University of Texas researchers develop SOT-MRAM for AI efficiency
Researchers at the University of Texas at Austin have developed a new magnetic memory technology called SOT-MRAM, which could significantly improve the speed and energy efficiency of artificial intelligence workloads.
As AI models grow in complexity, they face a computational bottleneck caused by the constant movement of data between memory and processors. According to a study published in Science Advances, the SOT-MRAM architecture addresses this by bringing memory and calculation closer together, allowing some operations to be executed directly within the memory devices.
Unlike conventional memory, SOT-MRAM is non-volatile, meaning it can retain information even when power is disconnected. By utilizing magnetic properties, the technology offers faster state changes and lower energy consumption compared to existing memory solutions.
Entities
Claims
What the coverage asserts, and how many sources carry each claim.
- [● 3 SOURCES] Researchers at the University of Texas at Austin developed SOT-MRAM technology. www.eldia.es · www.diariodemallorca.es · www.elperiodicoextremadura.com
- [● 3 SOURCES] SOT-MRAM is non-volatile and retains data without electricity. www.eldia.es · www.diariodemallorca.es · www.elperiodicoextremadura.com
- [● 3 SOURCES] SOT-MRAM uses magnetic properties to store information. www.eldia.es · www.diariodemallorca.es · www.elperiodicoextremadura.com
- [● 3 SOURCES] The technology aims to reduce energy consumption and increase speed for AI workloads. www.eldia.es · www.diariodemallorca.es · www.elperiodicoextremadura.com
- [● 3 SOURCES] The architecture brings memory and calculation closer together to perform operations directly in memory devices. www.eldia.es · www.diariodemallorca.es · www.elperiodicoextremadura.com
- [● 3 SOURCES] The research was published in the journal Science Advances. www.eldia.es · www.diariodemallorca.es · www.elperiodicoextremadura.com