Exploring Mortal Computation, Energy Efficiency, and In-Memory Computing

The concept of mortal computation, which focuses on energy efficiency, is gaining momentum due to physical limits on thermodynamic costs in computing. The lend our limit and Jarsinski inequality highlight boundaries in the amount of work required to write and store information. This leads to the emergence of in-memory computing where computational calculations, like neuronal simulations, are done directly at the chip level, increasing efficiency by avoiding the inefficiencies of using von Neumann architecture. In-memory computing allows for organizing processing and memory closely to achieve more energy-efficient and faster computations, unlike traditional computing systems that rely on slow long-term memory and volatile memory.