This $400,000 Project Grant awarded by the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE, CFDA 47.070) program supports a collaborative research effort led by Carnegie Mellon University (CMU). The project aims to develop a hardware-software co-design approach for high-performance in-memory analytic data processing. Key technical thrusts include mechanisms for in-place data analytics on dynamic random-access memory (DRAM), processing-in-memory (PIM) designs for static random-access memory (SRAM)-based caches, and a domain-specific language (DSL) framework to optimize analytics query execution across DRAM and SRAM-based PIM. The project also involves developing simulation infrastructure and benchmarks to support the broader architecture and database research communities. In addition to the technical research, the project features educational and outreach components, including training students across hardware-software boundaries and engaging with Native American high school students and Historically Black Colleges and Universities (HBCUs) in the Virginia area.
Generated 8/13/24, 2:41 AM