The 3 Key Elements of Modern IT Solutions: Centralized IT, Private Chains, Public Chains

Part 1 of the 3 Body Problem Series. See Overview and the whitepaper for more info.

Emergence of Blockchain in a Centralized World

New, and especially disruptive, technologies often position themselves as antagonistic to what has come before. In the early days of the public cloud, for example, it was common to hear cloud service providers assuring would-be customers that on-premise deployments were soon to be 100% eliminated, and that only fools would consider anything but a pure-play cloud architecture.

Didn’t quite work that way, did it?

As it turned out, public cloud adoption was a more incremental adoption pattern spanning decades, and even the most "all in" enterprise customers often still pursue a hybrid strategy that includes both cloud and on-premise elements.

Blockchain advocates took a similar stance when "their" technology emerged. Not only were the "private" and "public" chain camps at war with one another for supremacy, but both positioned themselves as a one-size-fits-all replacement for the entire IT stack. Of course, with some time and perspective, it became obvious that neither public nor private chains were a true substitute for the other, and that neither is a replacement for conventional (aka "centralized") IT architectures.

What did emerge is a playbook for how all three technology categories – centralized cloud and on-prem solutions, public blockchains such as Ethereum, and private/permissioned chains will interoperate to create best-of-breed IT outcomes over the next 20 years.

To understand these new design patterns better, let’s begin with a clear understanding of the continuum on which these technologies exist.

Understanding these technologies as points along a continuum

Understanding these technologies as points along a continuum.

It’s a Continuum...

Let’s start with the basics: Centralized IT implementations will always offer the highest throughput, lowest cost, and easiest governance for high-bandwidth data, because they’re at one endpoint of a continuum.

Conversely, public chains offer the broadest distribution model for limited amounts of data, but at the highest cost of transfer and storage. Between the two, private chains optimize for sharing large amounts of data in a reliable way with a fixed population – ideal for building partner data sharing solutions.

Blockchain advocates sometimes speak as if these distinctions will be swept away by "upcoming optimizations". But the fact of the matter is that this continuum is dictated by physics: Copying data to more places, further away from the starting point, costs more time, takes more infrastructure resources, and therefore costs more money.

To put it bluntly: Changing a row in 1,000 databases has to cost at least 1,000 times more than changing a row in 1 database. Move that lump around wherever you like under the rug, but it’s still there somewhere. However those costs may be presented or apportioned, there’s no way around the fundamental truth that, on a per-transaction basis, public chains cost dramatically more than a single, centralized resource holding the equivalent information.

This difference is exacerbated by the fact that not all information in a public chain is necessarily of interest to all of its users, while the overhead of worldwide replication is paid regardless of future consumption or access patterns.

So much for continuums. In the next blog post we’ll dive deep into the strength and weakness of each technology.