home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.arch:8990 comp.arch.storage:597
- Newsgroups: comp.arch,comp.arch.storage
- Path: sparky!uunet!mcsun!sunic!aun.uninett.no!alf.uib.no!newsroom.bsc.no!izahi
- From: izahi@bsc.no (Raul Izahi Lopez Hernandez)
- Subject: Re: Memory Hierarchy Questions
- Message-ID: <1992Aug21.071534.3737@newsroom.bsc.no>
- Sender: usenet@newsroom.bsc.no (Usenet News Administrator)
- Organization: Bergen Scientific Centre, Bergen, NORWAY
- References: <y6#nqyb.ken@netcom.com>
- Date: Fri, 21 Aug 1992 07:15:34 GMT
- Lines: 22
-
- In article <y6#nqyb.ken@netcom.com> ken@netcom.com (Ken Kofman) writes:
- >Hi,
- >
- >I have a few questions, and I hope this is the right place to post them.
- >
- >Along these lines, I have also developed an interest in learning about
- >the kinds of memory hierarchies being designed into real-world larger
- >systems. The Patterson/Hennessey book was a good start, but I'm also
- >interested in actual data. How does doubling main memory from 200Meg to
- >400Meg, from 0.5Gig to 1Gig, or from 1Gig to 2 impact the miss rate?
-
- It will not affect the miss rate.
- The miss rate of such a system is solely determined by the cache so
- increasing main memory will only decrease the miss penalty since the pager
- has to go to disk less often.
- In the real world of large environmental simulations it DOES help to
- increase main memory of workstations to 256-512MB+ levels.
- --
- -----> All opinions expressed here are my own, not IBM's <-----
- Raul Izahi Lopez Izahi Engineering
- izahi@bsc.no IBM Bergen Environmental Sciences and Solutions Centre
- Thormoehlensgate 55, 5008 Bergen, NORWAY (47-05)544653
-