Maximizing the efficiency of software products

Oricane’s green software engineering maximizes the efficiency of almost any software products. Moore's Law Over the last three decades, Moore’s law has had a major impact on the software industry. Moore’s law states that the performance of computer hardware doubles, at a fixed cost level, every 18 month. It has simply not been necessary to engineer efficient software because of Moore’s law, since Moore’s law takes care of performance. The outcome of this is that most software today is not energy efficient. Improving Decision Intensive Software In general, software consists of thousands or even millions of interconnected calculation and decision processes. In a calculation process, little or no stored information is taken into account, whereas, in a decision process, huge amounts of stored information may be used. Some software is very calculation intensive whereas other software is extremely decision intensive. Some examples of calculation intensive software are digital signal processing, graphics rendering and weather simulation applications. Some examples of decision intensive software is packet forwarding, packet classification and data storage, searching and indexing. The Internet is highly dominated by decision intensive software at all levels, with applications ranging from packet forwarding and packet classification to indexing and searching the whole World Wide Web. Oricane’s green software technology is built on a foundation of several key innovations to improve the efficiency of the processes in decision intensive software. Reducing Energy Consumption The performance of a computer is mainly determined by the speed of the processor and the amount of memory the computer have. Since memory speed typically is proportional to processor speed, the power consumption can be modeled as ENERGY = PROCESSOR × MEMORY. The processor and the memory decide how powerful a computer is, but also how much energy it uses. When software is added to this, the decision processes in software will use memory and processor power, so software also consumes energy. There is an interaction between software and hardware and in order to reduce power consumption in software, it is necessary to minimize these two hardware costs. Ideally, the hardware should be exactly as powerful as needed to meet the performance requirements, but not more powerful. Basically, a computer can be thought of as a box, where the width of the box represents the processor and the height of the box represents the memory: It is rather easy to optimize for low memory cost, i.e., squeeze together the box at the height, but the outcome of this is that the processor cost explodes: Similarly, it is rather easy to optimize for low processor cost, i.e., squeeze together the box at the width, but the outcome of this is that the memory cost explodes: In fact, immediately when limiting one of the hardware costs the other cost grows un-deterministically and can become very large in the worst case when using existing technology. As a result, it has been necessary to over dimension computers in order to cope with the non-deterministic behavior of the software. Over dimensioning of computers directly leads to higher power consumption and considerable waste of energy. One of the key inventions by Oricane is to achieve low processor cost and low memory cost at the same time: Furthermore, the costs are not only extremely low but also completely deterministic. By using Oricane’s green software technology, the need for over dimensioning is heavily reduced, if not completely eliminated. Optimizing all decision processes in software results in radical improvements of the efficiency of a wide range of applications, ultimately reducing the power consumption of the hardware on which the Internet is built.