Researchers at the University of California have made a groundbreaking discovery that could potentially double the power of PCs without the need for additional hardware.
By introducing a new method of parallel data processing, these scientists have unlocked the potential for significant increases in computing speed and reduced energy consumption in computers.
This innovative approach harnesses the combined power of the central processor, graphics processor, tensor cores, and other components for parallel data processing. The result is nearly a twofold increase in data processing speed compared to using just the graphics processor, while also reducing energy consumption by an impressive 51%. What’s more, these results were achieved through the optimization of existing components, without the need for new hardware.
The creators of this method, Hung-Wei Tseng and Kuan-Chieh Hsu, successfully integrated the central processor, Nvidia graphics card, and AI processor on a single system platform. They believe that current computers are not fully realizing their potential, and while the practical implementation of this technology requires code and hardware optimization, the potential benefits are vast.
This breakthrough has the potential to revolutionize the tech industry, benefiting both large IT companies and everyday users. It promises to accelerate the development of artificial intelligence, reduce costs, enhance computer performance, decrease energy consumption in data centers, improve the capabilities of mobile devices, and much more. The research conducted by these California scientists in the field of parallel computing could mark the beginning of a new era in computer technology.
Original from www.playground.ru
