A New Hardware Prefetch Method and Its Effect on System Performance
Deng Rang
Abstract
Deng Rang
Abstract
The large latency of memory access is a key obstacle to achieving high processor utilization.Prefetch is one way to hide memory access latency,it is normally performed by explicitly executing instructions to move data close to the processor before they are actually needed,but this kind of prefetch will add burden on the programmers.In this paper,a new technology,which is realized by a VPFB in memory controller design,is presented,and its efficiency is analyzed by simulation.
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.
The large latency of memory access is a key obstacle to achieving high processor utilization.Prefetch is one way to hide memory access latency,it is normally performed by explicitly executing instructions to move data close to the processor before they are actually needed,but this kind of prefetch will add burden on the programmers.In this paper,a new technology,which is realized by a VPFB in memory controller design,is presented,and its efficiency is analyzed by simulation.
Key concepts: Instruction prefetch, Computer science, Latency (audio), CAS latency, Embedded system, Key (lock), Computer hardware, Controller (irrigation)