It might not be lights-out for Big Switch yet, but the company seems poised for some big changes.

Thanks partly to VMworld, the Big Switch rumor mill was in high gear late last week, with some sources saying an acquisition by Cisco could even be in the works. The going theory was a $100 million deal — less than half of Big Switch's valuation as of its 2012 funding round.

Other sources shot down that theory; a competing rumor says Big Switch got an offer from somebody but turned it down.

Neither Big Switch nor Cisco would comment on rumors, both citing company policy. But it's clear Big Switch needs a change, and a spokeswoman did acknowledge that the company is making an "adjustment" that might be revealed in the fall.

It's hard to tell how big that switch might be. It's too mild to be considered a pivot or a reboot, the spokeswoman says.

Big Switch's change of fortunes reflects how the thinking around software-defined networking (SDN) has shifted in the past two years. Anyone selling an SDN controller has to contend with the open-source version that the OpenDaylight Project is developing — a controller based on Cisco and IBM technology rather than Big Switch's.

More generally, the OpenFlow obsession has given way to more general talk of network overlays. For the moment, the spotlight is on VMware, which last week announced more details around its NSX network virtualization platform.

"We've seen overlay-based network virtualization definitely on the rise, and there's no question NSX is, out of the starting gate, one of the leading platforms," alongside PLUMgrid and Alcatel-Lucent subsidiary Nuage Networks, says IDC analyst Brad Casemore.

On top of all that, the networking establishment has been surprisingly quick to join the SDN craze. Cisco, in particular, has come up its own SDN-related acronyms, including the Cisco Open Network Environment (ONE) of APIs and the Application-Centric Infrastructure model (ACI) that spin-in Insieme is focused on.

Big Switch, meanwhile, has continued to focus on OpenFlow as a means of programming switches. The company submitted most of its Big Network Controller to OpenDaylight but was rebuffed in favor of the Cisco/IBM option — a clear setback.

Big Switch has also lost a couple of executives recently: VP of Engineering Howie Xu (whose departure might not have been voluntary, according to some) is now at Cisco, and Senior Director of Product Marketing Andrew Harding recently joined VSS Monitoring.

It adds up to one big spot of trouble for Big Switch, but that doesn't mean the company has to sell itself just yet.

Big Switch still has a shot because the ecosystem is still in flux. Vendors have piled onto OpenDaylight, but really, they're hedging their bets, looking also at VMware's NSX and other network virtualization platforms, Casemore says.

"It's definitely an ecosystem game. You need hardware vendors to support your software for things like visibility and federation and automated configuration," Casemore says. "They did a good job of framing that ecosystem early. Their challenge is that a lot has happened since then."

Why would Cisco want to buy Big Switch, anyway? Whether convincing or not, a couple of reasons come to mind.

First, source agree Cisco's primary interest would be Big Switch's people. But Cisco can already get the people; it's rumored that the Insieme spin-in has already picked off some Big Switch engineers by offering huge salaries. But if Cisco values the executives, an acquisition might be the only way to bring them into the fold and convince them to not bolt as soon as possible.

What about the products? A deal would give Cisco two SDN controllers, but that might actually be helpful, says Matt Palmer, one of the Wiretap Ventures partners who runs SDxCentral. Cisco could then offer the OpenDaylight Project’s controller (based partly on Cisco’s own technology) as an open-source option and the Big Network Controller (including the Floodlight portion of code, which was initially open-sourced) as a commercial option.

“With Big Network Controller, Cisco would have an arguably more mature controller for their commercial offering,” Palmer says. “One could then speculate that driving Beacon to OpenDaylight and then M&A of Floodlight/BNC was a brilliant move to fend off anti-trust.”

Among the bigger problems with a Cisco/Big Switch theory, though, is that a deal would be a bitter pill for Big Switch. The company was founded with disruption in mind and an apparent eye toward toppling Cisco (even if they didn't say that out loud all the time). Outright failure would just be a fact of startup life, but an acquisition by Cisco might be considered a deep defeat.