The End Of Massive Data Centers?

Fractal technology is one of several new tech stacks challenging the narrative that data centers will need nuclear reactors. We do not need data centers.

The overwhelming narrative in the A.I. investment community is that Microsoft, Google and Amazon will go nuclear - meaning they need nuclear energy to power the massive data centers for A.I. systems.

As Fractal proves every day, there is not a commercial application on planet earth that needs a data center.

The reason these massive data centers are needed is because every tech stack today is horrendously I/O bound.

Not everyone on this feed is a techie - and we do not write for techies. If you are a super geek, go to the website FractalWeb.App and you can read more about things like architecture.

Back to I/O.

I/O means input output. It is the bane of all computer processing today and the Fractal team reduced it to such a level that it eliminates the need for data centers - period.

What’s I/O?

Every compute instruction when you hit the enter key goes down the tunnel to the database and looks for your answer. The answer in your corporate billing DB might get name from one field, a transaction from a second field, time of use from a third - and each one of those is an I/O event.

This is no big deal to you - it happens in seconds.

But if you are running the billing system for 10 million customers, those I/O events add up. They need a massive data center to run those bills.

Those data centers consume about 4% of the entire electric energy used in the world.

Here is a customer story.

An electric utility has a billing system for 1.2 million customers.

It runs on a massive Oracle cluster - costing tens of millions of dollars a year.

It takes 4 hours every day - using all their compute power - to run those bills.

So at the end of the month, they have 4 hours a day clock time, for 24 days = 92 hours to generate those bills.

The billing system was rewritten in Fractal in less than 60 days.

It runs all the same calculations - it creates every bill, and since it is a parallel system, it checks the main system so every bill is correct before it goes to the customer.

Instead of a massive data center it runs on an Intel NUC, costing $4,000, about 8 inches, x 8 inches.

Instead of using massive power, it uses the power of a table lamp.

Instead of running for 92 clock hours, it runs for less than 5 minutes.

That is why the Fractal team believes, and is ready to prove, there is no need for a data center if I/O is properly managed.

For you new readers, if you think it is easy to sell this to large corporations, you would be mistaken.

We are often asked “…..so why doesn’t everyone use Fractal if it can save them literally billions of dollars a year?”

That is why we are writing this Substack - because we are experts in bringing massively disruptive technology to market.

We are the guys who built the eBay fraud engine when nobody could stop on-line auction fraud. Our team built the TSA No-Fly List underlying technology and the fraud tech for about 70% of the property/casualty insurance industry.

Disruptive technology is, well, disruptive.

So the people with empires hate it - they will never adopt a technology that threatens their world view.

When we show a CIO (Chief Information Officer) how to eliminate their data center costs, it is like telling a Navy admiral how to eliminate the need for his aircraft carrier - it is a short conversation - ending badly for us.

Thus, we do not focus on the CIO with an empire and a career to protect.

One source of introductions is private equity.

Those are the guys who step up to Southwest Airlines or AT&T and buy a huge stake in the company. They well know the bureaucracy at these firms has stymied innovation - and they can profit from cleaning a lot of it out.

Private equity also loves to take out costs which a company cannot themselves remove.

If you have a recalcitrant CIO running a multi-billion dollar organization, if he or she could have taken out the costs, they would have.

Private equity guys love to take out those costs - and since they sit in the board room it is a truly top-down selling experience.

Our customer is often someone who has a problem that simply cannot be resolved with conventional technology.

When relational database and other current obsolete technology cannot get you there, Fractal becomes a pretty easy sell for us.

Thus our focus on the electric utility industry as they need to replace very old systems fast and there is not the funding to do it. With Fractal we replace legacy systems in 90 days or less and we do it for less than 20% of expected costs.

We even built a website showing the systems we deployed - in 90 days or less - across multiple utilities: TheFractalUtility.com.

One of the things about massively disruptive technology is it wins every time.

It does not win some of the times, it always wins because when you can lower costs 50% - 80% and run 1,000 to a million times faster and deliver an application in 90 days one would expect to take 2 years - it’s a matter of execution.

The amount of blowback is both refreshing and predictable.

We always get the cubicle-dweller, some SQL jockey who knows his job is obsolete but somehow keeps it because he can run some old legacy system - that guy says we cannot possibly deliver these benefits.

So we do Proofs of Concept where we just build the application for the new customer and let them use it for a bit. When the results are in, usually in a matter of a few weeks, the cubicle guy knows his obsolete talents are on life support.

Our journey is inevitable - Fractal technology is just too cost effective to lose.

And since we are very profitable, we do not need venture capital nor do we need investors who are more trouble than having two spouses.

The only thing we need is to get in front of more customers with intractable computer problems - like high cloud costs, billing systems that don’t calculate a bill correctly, systems that run for days or hours but would be optimal to run in seconds.

So join us here on the Substack and enjoy the ride - with bumps and bruises along the way - and watch us disrupt the entire IT industry and have fun with us in the process.

---

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.