Cisco Systems, one of the closely watched networking hardware companies on the planet, today made a pretty spectacular series of announcements: It's diversifying away from its core business model of selling integrated network systems and will independently sell a new proprietary switching and routing applications specific integrated circuit (ASIC), called Cisco Silicon One, which it will sell directly to customers with or without its own system, including its new routing software operating system, IOS XR7.
There are other parts of the news including big bets on integrated silicon optics, which result from acquisitions of Luxtera and Acacia Communications, as well as the new 8000 routing platform. But for the purposes of simplicity, I'm going to focus on the Silicon One announcement, which Cisco says required a $1 billion research and development (R&D) investment. This has big implications for Cisco's overall strategic direction, because Cisco may have just thrown its Applications Centric Infrastructure (ACI) under the bus.
First: The good news. Cisco says Silicon One will be an industry-leading chip platform. Better power consumption, better performance, and the first routing chip to scale higher than 10 Terabits/s. One of the stats it gave is 80% less power consumption, which will of course have to be independently verified. It’s a flexible architecture that can be used for different applications, whether its edge or core. It's also including hardware root-of-trust, an important security feature, in the chip. All good stuff.
But the shift in the business model —and certain powerful statements such as the embracing of Microsoft's open-source SONiC network operating system—pose serious risks to Cisco in the long term, because it's now placing all of its chips, well... on the chips.
Cisco, as everybody knows, is a sales and marketing machine. It has always focused on selling systems -- not individual components -- and has in recent years struggled to find new paths to diversify and grow. In fact, Cisco's revenue growth has been virtually flat over the past few years. So, the move to becoming a discrete chipmaker and seller is quite a departure. It’s a major shift in the narrative.
For the past several years, the story we heard was that Cisco's efforts in software-defined networking with its ACI would pay off as it shifted its revenue mix from hardware to annual software licensing. But Cisco is embracing cloud-native software that has grown faster than its own. It says it really likes Microsoft’s SONiC, which is another operating system other than Cisco’s. It also said that Silicon One would support P4 programming language, another popular engineering approach in the cloud community. Cisco is embracing cloud-native software approaches that have grown faster than its own.
All of this, of course, is the product of very difficult market changes. Over the past few years, Cisco and other networking systems players such as Juniper Networks, Nokia, HPE, and Ericsson (and many others) have struggled with large macro shifts in the market. The center of gravity in networking and compute infrastructure shifted away from service providers and enterprises to cloud and "webscale" players, who are now spending more on networking hardware and software. In addition, many cloud players, especially the largest ones such as Amazon, Google, and Microsoft, have become the innovators themselves, finding new ways to build infrastructure by piecing together their own hardware and software rather than buying from independent OEMs. At the same time, the service provider market has stagnated, as the workloads move to the cloud and capital spending budgets have flattened. That has made it real tough to be a Cisco, a Juniper, or an HPE.
Cisco highlighting Microsoft's SONic and programmable ASICs is no small thing. It means that Cisco has now acknowledged the cloud guys have the right approach and it is responding by making big changes by experimenting with selling separate OS and an ASIC, rather than completely integrated systems.
Cisco did a convincing job rolling out its largest customers—Microsoft, Facebook, AT&T, and Comcast, who embraced the new technology and touted benefits of Silicon One such as performance and power consumption. But there was almost no mention of Cisco’s software.
Half of those customers on the stage were service providers, which have been Cisco’s biggest financial pain points over the past years of its (declining) revenue. And the cloud guys seem to be doing fine with software. So that leaves Cisco trying to sell them optics and ASICs. What does that say about ACI and Cisco's software initiatives? In a sense, it's an admission of failure. And on a larger scale, it's also a huge shift in focus. Rather than transitioning to cloud software —an area that VMware and Red Hat (IBM) have mined quite well --—Cisco is doubling down on hardware, fortifying its DNA as a hardware player.
It's a huge bet. And quite a gamble. Cisco's biggest competitors are now chip companies. That ASIC had better work really well.
R. Scott Raynovich is the founder and chief analyst of Futuriom. For two decades, he has been covering a wide range of technology as an editor, analyst, and publisher. Most recently, he was VP of research at SDxCentral.com, which acquired his previous technology website, Rayno Report, in 2015. Prior to that, he was the editor in chief of Light Reading, where he worked for nine years. Raynovich has also served as investment editor at Red Herring, where he started the New York bureau and helped build the original Redherring.com website. He has won several industry awards, including an Editor & Publisher award for Best Business Blog, and his analysis has been featured by prominent media outlets including NPR, CNBC, The Wall Street Journal, and the San Jose Mercury News. He can be reached at [email protected]; follow him @rayno.
Industry Voices are opinion columns written by outside contributors—often industry experts or analysts—who are invited to the conversation by FierceTelecom staff. They do not represent the opinions of FierceTelecom.