Significant Step Toward Energy Savings in Artificial Intelligence Technology

IBM chip prototype could enable energy savings in artificial intelligence technology, according to reports. According to BBC reporting, the company's "human brain-like" chip prototype could reduce battery consumption in smartphones. Thanos Vasilopoulos, a scientist working at IBM's research laboratory in Switzerland, noted that the human brain accomplishes much with minimal energy, and that the chip in question could perform large and complex tasks with low power consumption. While concerns have been raised about the carbon emissions impact of devices controlling artificial intelligence technology, Vasilopoulos added that the chip could be used to reduce energy costs and "carbon footprint" of cloud computing systems.

Operating like the human brain

IBM stated that in the chip it designed, they used "memristor" components that can store intermediate values, unlike digital chips that only carry 0 and 1 values. It was noted that the chip's energy efficiency is due to "memristors" that operate like the "human brain." IBM stated that they hope chips used in phones and vehicles will become more efficient in the future by providing longer battery life and enabling new applications. Professor Ferrante Neri from Surrey University said he believes chips containing "memristors" that mimic the structure of the human brain will increase in the future, but emphasized that the widespread adoption of this technology could face challenges such as manufacturing difficulties and material costs. The proliferation of chips like IBM's prototype could enable substantial energy savings in computers controlling artificial intelligence technologies. Source