Hardware Canucks

Hardware Canucks (http://www.hardwarecanucks.com/forum/)
-   Storage (http://www.hardwarecanucks.com/forum/storage/)
-   -   FancyCache... Ram cache for SSDs/HDD's. (http://www.hardwarecanucks.com/forum/storage/41516-fancycache-ram-cache-ssds-hdds.html)

lowfat March 10, 2011 04:43 PM

FancyCache... Ram cache for SSDs/HDD's.
 
FancyCache Overview

Found this over @ XS. Did a bit of playing around with it. And it does work very impressively. DO NOT USE IF YOU ARE WORRIED ABOUT DATA LOSS/CORRUPTION. My gaming system is rock solid stable and I never have power outages so I thought I'd give it a go. It is a 90 day trial beta. The software even supports TRIM, although I have no idea how that works. It should help w/ lengthening NAND life as well due to eliminating most of the small file random writes.

This is on a single 30GB Intel X25-E. Cache is set to 2GB.
http://hostthenpost.org/uploads/878e...9b329c9f1c.jpg

This is a 4k 100% random read w/ a queue depth of 1. After 5 minutes.
http://hostthenpost.org/uploads/e185...eff0db9567.jpg

And this is a 4k 100% random write w/ a queue depth of 1. After 5 minutes.
http://hostthenpost.org/uploads/b508...8863740b30.jpg

If you are familiar w/ Iometer you will know that these are outstanding results. And the fact that it kept up after 5 minutes is pretty impressive. Probably around 10x higher than w/o using FancyCache.

Just a quick run I deleted to do a/ a mechanical drive. Drive is a Seagate 1TB 7200.12.
http://hostthenpost.org/uploads/5f94...047b23c586.jpg

ilya March 10, 2011 04:51 PM

In theory this should throw most of the Windows disk thrashing onto your RAM right? I've been looking for ways on top of AAM to reduce HDD noise, I can't find any RAM drive software I like, but this looks promising.

Supergrover March 10, 2011 05:34 PM

I am just starting to try this out but yes so far impressive results as you stated 10x read/writes on my ssd.
Only done 2 runs so far still getting familiar with the program but so far impressive results.

ilya March 10, 2011 06:16 PM

1 Attachment(s)
Just did a quick CDM test, reads are all through the roof, but writes are still bottlenecked by my HDD.

lowfat March 10, 2011 07:21 PM

Quote:

Originally Posted by ilya (Post 494362)
Just did a quick CDM test, reads are all through the roof, but writes are still bottlenecked by my HDD.

Did you turn on write cache?

ilya March 10, 2011 07:35 PM

Quote:

Originally Posted by lowfat (Post 494371)
Did you turn on write cache?

I did, and after the restart it shows that its enabled, but the included performance monitor isn't detecting any write caching. Going to experiment with it tomorrow, going to leave my system to fold for the night.

Supergrover March 11, 2011 09:02 AM

3 Attachment(s)
Okay ran some tests this morning.

First image is my Corsair P64 ssd (cmfssd-64gb2d)

Second image is a Seagate 500g (st3500320as)

Third image is a (wd5000aaks-00v1a0)

CMetaphor March 11, 2011 10:13 AM

Is this actually legit??

I mean, SuperGrovers #s all look the same, whether its on an SSD or a HDD. Not making sense to me somehow...

lowfat March 11, 2011 10:16 AM

Quote:

Originally Posted by CMetaphor (Post 494523)
Is this actually legit??

I mean, SuperGrovers #s all look the same, whether its on an SSD or a HDD. Not making sense to me somehow...

Because the files he are testing completely fits in the cache. Real world performance isn't going to be anywhere near like that. But it should still give a notable improvement.

LarkStarr March 11, 2011 10:29 AM

yeah these benchies don't apply to the application. there needs to be activity that's so random and/or large such that it doesn't always hit the cache. (IoMeter I think covers this well?)

But it does indeed show that if you're using your computer such that most of your activities hit the cache, the increase is nothing short of fantastic.

L2ARC for windows?


All times are GMT -7. The time now is 06:51 PM.