The Next War Will Be Waged With Software - Not Just Tanks

The next war will be waged with software - Fractal Is the force multiplier

The next big war, (and a bunch of little ones) will be waged by software-enabled weapons.

Advantages will favor whomever deploys them fastest, neutralizes adversaries’ combat power most effectively, whether kinetically or non-kinetically, and to the combatant that defends its critical systems from attacks in-kind.

Don’t take our word for it, here is an article from a group of military people, in this case the Navy, making the argument:

https://www.hudson.org/technology/software-navy-new-warfighting-advantage-bryan-clark

Software was a competitive differentiator in the Fortune 500 starting decades ago.

That kind of software required large data centers, central points of control, moving the data to the computing hub - not computing where the data resided.

Those data and compute models with reliance on robust data lines-of-communications (D-LOCs) remain to this day – and create critical weakness for exploitation by adversarial targeters.

Software determining winners in warfare is here - navigating drones, providing satellite reconnaissance and targeting, determining if a Navy destroyer should shoot a $3 million missile at an incoming object or take it down with $4,000 in Minigun ammo.

Software determines scale.

The more software can replace moving parts in a flying object - whether manned or not, whether a returning weapon system or a one-way missile - the smaller and faster the object becomes.

Speed, cost, size are all determined by software.

Just look at your cell phone - and the “bricks” of 30-years ago.

Modern phones have operating system software, and chip-control software (“firmware”), together with advances in manufacturing that drive miniaturization and put enormous capability in your pocket, on your watch, or into the swamp of micro-drones jamming an enemy’s air defense systems.

Technology is changing the character of warfare and software is disrupting the world of technology.

If one combatant has software that can locate, characterize, and neutralize an opponent’s weapon systems before those systems can be employed - it’s game over.

A multi-billion-dollar weapons platform can be annihilated by 200 lines of computer code, in a distributed mesh, creating thousands of independent tiny programs, each operating at silicon speed.

The vectors determining winners and losers - those who live or die - are simple: cost, speed, and size.

Let’s take a walk through the software available to warfighters today and analyze cost, speed, and size.

For cost, how much does it take to build a family of war-winning software applications?

That includes training, development, equipment, data center, network and scores of other variables. In the U.S. government, costs include the iceberg of hidden friction with current government procurement - keeping out innovation and doubling down on 30-year old software.

For size, what does it take to deploy that software?

Does it need a data center, or does it need a chip?

Does it need large water-cooling infrastructure and raised floors or can it operate with the power of a tiny battery?

Speed of course is everything.

Software needs to input millions of instructions - at silicon speed - process vast amounts of data - and deliver a target’s coordinates with real-time updates.

There are other criteria – connectivity, scale, flexibility, survivability and security.

Connectivity?

Does the software need a central point-of-control to issue or relay instructions?

If a hub or node is destroyed or degraded, is your offensive capability neutralized?

Do those platforms become hazards and induce collateral damage if already deployed?

Connectivity to other stuff is a pretty big deal.

Connectivity - if required, opens one combatant to having a central point-of-control - or vulnerability - where that glaring vulnerability becomes the enemy’s top target.

Like a data center or cloud (someone else’s data center). Take it out and everything reliant on it collapses.

You’ve all seen the sci-fi movies - take down the alien mother ship and the drones fall to earth. Those scary aliens would have been better served if they implemented distributed computing.

They were data center centric, so it was easy to stop their alien invasion.

If America’s armed forces continue to adopt a data center or cloud (someone else’s data center) software defense strategy, the country will be at major risk.

So let’s put down “connectivity” if it is mandatory - as a fatal flaw - the software better be able to be survivable - no matter how many nodes are removed, destroyed, every other node remains in the fight.

With a true distributed software weapons system - even one surviving node instantly reconfigures the rest and the fight is on.

Then there’s security.

The same weapons can be used by both teams - often in war, one army captures the cannons of the other and turns them on the former owners. Hacking a weapon system not only raises that possibility, but it can also render one side completely unarmed with a click.

Today’s data center centric strategy of moving the data to the processing, rather than the processing - to where the data resides - as in the firefight - increases vulnerabilities. Data centers use mostly a moat security strategy - big, deep, well protected moats protect - but if an adversary breaches, bad things happen - at scale.

Distributed security - every node expecting to live in a threatened, sovereign attack envelope - protected down to the lowest level - will quickly emerge as the war-winning software security paradigm.

Then there’s scale.

He who processes the most data, in the shortest time - yielding a target - wins.

That’s kind of it.

That takes us to A.I.

For those of us doing A.I. for 40 years, the stuff today remains quite rudimentary.

Just get on the chat for any government agency or corporation using A.I. in customer service and you can see it is a long way from accurately aiming a drone.

A.I. will get better - but A.I. is just software.

A.I. is one piece of software, processing gargantuan amounts of data to get some result.

People can take every article written on the Civil War, put them in a great big data center, build indices and you could get a compelling term paper on how General Stonewall Jackson, had he lived, would have taken Culp’s Hill at Gettysburg and today Jefferson Davis would have a holiday in half the country.

Now that we mentioned A.I., let’s revisit cost, scale, speed, and connectivity - none of those requirements went away.

Warfighting software, A.I. or other software cannot be data center centric. Period.

First there is the cost.

Data centers take forever to build Large Language Models - and our Chinese friends already showed they can deliver those without a large data center. So bigger data centers are not the answer - even though every legacy software company will disagree.

If the data center is in Virginia and the fight is in Yemen - you have the latency of sending signals back and forth - and other issues.

Data centers are hugely costly, requiring among other things, up to 450,000 gallons of water per day to cool them.

The reason governments require data centers is to support their massively large relational databases that are the foundation of much of our current warfighting capability.

Trying to deliver warfighting capabilities, back in the U.S. - with a data center using 40 year old relational technology is a sure way to fall behind - with catastrophic consequences.

Those same relational databases lack the inherent architectures to operate at the speed of modern warfare. This creates critical and strategic liabilities in the physical infrastructures required to support and operate them.

Adding to the conundrum are the speed and inflexibility at which they operate, update, adjust, and engage.

Data centers and relational technologies are slow, inflexible, difficult to change - but most of all, as many U.S. Government agencies are reporting this month - cannot scale to the level needed for even current administrative applications.

The Gartner Group, the leading technology organization who gives peeks around the next tech corner said the current tech stack - which is large data centers running relational databases - is obsolete in 2025.

That day is at hand.

Whether referencing John Boyd’s OODA Loop or the kill chains we invest in to beat opponents to the draw, a data center, relational database strategy will not ensure our dominance in modern warfare.

More frightening, it will impede America’s investment in what works, in distributed, super computing at the point of the fight.

The entire data center model and its 1980s software are not suited for the dynamic and nearly instantaneous data access, manipulation, dissemination required for Information Dominance, a core foundational requirement for our current and future C5ISR architectures and the weapons platforms that rely on them.

Further, those who capture this high ground will close key and critical capabilities gaps with us, and may put them into closer proximity to us as near-peer competitors in various warfare domains.

In a hypothetical conflict with China, allied data centers reduced to cold-iron are metaphorical carrier battle groups adrift and dark – and likely the initial targets to be compromised in the opening round of such an engagement.

So how does the United States win wars with the best software?

Or better, how do we preclude such an event by ensuring durability and resiliency so an attack on critical national defense infrastructure - taking out key capabilities - is simply unattainable for all adversaries, small and large?

Here are a few ideas presented this week to some very serious expert warfighters in Washington DC.

From their enthusiastic feedback, they liked what they heard - so we thought we would share it with you.

We aren’t allowed to discuss specifics about our visit, but if there were a common refrain it was:

Fractal is the A.I. force-multiplier and fundamentally transforms the offensive and defensive battlespaces. America must own it.

Our background isn’t in warfighting, but we took those comment as a positive sign.

Fractal was originally built for the federal government - to handle massive databases, faster than any other technology, and be completely survivable.

Here is why our new friends made such a comment:

1. Fractal is fast.

Fractal runs any application from 1,000 to a million times faster than on any current computing technology.

Fractal runs at quantum speed, on current hardware. Today.

Several of our previous Substacks about I/O optimization address the concepts and design. This is our core mastery, and it is based on 40 years of research and implementation.

When America’s software runs - on a chip - a targeting system - 1,000 times faster than its adversary, well, you want that.

2. Fractals are tiny.

A single Fractal instance can run on a commercial grade mobile phone.

A collection of phones running multiple instances increases that compute and response significantly. Adding phones (in this instance) around the world would harness them into the world’s largest super-computer.

This is not our core (current) focus, so don’t look for it any time soon.

It does, however, illustrate how Fractal runs on a tiny chip - delivering supercomputer functionality in the most hostile conditions, with the ability to massively distribute and survive – while making its collective power available at the computing edge.

No data center needed.

3. Fractal systems can be built for 1/10th the cost of any conventional system.

That’s a good thing. It has a flip side - that means our adversaries must invest and manage at least 10 times as much resource for the same functionality.

Even then, their functionality will not run anywhere close to Fractal speed. With these implementations, the U.S.A. should maintain the compute-centric warfighting advantage.

4. Fractal systems require virtually no power, drawing minimal resources, we often run them off of tiny batteries.

This paradigm shift is a far cry from the requirements of facility-reliant data centers and their large power, water, structural, security, and operational requirements.

5. A Fractal system is nearly indestructible.

Every Fractal is a 100% copy of every other Fractal, except for the data.

If a system has 1,000 Fractals, 999 may be destroyed while the single remaining Fractal can reconstitute the main system - in seconds.

That’s a pretty handy feature in a firefight.

In a warfighting scenario, we may be fielding thousands to tens of thousands of Fractals - thus none of the vulnerability of a data center.

Fractals are kind of like a swarm of tiny super compute drones - without the propellers - and almost impossible to destroy. Again, a pretty handy warfighting feature.

6. Fractal is foundational.

Fractal can take any corporate or government application and run it without a data center – at significantly faster rates. Check out our videos and some of our micro-sites and we describe some of our demonstration systems.

Fractal also has major uses in streaming, transactional security and validation, and other domains we are exploring in parallel.

That means Fractal has disruptive uses protecting communication networks, authorizations and scores of other emerging areas.

7. Fractal is secure.

Traditional computing security assumes there is a castle and a moat that keeps the barbarians outside the gates.

Fractal assumes it is on a compromised server, in a hostile environment, and with bad actors intent on breaking into it.

Fractal is “sovereign secure” - withstanding well-funded and resourced government actors - not “just” kids in the basement - intent on penetrating critical systems and profoundly valuable data.

8. Fractal can deliver the largest A.I. models without a data center.

So while the Chinese are touting building Large Language Models with smaller data centers, and American software companies are raising half a trillion dollars to destroy the Virginia landscape with hyper scale data centers - Fractal builds the equivalent of Large Language Models - distributed into Small Language Models - effortlessly - on 4 - inch by 4 -inch computers you can hold in your hand.

When we started this newsletter, we were working in a handful of utility companies and running some state voter rolls and showing the world how NGOs were undermining American elections.

Today we are working to add Fractal to some of the world’s largest distributed communication systems, working with several state governments to eliminate voter, Medicaid and other program fraud, and getting the acceptance from key government actors.

We have in a way, crossed the Rubicon. At least we are on the bridge.

We will be reporting here on upcoming implementations which are massive in scale and implication and simply cannot be done with any other available technology.

At the request of hundreds of subscribers - that’s you - we are going to do some podcasts with Fractal engineers to answer the questions you regularly send in. Make sure you subscribe and you will then have access to them.

We want you to be assured that as software takes an even greater role in shaping and dominating the battlespace, those in positions to recognize Fractal’s potential to reshape how we deter, defend, and defeat are taking note and we are pleased they are doing their jobs.

We can’t talk much about our new friends, but you can be assured they get it.

---

FractalComputing Substack is a newsletter about the journey of taking a massively disruptive, sustainable computing technology to market.

Subscribe at FractalComputing.Substack.com

Fractal Website: Fractal-Computing.com

Fractal Utility Site: TheFractalUtility.com

Fractal Government Site: TheFractalGovernment.com

Jay@FractalWeb.App

JayValentine.com

Portions of our revenue are given to animal rescue charities.