On the 13-April-2023 at 11:00 a.m. the 20th lecture of the Living Lab lecture series will take place. In this talk, ScaDS.AI scientific researcher Asif Ali Khan will talk about Compilation of Emerging Near-Memory and In-Memory Computing Architectures. Add this event to your calendar (iCal).

Compilation of Emerging Near-Memory and In-Memory Computing Architectures

Data-intensive applications have become increasingly popular, but they have exposed the limitations of conventional processor-centric von-Neumann architectures. These architectures struggle to meet the off-chip memory bandwidth requirements of these applications, resulting in hard performance and energy-efficiency boundaries. This is because data has to be moved between the CPU and memory via a narrow memory channel, which consumes a significant amount of energy. To address this issue, in-memory processing architectures have been developed, including near-memory computing (NMC) and computation-in-memory (CIM). NMC aims to process data close to where it resides, while CIM aims to reduce data movement completely by using memories with compute capability. These architectures can improve performance by reducing data movement and exploiting the parallelism of the processing units. However, compilation for these architectures must take into account the unique characteristics of the memory and processing units to optimize performance.