10x Performance Increase

Locality Optimization™ Technology

10x Server Performance

Locality of Reference

Optimizing for Locality of Reference is all about reducing I/O wait states. This optimizaiton occurs at two levels.

System level   Optimizing the distributed system so that within each running process, the process needs to only access data that it has stored locally. By avoiding network requests in the middle of computation loops, the overall speed of the distributed system can by improved by multiple decimal orders of magnitude.

Process level   Optimizing sequential data access so that pipeline processing/caching is utilized to efficiently move data from persistant storage to RAM, to layer-2 cache, to CPU registers.

10x Development Speedup

Locality of Logic

Locality of Logic has existed as a concept for quite some time. It is usually implemented in the form of object encapsulation (object oriented programming).

The basic idea   Is that a programmer only has to understand the inner workings of an object in the code that implements the object. When other programmers use the object as a building block in a larger application, they only have to understand how the objects works “from the outside” and do not need to understand the details of how the object is implemented internally.

Does not translate   While this is beneficial at the object level, object encapsulation does not translate equally well to the system level.

Reduce the number of DB schemes   In a System Oriented Programming™ deployment, both through the software run-time environment "plumbing" as well as design and implementation methodologies that are encouraged/enforced by the programming framework, the design of enterprise class applications is done with a vastly reduced number of database schemes. In System Oriented Programming, the number of encapsulations in an enterprise class application fall from 100's to typically fewer than 10 .

App logic is embedded in DB schemes   System Oriented Programming dictates that application building block functionality/logic is implemented locally in a few schemes in what would be referred to as "stored procedures" in the relational database world. These stored procedures have Locality of Logic; the application logic is co-located with the storage scheme definition and typically requires knowledge of only that scheme or just one or two additional schemes.

Analog to Locality of Reference   In much the same way that Locality of Reference improves computation performance by not requiring access to remote data during a calculation – Locality of Logic requires only local knowledge of a database scheme instead of knowledge of an intractable number of potential interactions between relational data tables.

Multi-model   Optimizing for Locality of Logic, means that database layer code presents multiple data models and compute models to the application developer. This enables the application developer to use the data model that most “naturally” fits the part of the application the developer is working on.

Eliminating source of complexity   A major cause of complexity in traditional enterprise applications, is the convoluted code that forces application data into a (relational) data storage model that may not be the most natural, intuitive, or concise way to represent the data. By making it easy for multiple data models and compute models to co-exist, the application code becomes more intuitive, expressive, and smaller. Building blocks of application logic can be embedded in the database scheme itself instead of having to sit above layers of code that massage the data into a form that is consumable at the application layer.

Development speed   The reduction in complexity that results from optimizing for Locality of Logic makes application layer code significantly easier/faster to write. Empirical results show about a 10x reduction in time to develop complex applications.

Architecture Innovation

System Oriented Programming™

Next evolutionary step in enterprise-scale system design and deployment. Distributed microApps™ beyond the cloud.

Fundamental building block   Is a "system" which has:

  • GUI widgets,
  • distributed processing middleware (essentially a MapReduce variant),
  • a web server,
  • a shard & partition manager,
  • a paged memory manager & stream processor, and
  • a multi-model database engine.


Each "system" instance   Is a small vertical software stack running in a loosely coupled process that is an independent processing entity (a "system") in a distributed processing environment. The software framework gives programmers explicit and granular control over sharding and partitioning of the distributed data. The software framework also constructs software data processing/caching pipelines to efficiently move data to the CPU core during run-time processing.

The performance   of System Oriented Programming applications is due to Locality Optimization™ technology:

  • At data prep time (essentially "system compile time"), for Locality of Reference at run-time when the distributed processing system is doing real work, and,

  • At software developmen time, for Locality of Logic to reduce and simplify application-layer code.

Best use cases   are enterprise-class applications that support mission critical repetitive business processes. Examples of this are: billing, forecasting, reconcilliation, and other customer care oriented applications.

Emperical results   Locality of Reference optimization can reduce I/O wait-states in application layer code by multiple decimal orders of magnitude (compared to a similar legacy application implemented in a relational database). We have customers with production applications that are now running 1,000x to 1,000,000x faster than their legacy application.

Simultaneous innovation   System Oriented Programming consists of simultaneous innovations in:

  • Distributed processing,
  • Database architecture,
  • Stream processing,
  • Object oriented programming,
  • Fractal System™ architecture,
  • Full stack development frameworks (at macro and micro scale), and
  • Compiler design.

Architecture Innovation

Fractal System™ Architecture

Next evolutionary step in enterprise-scale system design and deployment. Distributed microApps™ beyond the cloud.

Fundamental building block   is a "system". Each "system" instance is a small vertical software stack running in a loosely coupled process that is an independent processing entity (a "system") in a distributed processing environment.

Each system contains the entire application   A user, or another application, interacts with an individual system instance via web services. The system contains all of the data and code (including all user interface widgets) for the application that the user wishes to run. Each system instance is a smaller scale version of the overall application.

Billing example   if the user is running a billing application, an individual system instance might be responsible for on the order of about 25,000 customers and all of their associated information. If the user asks the system to "show a list of all customers" the system instance will respond with a list of 25,000 customers. If the system instance is running in a System Oriented Programming™ environment where it can communicate with all of the other system instances in the environment -- then before the system instance sends a response back to the user, it broadcasts a message to the other system instances asking them if they also manage any customers. The system instance takes all of their responses and folds them into its internal response before sending the final response back to the user. The mechanisms for dissiminating requests out to other systems and integrating their results into a final result is part of the infrastructural framework of the System Oriented Programming environment.

Spreadsheet example   If the user wants to build an application with a "spreadsheet" implementation style, each system instance is responsible for a section of the spreadsheet. Often these type of spreadsheets are organized as multiple "sheets" within the overall system-wide spreadsheet. Each system instance is response for it's assinged sheet in the spreadsheet. This System Oriented Programming approach to spreadsheet implementation makes is possible to have infinite spreadsheets that enables a simple application implementation style (spreadhseet programming) to scale to arbitrarily large data sets and to utilize massive amounts of parallel processing. Each system instance manages its iSheet™ sheet in the aggregate system-wide spreadsheet in parallel and the System Oriented Programming framework provides the programmer with Locality Optimization™ tools to insure parallelism for optimal run-time performance.

Fractal System™   architectures have this unique characteristic that is extensively exploited in System Oriented Programming. Each system instance contains a fully functional application with all of the data and application logic for performing some task. The user experience of interacting with an individual system instance or the entire network of system instances is exactly the same -- the result sets are either a fraction of the total data set (when interacting with only a single system instance) or the entire data set (when interacting with a system instance that, in turn, is communicating with the other system instances).

All system instances are the same   A Fractal System architecture is a true peer-to-peer architecture. A user can attach to any system instance in the network to initiate a request. Each system instance is capable of communicating with all of the other system instances and assembing a complete result for the user. A Fractal System architecture enables massive parallel processing, fault-tolerance, and high availability of the overall application since failover and backup logic is also embedded in each system instdance as part of the System Oriented Programming infrastructural framework.

System Oriented Programming

Key Concepts

Distributed MicroApps™

Beyond The Cloud