Brute Force Computing

In framing strategic plans for a business entity, product line, or service, a key consideration should be to make "brute force" computing work for you, as opposed to working against you by better arming your competitors or creating new competitors. Effectively applying brute force computing is an under-appreciated art, but very much part of the Colts Neck Solutions portfolio.

Looking ahead, it is remarkable is that this rate of increase continues to apply even though after many doublings the current absolute capacity/performance numbers are in the billions - PCs operate at gigahertz speed, disk drives offer gigabytes of storage, etc. Paolo Gargini of Intel recently predicted that Moore's Law is likely to pertain for at least another 15-20 years.

If one computer isn't enough, offerings variously
termed "computing on demand" and "grid
computing" - dynamically leverage multiple
computers with idle capacity - open up new
vistas of raw capacity, relevant if one can exploit
that capacity for business benefit and competitive
advantage. See "bucket reduction" for further
specifics as to how to exploit them..

The comparative surprise has been the presently
breakneck improvement in networking prices and
performance. Breakthroughs in wide area networking somewhat lagged computing trends, because changing telcom infrastructures involves complex processes and decision-making, but today networking is far more of an enabler than a constraint.

At some point, such changes in scale - bigger, faster, etc. - creates a change in kind. The jet air transport is conceptually not much different from the Wright Brother's early biplanes, but theie speed, reliability and capacity created the modern air transportation industry. It is only in recent years that computing speed, reliability and capacity have passed the point at which they can become widely transformational.

In Darwinian terms, evolution is a matter of species adapting to new conditions - warmer, cooler, wetter, etc. In the "punctuated equilibrium" view of evolution ("punk eq"), changes may come rather dramatically rather than being spread over millions of years, so species need to adapt equally quickly to reach a new equilibrium. "Time to market" was important long before humans.

In "punk eq" terms, a few years ago computer price and performance crossed the threshold at which computing became powerfully "democratic." Today, all players in the modern economy, almost anywhere in the world, can afford unconstrained business computing power. "Unconstrained" means that an entity can afford enough computing power and associated networking reach to achieve a competitive, or even leading edge, level of business process automation.

Democratic computing is overtaking a world order in which the big predators had "powerful" mainframes and could do things that little predators could not do at all or could do only in a very constrained way on low end mimicomputers and PCs. Of course large companies and other large institutions can and do buy bigger and more computers than small companies, but the small company can buy enough computing power to sustain its business, with plenty of headroom.

At today's prices, the pacing factor is intellectual - in defining how to apply raw computing power in ways not previously feasible. We are in many respects in a "catch up" phase, and thought and experimentation is needed to make creative, practical use of new capabilities.

In a world that rewards full factor productivity and "time to market," the road to success is based on trading off increasingly affordable computing and networking to economize on more expensive resources, land, labor and non-IT capital.