A high-performance DRAM controller based on multi-core system through instruction prefetching
Kang Li, Qing Guang, Lei Li, Yu-Jia Peng, Jiangyi Shi
Abstract
Kang Li, Qing Guang, Lei Li, Yu-Jia Peng, Jiangyi Shi
Abstract
In this paper, we propose a cost-effective way to improve the performance of DRAM based on multi-core system. A novel DRAM controller with instruction prefecthing mechanism is introduced. The controller dynamically selects Open Page(OP) or Close Page(CP) policy by getting some information of future accesses in advance. This DRAM controller with dynamic policy based on instruction prefetching(DP_BIF), can provide DRAM the lowest possible latency without increasing too many areas of chip when compared with the controller only with OP policy or CP policy. The analysis of the simulation results show that the access latency of the DRAM memory can be improved nearly 10.4%, and the throughput of the DRAM is also increased nearly 10.2% by adopting the DP_BIF policy.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper, we propose a cost-effective way to improve the performance of DRAM based on multi-core system. A novel DRAM controller with instruction prefecthing mechanism is introduced. The controller dynamically selects Open Page(OP) or Close Page(CP) policy by getting some information of future accesses in advance. This DRAM controller with dynamic policy based on instruction prefetching(DP_BIF), can provide DRAM the lowest possible latency without increasing too many areas of chip when compared with the controller only with OP policy or CP policy. The analysis of the simulation results show that the access latency of the DRAM memory can be improved nearly 10.4%, and the throughput of the DRAM is also increased nearly 10.2% by adopting the DP_BIF policy.
Key concepts: Dram, Memory controller, CAS latency, Computer science, Latency (audio), Controller (irrigation), Embedded system, Low latency (capital markets)