There Is No Cloud - It's Just Someone Else's Data Center

Timesharing had its run in the 70s, cloud computing is the same thing and will collapse the same way

Anyone remember Tymshare, Service Bureau Corporation and NCSS?

If you are a techie, over 65, you may recall them as the Microsoft and Amazon of the late 70s.

They were the timesharing giants - although giant meant something quite different in those days.

Timesharing was a way for corporations to get their computer work done - forecasting, financial simulation, personnel applications and they used software called Express, Focus, NOMAD and some others.

Timesharing was a growth business and those companies were doing well building vertical applications like a personnel management system or a rail car management system - then sold by the transaction or the hour license fees to corporate customers.

There were some real problems with timesharing that were not obvious early on.

One of them was “vendor lock in.”

When a company spent half a mil building a financial forecasting system in Express, there was no way to take it in house later. Express didn’t port.

In the early 80s IBM, DEC and a host of other companies started delivering mini-computers - and became the behemoths of their day. At least DEC did, IBM was already there.

Companies saw no reason to continue to outsource their computing, so they brought it in house and started the mad dash to building huge IT (info tech) staffs still in use today.

Timesharing died in a 4 year timeframe - from a very hot product domain to companies literally out of business - that fast.

This was disruption at scale - and it was something to behold.

Today you hear about the “cloud.”

People, there is no cloud - it’s just someone else’s data center.

The cloud is nothing but a massive marketing phenomenon which will level off, become marginally profitable in the late part of this decade. It has to because it is 100% the same technology that everyone has in house already - it’s just on someone else’s computer - at a far higher cost.

We see CIOs all the time bullshit their board that they are going to get their companies into the forefront of computing - with agile development, lower costs, the ability to get out of the data center business - and over the last two years, the reality has come home to roost - it’s not happening.

Companies who moved to the cloud found they could neither downsize nor eliminate their data centers.

They found cloud costs were much higher - maybe twice the costs of running the same application in house.

Firms learned about vendor lock in big time - using those “value added services” of the cloud vendors means they are stuck there - or they have to re-engineer those applications costing more than their original migration to get them back in house.

We at Fractal have a prospect - their Amazon bill is $40 million a year, the workflows necessary to make their portfolio work are slowing down app/dev (building new applications) and their board is mad as hell.

We will rebuild their application in less than a quarter, it will run 1,000 times faster - which is important in their business - and we will charge them $5 million a year - not $40 million.

When you go to your board, show them a $35 million saving - in a current budgeted amount - delivered in two quarters - you just changed your career trajectory.

We have another group who refuses our call - because they spent $350 million going “cloud-native” which means they built their applications to work only in the cloud - their costs are skyrocketing, their Fortune 500 company is financially struggling - and they will lose their jobs if the CEO hears about Fractal.

Unfortunately, while their CIO will not speak with our team, one of our team members’s brother is the chief pilot for the CEO executive jet!

These companies are hitting what we call the despair curve - shown below:

Cloud computing is slowing - or at least its upward growth is slowing - but it is still growing - although at a slower pace.

Nobody, and we mean that literally, has digitally transformed their business with cloud computing - something they mindlessly sold their boards.

Even McKinsey - who publishes reports from a rear view mirror - noted 70% of all “digital transformations” fail. Those are almost all in the “cloud.”

App/Dev is maybe 5% or 10% faster - because in the cloud you can provision servers with a click rather than filling out forms with a useless in-house IT help desk.

None of the benefits sold to their boards are happening and we see it because we get the call when they hit that despair wall.

The “cloud” requires data centers so large firms are talking about needing their own nuclear reactors.

Those of us in Fractal land just laugh at that nonsense.

We run applications that consume an entire data center on a 4 inch, by 4 inch by 4 inch Intel NUC costing $6,000.

A major hedge fund brought us in to test Fractal and see if we can actually do this - I guess we did OK since they are now taking us into deals.

You dear reader are at the early stage of a story that will be conventional wisdom in 36 months.

Data centers are essentially obsolete - as one of our senior engineers notes on our website Fractal-Computing.com, there is not a single commercial application we have ever seen that requires a data center - and we have 40 years of looking behind us.

In the early 1900s, there was the belief that the telephone could not grow beyond the numbers of switchboard operators necessary to plug in all those lines. You have seen the pix of the pretty girls, in white blouses and dresses, by the scores, taking the call, plugging in the wire to complete the call.

Well, someone invented the rotary phone and these operators disappeared. Or they went on to something else.

Those days are ahead for the cloud.

The cloud will never disappear - it is going the way it was originally intended - to be the elastic compute stack needed for retailers to use during that 90 day period when their compute needs quadrupled - in the holidays.

It makes sense to run those applications when needed on someone else’s data center - not have those resources unused 9 months a year.

The cloud as it is being sold today is obsolete - any technology that is built on the relational data stack is at end of life.

Here at FractalComputing.Substack.com we will publish case histories of firms who are moving off the cloud and finding real digital transformation - not the mindless despair they are currently getting from someone else’s data center.

You will be able to see real cases with metrics of the savings - and remember, we were the ones who first told you - there is no cloud.

It’s just someone else’s data center.

FractalComputing Substack is a newsletter about the journey of taking a massively disruptive technology to market.  We envision a book about our journey so each post is a way to capture some fun events.

Subscribe at FractalComputing.Substack.com

Fractal Website:  Fractal-Computing.com

Jay@FractalWeb.App

Portions of our revenue are given to animal rescue charities