Earlier this week, The Information reported that Amazon Web Services (AWS) was considering getting into the network switch business and competing with the likes of Cisco, Juniper Networks, HPE and Arista with open-source white boxes priced up to 80% below what Cisco charges for its network switches.
The share price of Cisco, Juniper and Arista plunged on the news – which is remarkable when you realize that the report is just a rumor (one that Amazon has officially denied). And even then, it’s a rumor that Amazon was simply entertaining the idea of selling its own network switches, not that it had already decided to do so.
It says a lot that the mere suggestion of Amazon simply pondering the possibility of entering your industry sector is enough to freak out your investors. We saw similar reactions when Amazon bought Whole Foods and entered the retail grocery business, as well as its recent entry into the pharmaceutical sector (via its purchase of online startup PillPack).
They arguably have good reason to be spooked. This is after all what Amazon does – it shoehorns its way into new industries and disrupts the hell out of them. Even AWS started off as a wannabe public cloud services challenger that Google and Microsoft initially laughed at. They’re not laughing now.
Meanwhile, an Amazon push into network switches is entirely plausible, partly because AWS already makes its own custom white boxes for its data centers (i.e. commodity hardware with custom software), and partly because open-source white boxes are a hot growth market right now – thanks, of course, to hyperscale players like AWS.
Another factor, it seems, is Jeff Bezos’ reputed desire to own the universe.
AWS: the Hotel California of cloud services?
Which brings to mind David Cheriton, a Stanford Professor, entrepreneur and investor who spent part of his keynote at a NetEvents conference in San Jose this past May explaining that he is currently backing Apstra (a start-up specializing in intent-based networking and automated data center network management) because he believes private clouds are a viable alternative to the AWS public cloud model that can also compete with AWS’ efficiency levels – provided they have the right hardware and the right automation and management solution.
However, he said, when he told people this, they replied that there was no point investing in a company that enables enterprises to automate and manage their own networks because everything is moving to public clouds – more specifically, Amazon’s public cloud.
Cheriton’s main beef with AWS – apart from his contention that it’s not a ‘public’ cloud any more than Disneyland is a public park – is that it offers additional services on top of its computing services that end up serving as a lock-in mechanism for customers.
“These start-ups and big companies end up where they develop some application and it’s using a hundred of Jeff Bezos’ little services that are unique to AWS,” he said.
And that makes it very hard to move that app out of AWS, he added. “You can’t get off of AWS without rewriting the application.”
As it happens, it’s possible that an AWS network switch play would accomplish a similar result by encouraging AWS customers to buy AWS switches designed to offer superior performance and functionality when used with AWS but not (say) Azure, as RadioFreeMobile’s Richard Windsor points out in this blog post.
That’s not to say an AWS network switch play wouldn’t have issues. For example, as NetworkWorld points out, one advantage Cisco, Juniper and other vendors have over AWS is tech support.
And yet it’s not so much about the practical details of an AWS switch business as the fact that Jeff Bezos can simply just go ahead and do it if he feels like doing it.
“If [Jeff Bezos] decided to start a bank tomorrow it would instantaneously be one of the largest banks in the world. There’s 100 million Prime customers at this point with Amazon,” David Cheriton said in his keynote. “He does want to own the entire universe here. He’s got incredible capabilities to move into almost any industry.”
That’s why Amazon is such a force of nature in the ICT world – it’s not just about what it does, but rather what it is capable of doing. Amazon can disrupt your industry if Bezos feels like it. And even where it can’t, it doesn’t matter because Amazon can afford to fail.
Talk about disruptive.