Tl:dr We have reached agreement that “supercloud” is not a thing.
A note on reader relevance: this series of posts is extremely niche and getting nichier. But consider staying for the gratuitous IBM joke.
It seems we agree the “supercloud” is not a thing today. That is breakthrough progress. So the debate shifts to whether it will ever be a thing, and specifically become a thing in 2022.
Transmuting the nebulous “supercloud” into “a thing” is a big, audacious (if characteristically indistinct) bet given the inability and/or unwillingness to answer or clarify any of the myriad previously posed questions. There are dozens and dozens of unanswered questions from our previous three installments seeking the basic contours of the “supercloud” . Usually the way the marketplace of ideas works is after unveiling your brainchild, you answer questions and attempt to rebut objections. That stress test makes good ideas better and puts bad ones out of our misery.
It doesn’t help that the “supercloud” continues to evolve. Another day, another definition. If anything, it is getting vaguer:
Supercloud, from a buyer’s perspective is about business transformation (aka digital).
Hmm. GPT-0.1 in action?
Goldman Sachs is out as the “supercloud” poster child. Snowflake is in. We’ll get a Poster Child of the Week whiteboard to make future updates easy.
I continue to show immense restraint around postulated “supercloud” features, particularly cross-cloud recovery and locality metadata.
The last thing this industry needs is more jargon, especially half-heartedly championed jargon. For “supercloud” to become a thing in 2022 (and the clock is ticking), it needs to get far more precise and address the questions raised. Start with the basics of What is it? What is actually new? How does it relate to the existing world?