company that developed an efficient model that doesn't require a massive number of high-end GPUs, would grow the enterprise AI market, Robbins said. "That is going to just facilitate a faster advancement of AI applications in the enterprise."

Revenue growth from Cisco's four product categories was led by security, which rose 117%. The second was observability, at 47%, and collaboration, at 1%. Networking revenue fell 3%.

Software from last year's $28 billion Splunk acquisition was the largest contributor to security and observability growth. Excluding Splunk sales, the pair grew 4% and 3%, respectively.

Antone Gonsalves is an editor at large for Informa TechTarget, reporting on industry trends critical to enterprise tech buyers. He has worked in tech journalism for 25 years and is based in San Francisco. Have a news tip? Please drop him an email.

Dig Deeper on Network infrastructure