Enterprise companies looking to build or upgrade their data center switches, or even get technical support, best be prepared to wait in line behind AI-first providers that are chewing up most of the available resources.

That’s the upshot of a recent Gartner report, which paints a grim picture of the data center switch market for the next year. Switch vendors are “aggressively pivoting resources” toward AI data centers for a simple reason: They are more profitable than traditional data centers.

“Spending on AI network fabrics will surpass general-purpose data center networks in 2026 and more than double through 2029,” the Gartner report says. “Consequently, the market is splintering into two distinct buying patterns: AI infrastructure networks and general-purpose data center networks.”

Resources at the disposal of switch vendors are not keeping pace with demand, says Andrew Lerner, distinguished VP analyst with Gartner and one of the report’s authors.

“Switch vendors have human resources in terms of engineering talent, and physical resources like memory and lasers for switches,” Lerner says. “All of those resources are constrained.”

Only about 200 companies operate high-end AI data centers, while some 100,000 organizations have more traditional data centers. But collectively, those AI players spend three times as much as the smaller companies do, Lerner says.

“If you’re a vendor and you’re doing what you’re supposed to do, you want to capture the growth,” he says.

Zeus Kerravala, founder and principal analyst with ZK Research, agrees. “Cisco, Arista, Juniper and those companies that build data center equipment, make no mistake, their resources are directed towards AI first because they want to be part of those big buildouts,” he says. “There’s a lot of money being poured into neoclouds, things like that. They’ve reprioritized the resources based on where market demand is.”

Price hikes, long lead times, sketchy support

The repercussions for companies with traditional data centers include higher prices, long lead times, and perhaps subpar support.

Gartner predicts switch price increases of 15% to 40%, largely the result of resource constraints, and lead times of three to nine months, up from one to two months in mid-2025. Constraints should ease by around the middle of next year, but don’t expect prices to come down.

“Generally speaking, vendors have no consistent track record of reducing prices in these networking markets,” Lerner says.

At the same time, with vendors dedicating scarce engineering talent to AI, they likely won’t invest in significant innovations for non-AI switch families. The same goes for support.

“As vendors reallocate their best personnel to support large AI build-outs, the support quality for general-purpose environments will decline,” the Gartner report says.

“Vendors are taking a ‘keep the fire running, don’t let the flame go out kind of mentality,’” Lerner says. They’re not abandoning those 100,000 traditional data center customers, but they will dedicate their best talent and resources elsewhere, he says.

Enterprise data center survival tips

Gartner’s advice for weathering the switch storm, or mitigating its impact, starts with sticking longer with the assets you have. “A lot of organizations refresh before they need to,” Lerner says.

Think beyond the vendor’s end-of-life policy to factors such as whether your current equipment meets your needs and is easy to operate, he says. Also consider the level of innovation in the market, and whether you’ll really be getting meaningful new features or functions with new switches.

“Most organizations generally err on the conservative side, so they refresh too early,” he says.

Another consideration is to achieve better utilization of your current switches. Lerner says it’s common to see only half of the 48 or 36 ports on a top-of-rack switch utilized. Similarly, companies can increase VMware density ratios to increase both server and switch utilization.

If you do need more equipment, check out certified refurbished programs offered by vendors such as Cisco and HPE. Such equipment comes from sources such as returned leases and bankruptcies and is available through authorized channels, Lerner says.

Another tip is to explore other vendors. “Not all vendors are impacted the same way, and not all vendors have the same portfolio,” he says. A company with a high-end switch portfolio that has access to only a limited amount of memory will dedicate that memory to its high-end switches. But a vendor that specializes more in lower-end switches doesn’t have to make that decision.

That means smaller enterprises may be spared the worst of the switch shortage. “The mid-market enterprises are probably in better shape than large enterprises as far as availability goes, because the large enterprise vendors are the ones that have had to pivot,” Kerravala says.

Companies should also explore various architectural options, including moving more workloads to the cloud or colocation facilities, Lerner says. Or be more flexible with respect to the class of switches you need, in terms of port count and speeds, perhaps using smaller switches in some areas of the data center.

If push comes to shove and you must buy new switches, Lerner has some simple advice: “Order early.”