You have an innovative idea in eBoostr where you can give certain apps priority in the cache (which is very helpful for small cache sizes). Since I'm not aware of any other HD Caching products that offer this capability, I think you could benefit from selling a cheaper version of this product as well that utilizes only system memory for it's cache. Since the findings in the forum seems to suggest that maximum USB2 performance isn't obtainable on systems not having Dual Core processors, this may be a better alternative for people having older less powerful systems. I know my preference is to use system memory for eBoostr's cache rather than USB2 Flash Drives.
There's a market for a Cheaper version of Eboostr that uses only system memory
(13 posts) (4 voices)-
chooky
Beta Veteran+1 Good idea!
Posted 2 years ago # -
mradr
Member+ 1 here too. I understand the sell part of what your talking about, but I dont follow when you say that USB cant rich their speed.. USB are cap at the 30-40 mark, but thats ually way faster then what older systems (hard drives) can even read at while system memory is usally scariest when it comes to theses older systems... Why take more ram up when the system could be using it? Usally middle to higher range will have the more memory that the system can't fully use so you could build smaller or larger amounts on that.
Actully... I think eboostr needs to sorta follow the MS rule and say you have to have a set speed on your USB to even have it work. Thats why you see people with soo many different speeds when it comes to USB. No USB is like.
Posted 2 years ago # -
chooky
Beta Veteranmradr it isn't the 30-40MB/s sequential read speed that is important with USB drives its the access time and 4K reads which are faster than a HDD even on cheap USB drives. As long as eBoostr only stores the smallest files this is where the performance would come from. The problem is I don't think eBoostr stores the files like this.
Posted 2 years ago # -
Robert Persons
MemberActually you can limit eBoostr to using only small files (which USB drives are best at ... as you say Chooky).
If you look at the help file under "Advanced configuration"/"fine tuning" you'll find you can add a value in the "settings.ini" file:
maxfilesize
This parameter is used to limit the size of the file that eBoostr is allowed to cache. The value is given in bytes (for example 100MB = 100 × 1024 × 1024 = 104,857,600 bytes.)Posted 2 years ago # -
mradr
MemberI know chooky, Robert Person got the right idea too. As for the limit, I wasn't really basing that off the sequential I/Os 100%... Sorry sorta tried ^^: I'm not sure what I was talking about really. Still dont... but I know I wasn't basing thos numbers on the sequential I/Os alone, or I hope not.. anyways maybe I was.
Also, Chooky did you ever figure out what your random low memory was from? If you need help, I'm more then happy to get a hold of you off the forum for more help.
Posted 2 years ago # -
chooky
Beta VeteranThanks robert didn't know that. But can I use that on a per cache basis to limit the size of files on USB only.
mradr, waiting to see what eBoostr support say.
I know there is nothing wrong with my memory.
Hard faults are normal and not a bad thing or a problem according to Microsoft"A hard fault (also known as a page fault) occurs when the page of the referenced address is no longer in physical memory and has been swapped out or is available from a backing file on disk. It is not an error. However, a high number of hard faults may explain the slow response time of an application if it must continually read data back from disk rather than from physical memory."
http://technet.microsoft.com/en-us/library/cc722173(WS.10).aspx
I am fairly sure a reinstall of eBoostr would fix it but am waiting in case eBoostr want to run some diagnostics to see what is going wrong.
Posted 2 years ago # -
byteboss
Membermradr wrote:
I dont follow when you say that USB cant rich their speed.. USB are cap at the 30-40 mark, but thats ually way faster then what older systems (hard drives) can even read atI tested the speed of a Patriot XT 4 GB USB Flash Drive on the following three systems and got the results shown:
System 1:
Intel 6550 2.33 GHz CPU with 2 GB RAM: Read Speed is 31.0 MB/S, Write Speed is 10.43 MB/S
System 2
Intel Pentium 4, 2.8 GHz CPU with 2 GB RAM: Read Speed is 26.3 MB/S, Write Speed is 10.2 MB/S
System 3:
Intel Pentium 4, 1.6 GHz CPU with 512 MB RAM: Read Speed is 16.0 MB/S, Write Speed is 8.3 MB/SSystem 1 Note:
System 1 is my newest machine having a Dual Core CPU and faster bus speeds and memory than the other machines tested, but processor speed is .47 GHz lower than that of system 2.System 3 Note:
System 3 is an older computer that has a Rosewill USB 2.0 5-Port PCI Adapter installed to give it USB2 capability.If reduced USB2 performance isn't responsible for not being able to achieve the same 31 MB/S read speeds on systems 2 and 3, then what is? The only thing I know for sure is that the speed of the flash drive didn't change between tests. My theory is that it's a combination of factors involving differences in bus speeds, the speed of system RAM, and CPU speeds. No matter what the cause, it points to the fact that older computers systems are not capable of the performance necessary to achieve the maximum USB2 performance specification.
I certainly don't consider myself an expert on computer hardware, so I'm definitely open to other ideas if anyone has some to offer.
Posted 2 years ago # -
mradr
MemberYou can't really count system 3 then due to a third 3 party hardware in the mixs, or I just cant comment on that one type of setup due to lack of information on that pci hardware.
**Update
Not sure if its the same card, but I found it and its rated speeds. Yes, it does havea lower cap then what you would find on a MB of 6-7 year old, a cap of around 16. So simply just getting a new USB controller will speed it up more. Just make sure you check the speed rates on it.system 2: what OS? I find that they have the bigest drop in speed vs the hardware (I still have a P4 2.2 1GB with xp on it and I still get the same speed as my Dual core 3.4 with 3GB of ram while my older ME hard drive put into the the same P4 build will cause some slow down to the USB it self.) I'm just not sure if there isn't some sorta trick to boosting that speed for the USB on older OS. Could it be that older OS had a bigger overhead on the USB?
Posted 2 years ago # -
byteboss
Membermradr wrote:
You can't really count system 3 then due to a third 3 party hardware in the mixs, or I just cant comment on that one type of setup due to lack of information on that pci hardware.
**Update
Not sure if its the same card, but I found it and its rated speeds. Yes, it does havea lower cap then what you would find on a MB of 6-7 year old, a cap of around 16. So simply just getting a new USB controller will speed it up more. Just make sure you check the speed rates on it.system 2: what OS? I find that they have the bigest drop in speed vs the hardware (I still have a P4 2.2 1GB with xp on it and I still get the same speed as my Dual core 3.4 with 3GB of ram while my older ME hard drive put into the the same P4 build will cause some slow down to the USB it self.) I'm just not sure if there isn't some sorta trick to boosting that speed for the USB on older OS. Could it be that older OS had a bigger overhead on the USB?
I should have stated that all three computer systems are running a fresh copy of Windows XP Pro SP3 (fresh meaning installed within the last 6 months or so). And the USB2 PCI controller that was added to system 3 was purchased and installed just three months ago. That particular Rosewill USB adapter was also given very good reviews by the customers who had previously purchased it. One other point worth mentioning is that none of the PC's tested are overloaded with either software or hardware. I avoid doing that. Instead, I dedicate different machines to different computing tasks in order avoid hardware and software conflicts and other issues that can cause PC performance to suffer. I also periodically clean the registry in all machines of left-over garbage and other bad entries. Also, all of these machines are performing well and are not experiencing any glitches in the OS that might point to other problems. All systems are also running the latest drivers, although I suppose drivers could still be source of the problem if they are poorly written and inefficient code.
You mention that the older ME hard drive would slow down the USB in your P4 system, but surely that has no effect on the read speed measurement of a USB Flash Drive. Read Speed measurements of a USB Drive do not require accessing the hard drive at all!
Posted 2 years ago # -
mradr
MemberSorry, not the hard drive, I mean the OS what was doing it.
The adapter I saw must have been yours then, around $10.00 on newegg, and the speed limit is 16.
As for system 2... I have no clue other then you have other hardware using a USB interface, unplug other USB interface devices then test it.
Posted 2 years ago # -
Robert Persons
MemberHi Chooky,
Refering to your earlier question:
"Thanks robert didn't know that. But can I use that on a per cache basis to limit the size of files on USB only."
As far as I know the maxfilesize parameter in eBoostr applies to all cache devices. Perhaps the developmen team will consider expanding it to set it individually for each cache device.
One way I've worked around this somewhat is to use my slower flash drive (I'm using three) as a ReadyBoost drive. Since ReadyBoost limits itself to 4K files and smaller, it is getting very good performance on those size files. It is on the larger files that eBoostr caches that the higher MB/s rating matters more. This is not a direct solution to your point, but at least it's a way to specify which flash drive gets the smaller files.
Posted 2 years ago # -
chooky
Beta VeteranThanks Robert! Would be great to get a response from the eboostr team on this.
Posted 2 years ago #
Reply
You must log in to post.