Sean
Guest
|
 Posted:
Sun Dec 05, 2004 10:38 am Post subject:
a problem about cahce replacement stratergy. |
|
|
Hi, I have a hard time to deal with the cache replacement policy.The
problem is basically ask you to think out a replacement stratergy(even
only suitable for just one case!) that can make the direct-mapped
instruction cahce perform better than a samll direct-mapped one.
(as the Smith and goodman found: for a given small szie, a
direct-mapped instruction cache consistently outperformaed a fully
associative instruction cache using LRU replacement)
thanks for your kindly read and help. |
|
FAQ
Memberlist
Register
Profile
Log in to check your private messages
Log in