$GTCH "Our computing world is in a constant need for performance growth and vast engineering efforts are made in order to enable faster computing systems. Microchip technology, which directly dominates computing power, affects nearly every aspect of our lives, including medicine, communication, security and business interactions. We are witnessing constant advancements in the IC arena that lead to extraordinary increase in computer performance. Microprocessor companies are in a constant race to design multi-core CPUs in order to provide faster computing power for the constant demand. But we need to think further. We need to push the limits into new paradigms in order to achieve a major leap in computing power, given the fact that we are still limited by the physical restrictions of electrons moving through matter," stated Danny Rittman, GBT’s Chief Technology Officer. "Since arithmetic and logic operations within microprocessors are the crucial factors for speed, we developed a new approach to the ALU. This new system and method introduce a different circuitry and machine language (Flow) with the goal of performing a much faster computation in an efficient algorithm, without any change in the computer hardware. We are seeking to introduce a different approach in the basic ALU calculations handling, that offers a possible way to make computers faster through parallel processing and pre-computation instructions execution, in order to achieve higher computing efficiency. We are taking into account the basic, traditional ALUs design concepts and changing them into out-of-the-box methodologies, so we can devise new kinds of arithmetic/logic operations that we believe are more efficient for computations. We are aiming to break the ultimate limitations of our traditional, decimal/binary approach within ALU by introducing new calculation methods. In order to keep up with the growing demand for high performance computing, hardware engineers are looking to alternatives like Graphics Processing Units (GPUs) and Tensor Processing Units (TPUs) which are mainly based on parallel computing to execute more instructions at once rather than do basic operations faster. As Moore’s Law is fading away, high performance computing will require a conceptual shift. As part of our R&D constant efforts to find new methods to improve our computing world, especially for advanced machine learning algorithms, we are seeking to develop new ideas with the goal of enabling high performance Arithmetic/Logic processing. The patent application seeks to protect a system and method. In parallel to filing the patent application, we plan to design an experimental, proof-of-concept microchip and test its performance against traditional ALUs."
(0)
(0)