Analytically Modeling NVM Design Trade-Offs


As considerable ongoing research is put towards developing emerging non-volatile memories, there is a growing interest in evaluating the potential of these technologies for enabling efficient compute in a variety of application spaces. While simulation tools for characterizing different NVMs at the array level exist, the increasingly complex design space introduced by emerging NVMs together with the need to assess the impact of application-specific performance and energy requirements demands the ability to conduct extensive and efficient system-level analysis. Our proposed, generalized framework provides an opportunity for analyzing the performance, power, and area implications of leveraging a particular proposed NVM technology under a specific use case.

Accepted as a poster in the Non-Volatile Memories Workshop (NVMW ‘20)
Akshitha Sriraman
Akshitha Sriraman
Assistant Professor

I am an Assistant Professor in the Department of Electrical and Computer Engineering at Carnegie Mellon University. My research bridges computer architecture and software systems, with a focus on making hyperscale data center systems more efficient (via solutions that span the systems stack). The central theme of my work is to (1) design software that is aware of new hardware constraints and (2) architect hardware that efficiently supports new hyperscale software requirements.