How Chips Are Using On-Chip Magnetic RAM for Non-Volatile Memory

In Episode 90 of The Hardware Podcast, Lucas and Luna dive into the emerging technology of on-chip magnetic RAM (MRAM). They explain how a new generation of chips is integrating MRAM cells directly into the logic die, enabling non-volatile memory that retains data without power. The episode focuses on a specific case: the 2025 deployment by a major foundry of 12-nanometer embedded MRAM for automotive microcontrollers. Lucas breaks down how MRAM works using magnetic tunnel junctions, why it beats SRAM and flash for certain applications, and the key metric — 260 picosecond read speed at 85 degrees Celsius. Luna asks about reliability concerns, and they discuss endurance beyond one million write cycles. The conversation also touches on how MRAM could reshape system architecture by enabling instant-on computing and reducing standby power. A natural segue into listener support leads to a brief, sincere mention of the show’s ad-free funding model, then back to speculation on MRAM’s role in edge AI. Packed with specific numbers and engineering context, this episode is a deep but accessible look at a quiet revolution in memory technology. #MRAM #EmbeddedMemory #NonVolatileMemory #OnChipMemory #Semiconductor #MagneticTunnelJunction #Foundry #AutomotiveChips #Microcontrollers #12Nanometer #EdgeAI #InstantOnComputing #Technology #Hardware #Engineering #ChipDesign #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo