John Chambers on the future of Cisco

Cisco hit the skids 18 months ago in the wake of the economic slowdown. The economic malaise hit the company hard and led to a restructuring that was completed in record time. Now, Cisco is ready to roll again, and to celebrate, CEO John Chambers discussed the company’s future with half a dozen reporters.

The big picture is that as the network becomes the pipeline running through modern society, his company stands to gain — if he can mange the transitions.

Lucky for Cisco, it’s not like the transition to an all-IP world connected by networks is a sudden shift. Change is happening in waves, and Chambers hopes to take advantage of as many waves as he can. Already, Cisco has managed to successfully predict and deliver products for the transition to video on consumer and corporate networks. In four years, Chambers believes video will dominate corporate communications — and even IT.

I’ve whittled down the 90-minute discussion into three key takeaways (and reported earlier on why Cisco didn’t buy Skype) that our readers will find most interesting. The topics range from the failure of the Flip buy to Cisco’s plans for the data center. Let’s begin in the home.

The changing set-top-box market.

Cisco is not, I repeat, not selling the set-top box business it purchased in 2005 when it paid $7 billion for Scientific Atlanta. Chambers denied it at least three times during the call, but he also said that in five years, he’s not sure content from service providers will be delivered from a box. Instead, it will be all done in the cloud.

Cisco's home videoscape product.

On it’s face, that seems like an admission that the set-top box is headed for the scrap heap, but Chambers is right that service providers are not going to give up on boxes inside their subscriber’s homes. None of the executives inside telcos or cable companies are eager to provide a pipe into the home, they all want some kind of hardware to ensure a device where they can deliver software, services and any other value add that will help them stay above the commodity pipe business. The only question is will that box be a modem, set-top-box or a residential gateway. Whatever it is, Cisco will be happy to provide it.

Speaking of the cloud, Cisco has a (secret) plan there too.

Chambers was mum on his plans to address the dual threat of an entirely new data center customer that cares mostly about price and scaling out, as opposed to the nifty software and plug-and-play aspects of Cisco’s UCS gear. He was also pretty quiet on the looming threat of software-defined networks, although he pointed out that Cisco is a member of the Open Networking Foundation and will support Open Flow. However, the real threat of Open Flow and SDNs is that they will undermine Cisco’s architecture play.

Cisco's UCS server.

In the data center, Chambers emphasized the customer wins at service providers (about 45 percent of the customers) and Fortune 500 companies. Cisco is betting that a system that goes from chip-based to software and services will win out. So far it’s doing well, but I also think it’s missing out on the next big wave of growth in the server market — companies like Amazon, Facebook and Yahoo that buy their servers by the racks (and even by the data center) as opposed to by the box.

Those companies are trying out merchant silicon in switches such as those sold by Arista, or even lower-level gear from the likes of companies like Huawei or Quanta. Chambers’ response to the rise of merchant silicon and commodity boxes is to bring on the ASICs. That’s right, Cisco plans to keep designing its own chips as part of its architecture play — but it also plans to design them faster. When I asked Chambers to go deeper into this, he declined.

It’s possible a custom ASIC could bring value to Cisco, even in the webscale world, if used a little bit more strategically. For example, SeaMicro is a hot startup in the server space because it’s building a box that could replace 500 machines from five years ago and run at 96 percent of the power. SeaMicro can do this because it has developed its own ASIC to handle the networking inside that box.

When I pointed this out, and asked if Cisco could deliver a fundamental rethink of its ASIC to deliver that type of functionality, Chambers said he wasn’t willing to discuss this further, so as not “to let all his competitors know where Cisco was going.”

But M&A is in the air.

Chambers seemed frustrated that no one asked about Cisco’s recent announcement that it planned to start buying other companies. He didn’t really offer any insights about who or what he wanted to buy, but on Friday Cisco announced the planned acquisition of LightWire, a company that makes an optical transceiver.

He also acknowledged Cisco’s mistaken efforts to get into the consumer market with the Pure Digital acquisition (otherwise known as the Flip camera buy.) He said when his team realized the deal wasn’t working, they quickly decided to shut it down, and did so, which is possibly the best way to spin a $590-million mistake.

Indeed, Chambers appeared to be the master of being able to admit his mistakes, and then spinning them quickly to an asset. Whether it was his ability to recognize quickly that the $590 million buy of Flip wasn’t going to work or his realization that an effort to change the pricing and product mix on switches led to a sudden drop in margins that he has since fixed, Chambers hasn’t met a market transition he couldn’t either see coming or fix in retrospect with quick action or an acquisition.