There was an incredible amount of hype leading up to Cisco’s announcement of a new core router, the CRS-3. Typically, router announcements are not covered in detail by the mainstream press, but with Cisco’s marketing acumen, they were able to raise the level of interest to a crescendo.
Big deal? sure. Was it the revolutionary thing that will change the Internet? No. A revolutionary innovation would be if somebody suddenly announced an entirely new technology, like optical routing. Or that routers weren’t even necessary anymore. A true innovation is unfathomable, such as trying to imagine the World Wide Web in 1982.
This is an incremental upgrade of the traditional routing technology. It is, in fact, a crucial and necessary upgrade in capacity to an existing product that helps Cisco keep their edge in the core router market, but not an entirely new technology.
Here are the pertinent facts of the announcement:
* The CRS-3 will triple the capacity of its predecessor, CRS-1. Cisco says the new router will have a capacity of 322Tbps (terabits per second)
* Cisco said it spent $1.6 billion developing the router
* The product is in field trials and will ship toward the third quarter of this year. The price will start at $90,000
* So far, AT&T is the only service provider that has announced that it is testing the product in field trials.
This announcement is important because Cisco is the leader in the core routing category and it needs to be constantly improving the products to maintain its edge. There are three major core router competitors: Cisco, Alcatel-Lucent, and Juniper. Cisco and Juniper are considered the leaders, with Cisco holding the edge in market share. All of these competitors vie with similar technology, and they are in a constant race to leapfrog each other every few years with “upgrades.” You often see a pattern where one vaults ahead, then the other catches up.
Cisco made the leapfrog move to the CRS-1 in 2004, beating Juniper to 40 Gbps (Gigabit per second) interfaces. But since then, both Juniper and Alcatel have been more aggressive moves to 100 Gbps interfaces, once again putting the pressure on Cisco to upgrade, and that’s why it was time to hear Cisco’s upgrade plans. So the fact that Cisco is making this announcement now should not come as a surprise. The most important announcement in the upgrade is that Cisco will be moving to 100Gbps (Gigbabit per second) line cards, as this is an area where Cisco was actually trailing competitors.
One major claim where Cisco has said it’s the leader is overall capacity: It says it now has 12X the capacity of any competitor. I am waiting to hear back from competitors, but verifying such claims is extremely difficult. Companies can make all sorts of claims about overall capacity but you have to make sure you are comparing apples to apples, and not to oranges. In the comparison, would the routers be configured the same way? Often routers are strung together to boost capacity, and different companies count capacity in different ways. Really the only way to verify it would be through a series of long, complex tests.
Furthermore, routers are often evaluated by other sets of features, including reliability and software control. The analogy I would make is: Do you buy a car just based on the engine size?
In the long run, it’s funny to watch the financial press and mainstream media report on this. The announcement is not likely to have any meaningful impact on Cisco’s numbers for at least six months. After all, the product doesn’t ship until the Q3 of 2010. We are not likely to see a revenue impact until the end of 2010, more likely the beginning of 2011.
Where this is important is on the marketing front, where it reinforces the perception that Cisco is the leader in core routers and is not backing down. It could have a “freeze the market” effect on core routing by postponing the decisions of large service providers who may put off such decisions until they can test Cisco’s routers against competitors. If anything this would have a detrimental effect on the sales of routers by Juniper or Alcatel, but even this is not certain. This will be interesting to watch in the coming months.
After the news, Cisco’s stock was trading up .14 (.5%) to $26.26. Juniper was close to unchanged, trading up .07 to $29.67. Alcatel-Lucent was trading up .03 to $3.42.