Fujitsu promises accelerated deep learning

While using less memory. Fujitsu has introduced a new technology to improve the internal memory of GPUs in order to increase the machine learning accuracy. Fujitsu claims that the new technology has doubled the efficiency of machine learning compared with previous technology. According to the company, in recent years the use of GPUs in machine learning has increased. GPUs have been providing the raw power to complete the complex calculations required for machine learning. The calculations can be done seamlessly when the data is stored in the GPU chip, but there is a limit on the GPU chips to store information. So, Fujitsu has been working to improve internal memory in GPU. Fujitsu says that simply increasing the internal memory would not be sufficient. It says that this can achieved by improving…


Link to Full Article: Fujitsu promises accelerated deep learning