Artificial intelligence (AI) might become more energy-efficient because to a prototype “brain-like” processor, according to IBM, a global leader in technology.
Artificial intelligence (AI) might become more energy-efficient because to a prototype “brain-like” processor, according to IBM, a global leader in technology. Emissions from computer warehouses that fuel artificial intelligence systems have drawn attention.
IBM noted that the prototype might result in more effective, battery-saving AI circuits for smartphones. It claimed that its effectiveness was due to parts that functioned similarly to connections in human brains.
Physicist Thanos Vasilopoulos said that who works at IBM’s research facility in Zurich, Switzerland, “the human brain is able to achieve remarkable performance while consuming little power” in comparison to conventional computers.
According to him, greater energy efficiency would allow for the execution of “large and more complex workloads in low power or battery-constrained environments,” such as automobiles, mobile phones, and cameras.
Furthermore, he continued, “cloud providers will be able to use these chips to lower their energy costs and carbon footprint.”
It demonstrates how advancements in hardware design can contribute to making AI more energy-efficient, which is crucial for reducing the environmental impact and operational costs of AI systems.
The majority of chips are digital, meaning that information is stored as 0s and 1s, but the new chip makes use of memristors, which are analog components that can store a variety of numbers.
The distinction between digital and analogue may be compared to that between a light switch and a dimmer switch.
Since the human brain is analog, the operation of memristors is analogous to that of synapses in the brain.
According to Prof. Ferrante Neri of the University of Surrey, memristors are a type of computing that is “nature-inspired” and imitates brain activity.
In a manner analogous to a synapse in a biological system, a memristor may be able to “remember” its electric past.
“Interconnected memristors can form a network resembling a biological brain,” he claimed.
He expressed cautious optimism over the potential of chips utilizing this technique, saying that “these developments signal that we may be on the verge of witnessing the appearance of brain-like chips in the near future.”
However, he cautioned that creating a memristor-based computer is not an easy operation and that there would be many obstacles in the way of general adoption, such as high material costs and challenging production processes.
The new chip’s use of these parts increases energy efficiency for AI, but it also incorporates digital components. This makes integrating the chip into current AI systems simpler.
Nowadays, many phones are equipped with AI processors to assist with tasks like photo processing. A “neural engine” chip, for instance, is present in the iPhone.
Future processors for phones and automobiles, according to IBM, might be more effective, providing greater battery life and new uses.
In the future, chips similar to the IBM prototype may help save a lot of energy if they were to replace the processors in the banks of computers driving potent AI systems.
The water required to cool the power-hungry datacenters might also be reduced. A big data center will need as much power as a medium-sized town to operate. Data centers require enormous quantities of electricity to operate.
James Davenport, a professor of IT at the University of Bath, called IBM’s findings “potentially interesting” but cautioned that the chip was more like “a possible first step” than a “easy to use” solution to the issue.