5 Responses to “Some suggestions”

1. 

   mradr
   Sep 09, 2009
   

   No. Actully i’m not sure if we really need to compress the data in ram. Anything in ram is going to be as fast as its going to get. Compression would hurt cpu cycles for the cost of only a few kb of space. Now, lets say we compress the data on a usb, we should get a increase in data bursting and for a few cycles of cpu that wont be as bad. Why? The limit of a flash drive is its bandwith, not speed, so by compressing that data going into that bandwith and into the ram we should see a slight or big jump into the data we arry form the hard drive at the cost of a few cpu cycles. Or aka, compression lets us grab more for little cost while ram is the uncompress version of the data that the cpu had to decompress from the usb. So see? If we compress the ram cache, the cpu would still have to uncompress that then store it into the unused ram, so we didnt save any space or cpu cycles at all and actully just caused more harm. While the usb is limited by how much it can push out, so we can increase this limit by making the data smaller and more compack sending more data to the cpu.

   
   
2. 

   Handojo
   Sep 10, 2009
   

   I think compression is very good choice. Remember when I use DOS and stacker. Compressed disk is claimed make computer faster when using faster CPU with slow disk.

   Eboostr often read than write. Decompress usually faster than compress.

   I think eboostr will got faster when using fast decompress algoritm. Eboostr must choose compression algorithm with fast decompress. And compression speed is not very urgent issue.

   
   
   • 

     mradr
     Sep 10, 2009
     

     Down side to a fast decompress is the less it can compress, so yea we wont see to much gain with space or speed. But we would see a larger increase in burst reading witch is differnt then normal reading/speed.

     
     
3. 

   Colonel ONeill
   Sep 11, 2009
   

   I agree, RAM is fast enough. Although compression can certainly be used to fit more, I’d think that the problem arises in chewing up CPU cycles to decompress data. Once you consider that if the amount of RAM you have requires eBoostr to compress stuff, you’d probably have a processor that won’t be able to keep up with everything, effectively negating the benefits.

   Unless you really want to cater to the niche of those who have an astonishingly low amount of RAM but a really fast processor.

   
   
4. 

   mradr
   Sep 11, 2009
   

   Would would only a need a (give or take for windows xp)
   P4 3.0 or higher @ 1GB ram or more (This setup may not work with vista or would need 2GB of ram, may work with windows 7)
   C2D 2.0 or higher @ 1GB ram or more (Will need 1.5-2GB of ram or more for vista, 1Gb for windows 7)
   QD4 1.5 or higher @ 1GB ram or more (Will need 1.5-2GB of ram or more for vista, 1Gb for windows 7)

   The number gets lower cause we can thread some of our compreission code so taking less cpu power/time. We’re looking at “time it takes” more then the cpu power when it comes to decompreissing. So really it wouldn’t be a “niche” but a greater number. Usally when I go help people with their computers they have a ok cpu but only like 1GB or less of ram.

   Now i’m not saying Colonel wrong, actully hes right in a way. Theres still no way off getting around “low amounts of ram” or a “slow cpu” so even with eboostr help that system would never be like if it had 2GB of real ram in it. But with compreission and a ok cpu to handle it, the system should “feel” like that computer has more then 1GB of ram.

   So unless you fall into the niche of people with a low amount of ram and cpu, it should work fine… but again it could be one of thos things it looks good on paper untill you try it in the real world and it starts to hurt more then it helps.