SHRIMP: Efficient Instruction Delivery with Domain Wall Memory
Multanen, Joonas; Jääskeläinen, Pekka; Khan, Asif Ali; Hameed, Fazal; Gastrillon, Jeronimo (2019-08)
Lataukset:
Multanen, Joonas
Jääskeläinen, Pekka
Khan, Asif Ali
Hameed, Fazal
Gastrillon, Jeronimo
IEEE
08 / 2019
International Symposium on Low Power Electronics and Design, ISLPED 2019
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-201910073733
https://urn.fi/URN:NBN:fi:tuni-201910073733
Kuvaus
Peer reviewed
Tiivistelmä
Domain Wall Memory (DWM) is a promising emerging memory technology but suffers from the expensive shifts needed to align memory locations with access ports. Previous work on DWM concentrates on data, while, to the best of our knowledge, techniques to specifically target instruction streams have not yet been studied. In this paper, we propose Shift-Reducing Instruction Memory Placement (SHRIMP), the first instruction placement strategy suited for DWM which is accompanied with a supporting instruction fetch and memory architecture. The proposed approach reduces the number of shifts by 40% in the best case with a small memory overhead. In addition, SHRIMP achieves a best case of 23% reduction in total cycle counts.
Kokoelmat
- TUNICRIS-julkaisut [17007]