Dive Brief:
- Enterprises are grappling with overtaxed networks as they attempt to scale AI initiatives, according to a Broadcom report published Monday. Dimensional Research surveyed more than 1,300 networking, operations, cloud and architecture professionals for the semiconductor and infrastructure software company.
- More than half of respondents acknowledged their networks couldn’t support AI’s bandwidth and low latency needs. Nearly 9 in 10 of the organizations surveyed reported inadequate network visibility due to blind spots in internet and cloud environments.
- While network readiness hasn’t put a damper on AI enthusiasm, it’s undercutting adoption efforts, according to the research. “Ready or not, enterprises are going full throttle on AI,” Mike Melillo, senior director of network observability solutions at Broadcom, said in a release accompanying the report. “For network professionals, this is both good news and bad. The network is the foundation of AI strategy, but 95% of teams can’t see what matters, putting their AI initiatives at risk.”
Dive Insight:
Despite an abundance of deployable generative AI use cases, the technology is pushing legacy networks to the limits of their capabilities. As enterprises adapt to high-capacity compute needs, they are looking to service providers for assistance.
More than three-quarters of organizations have turned to outside firms for operational support, Broadcom found. Nearly half of respondents said their company has an outsourcing strategy for networking, and 43% acknowledged they lacked in-house expertise.
While Big Tech is responsible for the bulk of global spending on AI infrastructure, enterprise investments in hardware to power the technology are growing, John-David Lovelock, Gartner distinguished VP analyst said. The analyst firm expects global AI spend to approach $1.5 trillion this year, with hyperscaler spending on chips nearly doubling the size of the AI server market, and enterprise adoption pushing the market for AI services to more than $280 billion.
Enterprise networking investments have gained momentum over the past year after a prolonged post-pandemic slowdown, according to Dell’Oro Group market research. After a Q4 2024 rebound, data center switch sales surged mid-year, as enterprises resumed networking projects, Dell’Oro Group Research Director Siân Morgan said in the October report.
Cisco saw orders for its AI-optimized networking gear mount during the first quarter of its 2026 fiscal year, which ended Oct. 25. The company’s revenue increased 8% year over year to $15 billion, with networking leading the way, CEO and Chairman Chuck Robbins said during a November earnings call.
“We know many customers still have a lot of work to do to ensure they have the modern, scalable, secure networking infrastructure to support their AI goals,” Robbins said. “Only one-third of organizations feel their IT infrastructure can accommodate the needs of their planned AI projects.”
