Are the big clouds getting rich by minimizing R&D in new tech, all while optimizing their infrastructure to squeeze fatter margins out of thinner innovation? That seems to be the contention made by Streamlio’s Jon Bock, and he might have a point. Certainly one of the chief complaints from open source vendors is that the clouds profit from their innovations without helping to fund it with cash or code.
On the other hand, as Jay Gordon has argued, each of the major clouds is going “far beyond maintain[ing]” old tech and is creating “a lot of building new blocks for devs/ops/etc.” Indeed, it’s increasingly difficult to point to innovation that is emerging from legacy tech vendors, with the innovation (and the spoils) seemingly centered on the big clouds.
SEE: Cloud providers 2019: A buyer’s guide (free PDF) (TechRepublic)
Nothing new under the sun
It’s perhaps too easy to overlook this innovation, given that some (much?) of it has been buried in older technology. As Gordon calls out, “A new way to host a database…isn’t so fun or interesting to all but damn can it make a difference for someone struggling to manage scale/costs.” Hence, one could dismiss Amazon Aurora as “MySQL or PostgreSQL running at cloud scale,” but that overlooks the reality that while Aurora may have much in common with those venerable databases, running the way it does and at the scale it does makes it a completely different product.
And yet there’s a ring of truth to Bock’s suggestion:
Another challenge is the business models of most cloud providers are generally high volume with lower margin. Those margins are viable because there’s less money invested in core technology development, most of the R&D investment goes to operations & automation.
As such, he posits, “in a large part of the general technology market, R&D investment in developing new product technologies has shifted from big companies to smaller ones.” In the open source world, this has been true of MongoDB, Confluence (though Kafka was born at LinkedIn), Cloudera (Hadoop born at Yahoo), etc. Outside open source, the more interesting companies doing new things with streaming data, for example, are also startups.
But is this truly new?
Making it work at scale
After all, hasn’t it generally been the case that startups often sprout the latest and greatest, then larger companies step in to commercialize it at scale? Despite enormous R&D budgets, it generally wasn’t IBM, Oracle, etc. who gave us the best new tech of the 1990s. If Microsoft, AWS, and Google aren’t releasing cutting-edge innovation, it wouldn’t seem fair to call them out.
SEE: Microsoft Build 2019: The biggest takeaways (free PDF) (TechRepublic)
And yet they are innovating. One can engage in an academic argument about how new containers are, for example, but it’s Google that wrote Kubernetes, which most everyone is using to manage containers at scale. And what about ML/AI? Nearly all the interesting tech is coming out of the clouds, not least of which is Google’s TensorFlow. Or how about serverless? While there is great open source technology here (OpenWhisk, for example), it’s the clouds that are really driving innovation.
Such innovation, in turn, is making it much easier for developers to do their own innovation. Without high upfront software or hardware costs, coupled with a smorgasbord of rich services, developers are spoiled for choice, and are innovating at a frenetic pace.
The cloud is not innovating? Hardly. Rather than bemoan a lack of cloud innovation, the inverse is probably more worrying: The tremendous amount of innovation, and the lock-in to which it leads.