Traditional Computer – Memory and processor relationship
Imagine you have a kitchen where the stove is in one room, and the refrigerator is in another. Every time you want to cook something, you have to run back and forth between the two rooms. This takes a lot of time and energy. Traditional computers work in a similar way, with the “stove” (processor) and the “refrigerator” (memory) being separate. Constantly moving data between them consumes a lot of power and slows things down, especially for complex tasks like artificial intelligence (AI).
New Approach: Computational Random-Access Memory (CRAM):
Researchers have come up with a new approach called computational random-access memory (CRAM). Think of it like having a kitchen where the stove and refrigerator are combined into one appliance. With CRAM, the computer can perform calculations directly within the memory, so data doesn’t have to move around as much. This makes the process much faster and uses less energy.
Experimental Demonstration:
To see if this idea works in real life, scientists built a small experimental model using a special type of memory called magnetic tunnel junctions (MTJs). They tested basic operations like storing and reading data, as well as more complex tasks like adding numbers. They found that CRAM could perform these tasks accurately, which is a big step forward.
Benefits of CRAM for humanity:
The benefits of this new technology are significant. For tasks that require a lot of data processing, like AI, CRAM can make computers faster and more energy-efficient. This means we could develop more powerful AI systems, improve technologies like speech recognition and image processing, and even create smarter devices for everyday use. Overall, CRAM has the potential to revolutionize computing, making it more sustainable and capable of handling the demands of the future.
Disclaimer: This content was simplified and condensed using AI technology to enhance readability and brevity.
Article derived from: Lv, Y., Zink, B.R., Bloom, R.P. et al. Experimental demonstration of magnetic tunnel junction-based computational random-access memory. npj Unconv. Comput. 1, 3 (2024). https://doi.org/10.1038/s44335-024-00003-3
