Modern parallel computing techniques are subject to poor scalability. Their performance tends to suffer diminishing returns and even losses with increasing parallelism. Some methods of intelligent computing, such as neural networks and genetic algorithms, lend themselves well to massively parallel systems but come with other drawbacks that can limit their usefulness such as the requirement of a training phase and/or sensitivity to randomness. This thesis investigates the feasibility of a novel method of intelligent parallel computing by implementing a true multiple instruction stream, single data stream (MISD) computing system that is theoretically nearly perfectly scalable. Computational dreaming (CD) is inspired by the structure and dreaming process of the human brain. It examines previously observed input data during a 'dream phase' and is able to develop and select a simplified model to use during the day phase of computation. Using mazes as an example problem space, a CD simulator is developed and successfully used to demonstrate the viability and robustness of CD. Experiments that focused on CD viability resulted in the CD system solving 15% of mazes (ranging from small and simple to large and complex) compared with 2.2% solved by random model selection. Results also showed that approximately 50% of successful solutions generated match up with those that would be generated by algorithms such as depth first search and Dijkstra's algorithm. Experiments focusing on robustness performed repeated trials with identical parameters. Results demonstrated that CD is capable of achieving this result consistently, solving over 32% of mazes across 10 trials compared to only 3.6% solved by random model selection. A significant finding is that CD does not get stuck on local minima, always converging on a solution model. Thus, CD has the potential to enable significant contributions to computing by potentially finding elegant solutions to, for example, NP-hard or previously intractable problems. / Master of Science

Computer Architecture is entering an era where hundreds of Processing Elements (PE) can be integrated onto single chips even as decades-long, steady advances in instruction, thread level parallelism are coming to an end. And yet, conventional methods of parallelism fail to scale beyond 4-5 PE's, well short of the levels of parallelism found in the human brain. The human brain is able to maintain constant real time performance as cognitive complexity grows virtually unbounded through our lifetime. Our underlying thesis is that contextual categorization leading to simplified algorithmic processing is crucial to the brains performance efficiency. But, since the overheads of such reorganization are unaffordable in real time, we also observe the critical role of sleep and dreaming in the lives of all intelligent beings. Based on the importance of dream sleep in memory consolidation, we propose that it is also responsible for contextual reorganization. We target mobile device applications that can be personalized to the user, including speech, image and gesture recognition, as well as other kinds of personalized classification, which are arguably the foundation of intelligence. These algorithms rely on a knowledge database of symbols, where the database size determines the level of intelligence. Essential to achieving intelligence and a seamless user interface however is that real time performance be maintained. Observing this, we define our chief performance goal as: Maintaining constant real time performance against ever increasing algorithmic and architectural complexities. Our solution is a method for Malleable Contextual Partitioning (MCP) that enables closer personalization to user behavior. We conceptualize a novel architectural framework, the Dream Architecture for Lateral Intelligence (DALI) that demonstrates the MCP approach. The DALI implements a dream phase to execute MCP in ideal MISD parallelism and reorganize its architecture to enable contextually simplified real time operation. With speech recognition as an example application, we show that the DALI is successful in achieving the performance goal, as it maintains constant real time recognition, scaling almost ideally, with PE numbers up to 16 and vocabulary size up to 220 words. / Master of Science