1. 

   Sugar
   Beta Veteran

   Isn't it convenient if eBoostr can free up unnecessarily allocated memory?

   Posted 2 years ago #
2. 

   wayne
   Beta Veteran

   freeing up memory allocated to a program is counterproductive. unused (or 'free') memory is wasted memory. all those utilities that supposedly reclaim allocated memory and return it to the free pool are useless. i'd rather have it allocated and used than free. xp, vista, and now win7 make much better use of memory and will release any allocted memory if need developes a lot more efficiently than so-called memory optimisers. if microsoft's hi priced os programmers could have gained anything by this freeing up, they would have. do not waste time on freeing up memory, it will actually slow you down or at nbest make no difference.

   ebooster however is another story, it uses memory and USB solid state drives to cache data similar to vista and win7's readyboost, but in a complimentary and more efficient manner and speeds up disk access as opposed to a system without it or with just readyboost... a much better use of memory and solid state drive technology.

   Posted 2 years ago #
3. 

   mradr
   Member

   True but I remember a while back with a program called FreeRam XP Pro it had a way to compress the ram data instead of actully "freeing" up the allocated memory. Witch would be better for any solid state, being solid state does have a wright limit. With it compressed you can also store that much more data onto one usb/ram instead, at the cost of a few cpu cycles to uncompress it of cours. With most cpus being so fast and having 2 or more cores, I dont see this being to much of a hit in proformace at all even with a game running in the background (or any cpu vampire program).

   Really, the limit is in how much data can be pulled off the hard drive at any givien time, not really how much ram there is. Thats why SSD and Eboostr shines, cus it has a "build in array 0" for reading the data.

   Thats why I think it would be better if somehow (if they do this all ready idk.. and it would only work with a usb, so using ram wouldnt work and would actully slow the system start up down, cus it has to be put back into ram if the pc restarts (looks up what happens to ram when you restart your pc to know why if you dont all ready)) could put the drivers and other start programs on the usb, trick windows into thinking the drivers are on the usb and the drive (SSD or HD) that way you get the speed of array 0 or aka faster start up. So if you pc starts up in about 4mins, it should start up in 2-2 & half mins.

   Posted 2 years ago #
4. 

   hugh
   Member

   I've used the free memory programs you refer and I'm reasobly happy with them. Probably because I use them in low memory computers. Firefox, for example, can take a lot of memory by itself, and the problem becames worst if I have I lot of windos opened...
   I can't afford having memory not needed allocated and having problems running the program I want.

   Sugestion: Create a menu item under "options" for "low memory" or "low specifications" computers.

   I don't know what alghoritms or information from the computer You use to estimate future hits, but I suppose that in low specification computers the strategie might be different from the more fast cpu with large memory...

   Posted 2 years ago #
5. 

   mradr
   Member

   Hows so? I mean windows does a perty good job on this alone. Windows will not go out of its way to allocated memory to something not needed and eboostr lets you set priorities to what applications that should be cache. Actully, FreeRam XP Pro was only good with its ram compression, other then that, it work like the others in the way it would force the data back into your paging system and forced you to reload back from the hard drive sadly. Now if Eboostr could unallocate the page system or even trick windows into thinking its the page system... that would help out more then the other way around.

   Posted 2 years ago #
6. 

   chrislaine49
   Member

   I have to agree. A feature similiar to the FREERAM XP PRO would be of great benefit to this already very good program.

   Posted 2 years ago #
7. 

   PangOS
   Beta Veteran

   Just butting in, but won't the time and resources taken to "decompress" this data be taking away from the performance gains we are looking for?

   Posted 2 years ago #
8. 

   mradr
   Member

   Yes and no, yes it will take a few more cpu cycles to do this but depending on how deep the compression code is, not really. Actully we are comparing time to cpu cycles, not much else. So how much x time does it take to decompress data then it is x time to get the same uncompress data.

   USB are limited at max 35mb/s (give or take). So with compression we still grab 35 mb of data and send it to the cpu but when the cpu decompress we get maybe 40 to 50mb of data, thats almost 2 times more data then if we were to pull it off uncompress.

   USb 35 at 1sec = 35 at 1 sec
   Usb 35 at 1 sec + 2 x decompression at .2 at sec = 70 at 1.2 sec

   Again this all depends around how deep our compression code is.

   Posted 2 years ago #
9. 

   PangOS
   Beta Veteran

   I see what u mean, and thinking about it, it will take less time to decompress data as opposed to compressing it.

   Posted 2 years ago #
10. 

    mradr
    Member

    I would like to say, if eboostr does do this, maybe look into using .7z engine? I think they are free for use and you wont have to actully make your own compression engine this way. They usally have the highest compression rate too compared to others. I did some test also...

    1.5GB cache data took 7mins to compress
    Uncompress = 1.5 GB
    Compress = 595 MB
    Compression Ratio = 38%
    Savings = 941 MB
    Decompression time as a whole = 2 mins
    Decompression time as singal files = ? (Less then a few milli-sec I bet if you were to rewrite it in this way)

    so in the end, with compression in this format, we will see 2x more space for us to put more into and almost 2x more brust reading from a USB. Down side is setup will take 5 to 7x more time, so if it took 2mins for setup, now it will take 9 to 10mins for setup and even more with larger cache sizes.

    So using the data above time will take (give or take)..
    not compress data + max size cache file = 2mins
    compress data + max size cache file = 16mins

    14mins more time needed... but at almost 2 times more data cache and brust rate reads.. is it worth it?

    Posted 2 years ago #

RSS feed for this topic

Reply

You must log in to post.