Winrar hardly using cpu?
- Thread starter Energize
- Start date
Work out the MB/s you are getting, or try 7Zip as that should tell you the speed you are getting. 7Zip also has a bench mark option and I think winrar may, seem to remember there being a "post your scores" type thread on here once.
And if unzipping to the same drive it'll be slower.
And if unzipping to the same drive it'll be slower.
But at the figures you see above extraction works out at 11.7MB/s, which implies a cpu bottleneck considering the hdd has an average read speed of 65MB/s.
Last edited:
Almost certainly you are being limited by media write speed. With a single HD you are probably talking under 100MB/s at best and modern CPUs are capable of decompressing way faster than that. Even 5 years ago CPUs weren't getting maxed out.
If you want to see better CPU utilisation, then you really need to get yourself some faster storage, i.e. extracting from one unfragmented RAID0 array (Velociraptor/SSD) to another.
65MB/s read speed is all well and good, but that's not the bottleneck, it needs to write as well, and reading and writing from the same drive will slow things down even further (it's a mechanical device and as such isn't 'full duplex')
If you want to see better CPU utilisation, then you really need to get yourself some faster storage, i.e. extracting from one unfragmented RAID0 array (Velociraptor/SSD) to another.
65MB/s read speed is all well and good, but that's not the bottleneck, it needs to write as well, and reading and writing from the same drive will slow things down even further (it's a mechanical device and as such isn't 'full duplex')
Yeah but were talking about a factor of 6, I have moved files accross the hdd much faster than this!
Well, of course you've then got to consider that copying files and extracting an archive aren't the same thing - the drive cache (and possibly some OS tomfoolery) will likely have a big benefit for copying because it can stream data much faster in large chunks. Whereas extracting from an archive (even if it's a single file) it's kind of coming in dribs and drabs if that makes sense.
I used to wonder why I was getting like 15-20MB/s extract speeds in the past, but I found that extracting to a brand new hd was much quicker, around the 50MB/s level.
As a test you could set up a RAM drive and try extracting to that. I bet you get much better performance and higher cpu utilisation (although read speed would then become a factor). I'd expect u something approaching full cpu utilisation if both the source and destination were on the ramdrive though.
I used to wonder why I was getting like 15-20MB/s extract speeds in the past, but I found that extracting to a brand new hd was much quicker, around the 50MB/s level.
As a test you could set up a RAM drive and try extracting to that. I bet you get much better performance and higher cpu utilisation (although read speed would then become a factor). I'd expect u something approaching full cpu utilisation if both the source and destination were on the ramdrive though.
Last edited:
By default yes it will be the same disk as WinRar just creates a sub-folder under the archives current folder.
As a test could the OP try extracting the files to another HDD and see if that makes a diff?
I'd be more concerned with your ram usage on firefox and svchost on that screenshot
Lolz. My Firefox RAM usage normally ranges from between 400mb to over a gigabyte.
But then I do tend to float around with lots of tabs, currently only on 77, positively frugal!
As for svchost, well any number of things can cause one of them to up its RAM usage, not least copying files.