First things first

If rumors are to be believed, Amazon Web Services is building its own networking gear, presumably to sell to customers that want a hardware-based foundation on which to run a hybrid cloud environment. True or not, I think it's a smart idea (and one I'm pretty sure I've advocated here before) as a way to truly own the hybrid cloud story -- a story that, despite years of public-cloud-only rhetoric, absolutely resonates with many enterprise CIOs.

A hardware-based approach to hybrid cloud is smart for two main reasons: (1) It gives the cloud provider the opportunity to optimize the entire package to its own specs (a la Apple products), and (2) cloud providers already know a ton about designing hardware and getting it built on the cheap. If, more than a decade into the cloud computing era, you still can't bring every workload over, then perhaps it's a good idea to take what you've learned about hardware and data centers and go to those workloads.

Networking gear would be an interesting first step (I assume most people's minds immediately jump to servers), but I can see the logic. The obvious point being that the network is the most natural place to manage traffic between private and public cloud resources. And despite the rise of SDN over the past several years, networking gear still has not been commoditized the way servers and storage have been. More room for differentiation means a greater opportunity to convince buyers they should lock into your platform at this layer, too.

But I also wouldn't be surprised to see cloud providers go down this route even deeper over time, selling servers, storage systems and perhaps even optimized gear for workloads like AI (complete with custom chips). It might sound crazy, but Microsoft, Amazon and Google are already selling their own consumer hardware right now, and their cloud businesses are no longer judged on how many just-bring-your-credit-card developers they sign up. With enterprise cloud sales taking on a familiar, difficult legacy feel as providers fight over large contracts, maybe cloud-branded hardware really is the next big opportunity to blaze new trails.

There are, of course, dozens of caveats one could bring up here -- including are a big one around the fact that CIOs are already fearful of cloud lock-in, hence the success of open source platform-level technologies like Kubernetes and Cloud Foundry, and Docker at the lower level -- but that's a topic for another post. For now, just revel in the idea of cloud providers selling hardware, and of the head explosions that will happen if they really do.

Facebook hires a chip expert from Google

On a related note, Facebook hired Shahriar Rabii from Google and has named him vice president and head of silicon. It's extremely unlikely that Facebook would ever start selling its own data center hardware, but it almost certainly is building its own chips for virtual reality headsets and possibly for its own data center workloads (including AI). There was a time not so long ago that a Facebook phone or other consumer hardware might have seemed possible, but I think that ship has sailed for privacy reasons.

Microsoft wants facial recognition regulated

And, building off of Facebook's privacy problems, Microsoft made a public stand on Friday by calling for facial recognition to be regulated by the government. Its blog post, bylined by Microsoft president and chief legal counsel Brad Smith, makes some good points and is definitely worth reading if you haven't already. Among other ethical rationale, he points to recent backlash over tech-company contracts with U.S. immigration agencies, and he lays out a series of tough legal and ethical questions that are far from answered right now.

But as with everything in the days of mega-platforms, it's difficult not to look at these types of proclamations with at least a bit of skepticism. You want to look for ulterior motives (for example, that Microsoft is better suited to handle heavy regulation than some of its competitors) or see how serious a company is about its moral code when huge contracts are at stake.

Still, I think Microsoft is right to point out the potential problems with facial recognition, and I think that companies considering rolling it out at any real scale should be very cognizant of what could go wrong. If violating privacy isn't bad enough, the legal landscape around data privacy is finally getting some teeth. Most CEOs can probably think of many things they'd rather do with 4 percent of their annual revenue than pay a GDPR fine.

ARCHITECHT

AI and machine learning







Cloud and infrastructure









Data and analytics