The most popular and comprehensive Open Source ECM platform
Let us know more
Machine Learning: One Step Closer to Supercomputers that Can Fit in your Pocket
Machine learning is a branch of computer science related to artificial intelligence that attempts to enable computers to learn and make predictions and decisions based on collected information.
Increasingly, machine learning is being done with machines equipped with GPUs. The current generation of GPU machines are computers with very many core processors. A typical CPU today has two or four cores. GPUs may have 16 or more cores. The GPU cores currently run significantly slower than CPU cores and don’t have as many features built into it that are needed by most operating systems. The GPU cores have been tuned to work on very compute-intensive processes like video and image processing and scientific simulation.
Yann LeCun, who founded the NYU Center for Data Science, said that “multi-GPU machines are a necessary tool for future progress in AI and deep learning. Potential applications include self-driving cars, medical image analysis systems, real-time speech-to-speech translation, and systems that can truly understand natural language and hold dialogs with people.”
Work at MIT recently announced the processor that they’re calling the ‘Eyeriss’ which is a GPU with 168 cores. But not only does it have many cores, it is also very small and requires low power to operate. A processor like the ‘Eyeriss’ could revolutionize today’s smartphones. Smartphones, for example, would be able to run artificial intelligence algorithms.
Vivienne Sze, an assistant professor in MIT’s Department of Electrical Engineering and Computer Science, said that “right now, the networks are pretty complex and are mostly run on high-power GPUs. You can imagine that if you can bring that functionality to your cell phone or embedded devices, you could still operate even if you don’t have a WiFi connection.”
