If you have to install an app to speed your machine up, you're doing something wrong.

holymogwai

@daybringer: I agree. Considering how you could get 4 gigs of DDR800 for lest than a benjamin, paying $30 for the software, and then having usb devices wasted on a slow boost, is way too much. Unless there's an open source or free version that isn't crippleware, I can't see myself using it, unless I were to use the built in one in Vista, if I had Vista.

nman

For how much that costs it might just be better to buy more RAM.

daybringer

I might give this a shot if and only if it were freeware.

ffantasyv

I'm considering buying this but I have a question for the experts...

I recently got a Lenovo ThinkPad T61 which comes with the new Intel Turbo memory. Since I use Windows XP, I had to disable the Intel Turbo drivers and followed these instructions:

[www-307.ibm.com]

My question is (a) will the eBoostr work with the Intel Turbo? and (b) if so, does anyone knows how we re-activate the Intel Turbo drivers on XP? (the "PCI Memory Controller" disappeared from the Device Manager once I installed the disabling tool :(

Much thanks, looking forward to try this!
a.

avniassa

Potentially avoids voiding a warranty or angering the IT gods by opening a case though.

jb008

I've got 2 GB RAM and allowed eBoostr to use the full capacity of my 2 GB flash drive. I don't really notice any performance boost except on startup after I login. Startup programs and programs I use frequently seem to startup a little faster. Windows uses less virtual memory while using eBooster, so I assume it's caching to my flash drive first before using the pageflie. RAM > eBooster cache > virtual memory is faster by principle, so it may be worth it if you have 1 GB or less RAM using XP.

yochanan_marqos

Anything like this for Mac OS? My old G4 800 iBook surely could use some help from my spare 1G thumbdrive...

thurana

Ironic that I'm reading this after just ordering 2 gig of DDR2 ram for my laptop I might use it if it was freeware but buying a program for almost as much as the ram is pointless.

Darth_Ninja

@jb008: The only thing you'll possibly void a warranty opening a case on is a laptop, and at least one or two laptop manufacturers have said that breaking the "warranty void if broken or missing" seals hiding the screws to get inside the case will not void the warranty. Try opening your computer case some time; it's not that scary.

Jarett

@UnStatusTheQuo: The purpose of ReadyBoost isn't to replace disk caching, it's purpose is to make up for the lack of speed when it comes to random access on the hard disk. The USB would be faster accessing the memory, but it would bottleneck when it came time to transfer data. It depends on what you're doing, but chances are doing what you described would slow down the system.

maverick0

@UNSTATUSTHEQUO: You wouldn't want to do that. Flash drives are actually 5 to 10 times *slower* at sequential reads than a hard disk. The advantage of flash is very low random access times (~1ms, vs. 8-10 ms on even the fastest consumer disks). ReadyBoost, and presumably eBoostr, work by caching only smaller, spread out, frequently-accessed files, where the access time, rather than the read time, is the bulk of the retrieval time.

As others have mentioned, adding RAM is almost always going to be a better bet. ReadyBoost is said to only make an appreciable difference on a Vista boxes with less than a gig of RAM. That said, this thing might be just the thing to breathe new life into an older laptop with a slow 5400 RPM hard drive if a memory upgrade isn't feasible.

ras07

Correct me if this wouldn't do anything, but couldn't you also just remove the paging file from the C: drive, and stick a permanent USB stick into your computer, then assign a static-sized paging file to the USB stick "drive"?

I mean, isn't the point to speed up the cache and not use disk caching?

That will be $29.99, please. Ha!

UnStatusTheQuo

@Jarett: wrong, a lot of computers now have the "intrusion" detection that is triggered the first time a case is opened. Not just laptops.

holymogwai

Readyboost in Vista does not work, so it is safe to say the same of the same

ss6sam6

Soo close, but if it ain't freeware, I ain't using.

angryrider

Couldn't one get the same effect by simply making your swap position live on a USB thumbdrive?

chicofelipe

After [my personally experienced] letdown of readyboost in Vista (4GB A-Data PD7, tested with 512MB and 2GB RAM setups), $29 seems pretty steep to try. I'll wait for an open-source version, kthx.

capsaicin

@HOLLYMOGWAI: "PC" = "Personal Computer". If opening the case voids the warranty, then I suggest taking your money somewhere else.

IAmMarchHare

Has anyone tried this with a virtual machine? (e.g., Parallels on Mac)

wqwert2

isn't there a maximum number of write cycles for flash memory? and while high, using it as active memory (constantly being accessed) would likely rapidly advance towards that maximum number
-
maybe i'm thinking of something else - can anyone validate this? i can't seem to find the writing i once read on it

m-r-v

I'd also add that I've tested readyboost in vista and it DOES work (there are many ways to benchmark and/or monitor the usage), but don't think that taking some off-brand USB key is going to improve anything because USB key performance varies widely.

Biggrz

Some of you are missing the point of Readyboost. RAM does a great job with things that belong in RAM but there will always be very small files used by your pagefile but your HD sucks at dealing with those. The readyboost drive takes over caching those files and gives a performance benefit in that case.

Biggrz

Ok guys, I just tested my Lenovo T61 laptop with the eBoostr w/Cruzer SanDisk U3 2m memory stick.

For my full system specs, see this:
[screencast.com]

Here's the benchmark WITHOUT eBoostr:
[screencast.com]

Here's the benchmark WITH eBoostr:
[screencast.com]

I ran both right after system reboot with no major apps open other than my usual (light) startup list (anti-virus, and a very short list of must-load startups).

As you can see, there is some improvement but it doesn't seem to be major...

what do you think, should I keep it?

avniassa

Is there any nice way of testing if there is any overall increase in performance? Sam

bugmenot21

Would this be faster if cache was on a separate hard drive, as opposed to a flash drive?

Dooga

Is there anything like this available for Ubuntu?

snwbrder0721

I installed and tried their benchmark thing. The before was about 10, and with using the flashdrive was 20!
So that's double. I might try it for a few days to see how I like it.

da5id_nz

I would never add this before adding real ram. Not to mention that ram may be cheaper depending on type. 1+gig of ram vs a program like this isn't really a choice at all. I guess if you are maxed out on the ram that your motherboard/bios/OS can handle, and you already run your swap on a separate drive, then it might be worth it, maybe.

@da5id_nz:
Might want to try a third party benchmark instead of one from the company trying to sell you on a product.

Will

I've noticed some degrading as the service runs... my startup seems slower after a few weeks.

Dooga

I'm currently testing eBoostr on what it seems to me a perfect candidate - my trusty old Celeron 450 (overclocked) running XP SP2 on 384 MB of RAM (go figure :-)
After plugging in a 2 GB USB key I noticed a significant performance increase as disk activity dropped for many tasks. The cache itself claims to fill up 35% of available space, thus floating around 775 MB.
The only problem so far is eBoostr is caching some big files that are often updated, like AVG's incavi.avm database (20 MB on my system), and I can see a serious performance hit as copy/delete/update operations are throttled by the key sequential write speed (which is rubbish).
If it was possible to explicitly blacklist certain files the product would be better. I just wrote them asking about this feature.

On the other hand I perfectly agree, the only target market for eBoostr are old PCs lacking RAM or laptops with slow disks. It's an interesting proof of concept without a real business case.

DriftSK