Behind Cisco's bid to buy Acacia for $2.6 billion is a company that wants the benefit of owning one of its significant technology suppliers. But the story doesn't stop there.
As much as vertical integration and the ability to drive down costs are factors, so too is Cisco's ability to provide technology for newer markets that don't necessarily buy in the same way as telcos or large enterprises.
According to Cisco's head of optical, Acacia's coherent optical interconnect components business -- sold as embedded modules, pluggable modules and semiconductors -- gives Cisco more networking options to offer its web-scale customers that are building data centers and adding cloud capacity.
"Over the last couple of years we've certainly shifted our thinking and said, 'Look, we got to figure out how our customers want to consume this stuff and make sure we're delivering it in that way,'" said Bill Gartner, SVP and GM of Cisco's Optical Systems and Optics Group, in an interview with Light Reading on Tuesday.
Around 2010, Cisco purchased a transponder vendor called CoreOptics for $99 million in cash plus incentives. For a few years, Cisco was using that technology across its optical networking product lines, as well as absorbing those optical transponders into its routers. But the tech was starting to show its age. "Acacia has, over time, become more significant as a supplier to us," Gartner said.
Cisco accounted for 14% of Acacia's revenues in 2018. It wasn't even a 10% customer before then.
Indeed, Cisco is following optical networking trends by pursuing Acacia. "The coherent technology that has traditionally been delivered as part of an optical system -- on a line card that goes into a chassis, or on a purpose-built pizza box, like a DCI platform ... That technology is finding its way into pluggables," said Cisco's Gartner. "And when it finds its way into pluggables, it starts to become very interesting from a routing perspective because now you think about how you can simplify operations, or reduce complexity, for our customers," he added.
"Cisco wants to shake up the optical industry by building their own component business for inside and outside the data center by acquiring Luxtera and then acquiring Acacia Communications," said Don Frey, principal analyst covering transport and routing at Ovum, in an email to Light Reading. "The merger shows the strategic nature of pluggable modules with switching and routing as well as the future of the optics industry, as we see strong prospects in the ZR modules."
For interconnecting data centers and in metro applications, having pluggable optics in the router will, in some cases, eliminate the need for a separate transponder. That gives Cisco a seat at the table when competing against traditional optical transport vendors.
"Cisco's a relatively small player in the optical systems market segment because it's pretty fragmented, but we're a significant player in the routing market. So if that TAM [total addressable market] could shift [from optical systems] into routing and we could affect that shift, then we have an ability to participate in that market as a routing vendor," Gartner said.
Also driving Cisco to buy Acacia is the business of keeping up with the trends in data center networking, where the largest webscale cloud providers are the biggest customers -- and each one does things their own way.
"I think one of the things that's interesting about Acacia to us is that they sell their technology in a way that the customers really want to consume it," said Gartner. "So it's not like one answer for everybody. If a customer wants to buy DSPs or PICs and put them on their line card, they can do that. If they want to buy an embedded module that has it all integrated, they can do that. If they want to buy a pluggable, they can do that."
Webscale players aren't telcos and they aren't traditional enterprises, either. "They may want to pick and choose what technologies to use and how to integrate them," Gartner said. "And that was definitely something that was attractive to us about this [acquisition], that that model already exists."
The deal might not be as straightforward as it looks, though. "This merger has a few risks," Ovum's Frey writes to Light Reading. "Cisco will need regulatory approval from China as ZTE is a key Acacia customer. The current state of US and China’s trade talks may pose a risk to this closing. Building these types of modules requires scale, including volume greater than Cisco, so Cisco needs to convince the industry that it is a neutral third party, which they have started by keeping Luxtera on its own."
Gartner said the regulatory approval process shouldn't be an issue. "We do believe that we can we can get through that process," he said. "We don't compete with ZTE in the optical world, so I feel confident that we can give them very good comfort that we're going to continue to support their business. In fact, we'd love to expand that business."
Should current optical transport vendors who buy Acacia be worried? Not at all, Gartner said. "I think they can take comfort because Cisco will bring a very strong supply chain discipline and I think we can help with scale factors as well."
- Cisco to Buy Acacia for Pluggable Optical Goodness
- In Search of Optical Networking's 'Holy Grail'
- Arista Ships Smarter 400G Routers
- A Coherent Take on NGON & DCI World
Ray Le Maistre, Editor-in-Chief, Light Reading, contributed to this story.