Hi Neha,
For this thread, and all other threads where people want to evaluate
various memory types for the L2, L3, LLC, my pragmatic suggestion would be
to start by getting this patch ready to go out:
http://reviews.gem5.org/r/1809/
Once that is in place, the cache captures different read and write
latencies, and models the contention and waiting time for the banks. I
would argue that for most cases this is probably a sufficient model for
the performance impact (but I¹d be keen to hear if anyone disagrees).
The next step would be to add sufficient stats to capture how many
reads/writes and how long time is spent in various states (idle, refresh
etc), and use these stats for a high-level power model.
At this point, I do not see how adding more detailed memory models would
help. Could someone elaborate on why they think this is a good idea?
Thanks,
Andreas
Post by Neha KharwadkarHi!
I am mediocre user of gem5.
I am implementing following different configurations in gem5
L1 cache L2 cache
SRAM SRAM
SRAM DRAM
SRAM MRAM
SRAM STT-RAM
SRAM PRAM
The aim of this implementation is to find out power efficient and
performance wise best cache configuration. I know tools like NVsim and
NVmain with which we can implement this. But any idea how can we start with
this in gem5?
A detail reply would be appreciated.
_______________________________________________
gem5-users mailing list
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
-- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.
ARM Limited, Registered office 110 Fulbourn Road, Cambridge CB1 9NJ, Registered in England & Wales, Company No: 2557590
ARM Holdings plc, Registered office 110 Fulbourn Road, Cambridge CB1 9NJ, Registered in England & Wales, Company No: 2548782