Cloud Computing

The term cloud computing means something different to everyone. To a few it relates to shared disk storage, to others it means software applications available via the internet, and to some a large numbers of processors that are loosely tied together to run distributed applications. Then there are those who insist that it is a combination of all of the above.

In terms of decision automation, cloud computing refers to the parallel execution of software processes, across a large number of machines, with some type of centralized management system to dynamically allocate resources and schedule the execution of rulesets and predictive models. By their very nature, intelligent systems require this newer cloud computing infrastructure. The existing software architectural model, upon which all of today’s transactional  systems are built, is that of a single process executing in a procedural fashion. But the  human mind works in parallel. Our very ability to think, rationalize, and even to dream, is predicated upon this basic fact.  It is just human nature,  that when trying to build intelligent systems the  cognitive computational model that engineers tend to mimic, is the human  mind itself.

There is one fallacy being perpetuated in the field of cloud computing today. That being the best (and most cost effective)  approach to adopting this new technology is via a hosted solution on the internet. One prime example is Amazon’s elastic compute cloud offering, dubbed EC2. Clients can provision and run applications on an unlimited number of virtual servers, on a pay as you go basis.  In effect, allowing any organization to increase their processing capacity based upon ever changing workloads.

What these vendors like Amazon omit telling you,  is that most organizations have a super computer sitting idle ninety-five percent of the time, right on their own employees desktops. Studies have shown that the average desktop PC is utilized only five percent (5%) of the time. This means that an organization with two hundred personal computers, having 64-bit dual core CPU’s (the most common PC’s built between 2006 and 2008), has over one teraflops (trillion floating point operations per second) of unused computing power. If you are a Fortune 100 company with a hundred and twenty thousand thousand desktop PC’s, you have about as many available CPU cycles as Blue Gene/L, the forth largest supercomputer in the world.  The moral of the story? Find yourself some cloud computing software that can re-purpose those dormant desktop CPU cycles.  Don’t let some IT vendor with a fancy two, three or four letter acronym, steer you any other way.

Herbert A. Lowe

Open Source Business Model

There was an interesting series of articles on Businessweek.com last month about the financial benefits of using enterprise level open source software.  It appears that the software industry is moving towards a business model that is very similar to that of pharmaceuticals. It takes a great deal of time, money, and effort to research and develop new software technologies, as is the same with FDA approved medicines. Innovative software products, that provide companies with a true competitive advantage, and which are able to be patented, should command a premium price in the marketplace. On the other hand, enterprise software whose features and functionality were first delivered, or whose underlying algorithms were first published, over twenty years ago, should be priced like generic drugs.

The open source model of freely distributed software, created by potentially thousands of programmers across the globe, for which minimal fees are charged for support, bug fixes and perhaps a warantee, is one whose time has come. Many commercial vendors today are going to need to rethink their pricing policies, as open source initiatives reach a critical mass, and the level of stability of the underlying source code becomes acceptable to business users. Open source packages now provide a significant portion of the features and functionality of more established proprietary enterprise offerings, for a mere fraction of the cost.

One example of the changing business landscape of the software industry, is the rise of innovative start-ups like Mobile Agent Technologies.  They have  aggregated five separate open source initiatives, including a rules engine, data mining and predictive modeling tools, a complex events processor, relational database, and content management system, and combined them with their own patented cloud computing environment. Together the subsystems form the foundation of a platform for building and deploying intelligent software applications, called Einstein Enterprise. Like Cirque du Soleil in the entertainment industry, who combined the most vital aspects of the circus with features normally found in broadway theater,  they have created a brand new market segment without any competitors. But at a price point at which they can be considered true “value innovators”.

Herbert A. Lowe

Machines That Think

As we move ever deeper into the worst economic crises since the great depression, we have to make the realization that bad decisions made both by mortgage originators, as well as securities firms underwriting complex debt obligations (CDO’s) , triggered the economic tidal wave that we are now experiencing.  Common sense would dictate that a mortgage originator should ask for employment and income verification when making a home loan, and that it would be an obvious trading strategy, to hedge ones bets by using credit default swaps when investing in risky mortgage backed securities.

Was it just pure greed? Was it simply poor judgment? Was it a failure to ensure that a company’s core business strategies were being followed by their employees? Perhaps it was a combination of all three.  The Wall Street Journal in December of 2008, had a complete itemized list of all of the “bailouts” and “economic stimulus packages” that the federal government instituted during the course of the year. Over a trillion dollars of taxpayer money has already been allocated to these programs with a good portion already having been spent. One CTO of a major investment bank confided in me recently that their business was no longer viable.

The financial crises which triggered the collapse or acquisition of a number of well respected Wall Street firms, could have been avoided by the use of automated decision making systems. Intelligent software platforms now empower computers with the ability to make faster, more accurate, and unbiased decisions, than their human counterparts. Decisions that allows follow company policy, strategy, and governmental regulations. The cost savings are significant, but what is even more important is that good decisions lead to increased profits,  corporate growth and economic prosperity, while bad ones oftentimes precede losses, downsizing, and recessions.

Herbert A. Lowe