In-Memory Processing Architectures and Applications

In-Memory Processing Architectures and Applications, published as Volume 144 of the serial Advances in Computers, provides a comprehensive and up-to-date examination of processing-in-memory (PIM) and compute-in-memory (CIM) architectures, addressing their architectural foundations, circuit-level innovations, security and reliability challenges, and emerging application domains. This volume brings together authoritative contributions that span DRAM- and SRAM-based in-memory computing, ferroelectric and other emerging memory technologies, modelling and simulation frameworks, and system-level interconnect innovations, offering a balanced perspective on both the promises and pitfalls of PIM. Collectively, the chapters position in-memory processing as a key enabler for overcoming data-movement bottlenecks in modern computing systems while highlighting the reliability, energy efficiency, scalability, and security considerations that must be addressed for real-world deployment.

• Explores DRAM-, SRAM-, and emerging-device-based processing-in-memory architectures at circuit, architecture, and system levels
• Examines reliability, energy efficiency, and security as first-class design concerns in PIM/CIM systems
• Presents modelling and simulation frameworks to support design-space exploration and evaluation
• Discusses architectural trade-offs, scalability challenges, and interconnect innovations, including 3D-stacked memories
• Balances visionary perspectives with critical analysis of practical limitations and deployment challenges