• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Enter The Matrix: Slice out and get the best part from your hard drives

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
At last...

For the sake of posterity (I think it is)... here is a direct comparison between a 3 x 36Gb Raptor Raid0 matrix array and a 3 x Seagate 500Gb 32Mb cache perp matrix array.

I left the 3 x 36Gb Raptor C partition as it was in the Matrix (49Gb) but shrunk it using Vista to 25Gb and ran fresh tests to compare those results with the Seagates, the only major difference being the size of the raid0 partitions... the Raptor one is 49Gb in the matrix with a 25Gb C drive in that. The Seagate one is a 12Gb matrix partition, used fully for D drive.

AttoRaptorVs32MbCachePerps.jpg

HD-TuneRaptorVs32MbCachePerps.jpg

HD-TachRaptorVs32MbCachePerps.jpg

I have to say, I am very impressed with the Seagates, even though the 12Gb partition is completely empty (I'm gonna install XP-Pro in it) and all the tests were run from my Vista 64-bit C-drive install.

Enjoy!
 
Last edited:
Makes me wonder why all raptors are sata150... when i enabled sataII on my 320 perps i saw an increase in performance... it was somewhat sufficient too, but more in burst speed, though it did have a slightly significant effect on avg read speed.

fritz, isn't a smaller raid0 partition going to have much better performance on a matrix array since you're using the fastest part of the disk? 25gb vs 49gb is still a difference... but nice comparison.

now, do i sell my 3x320s and get 3x250s 410as or no... i would even do a straight up trade with someone if i could...
 
fritz, isn't a smaller raid0 partition going to have much better performance on a matrix array since you're using the fastest part of the disk? 25gb vs 49gb is still a difference... but nice comparison.

Thanks.

I agree... as soon as I get back, I intent to image the C drive, then re-create using Ctrl+I an actual 25Gb partition, and then send the image back to that.

I am picking it will be quicker.

B
 
Hardware: Conroe E6600
HDD: 2x Baracuda 320gb

Raid 0: 80 gb (Should have made smaller, but I still got amazing results)
Raid1: Rest



Raid 0 Benchmark:
raid0ix8.jpg


Raid 1 Benchmark:
raid1ix5.jpg



:drool::drool:
 
Seeing as I might not always be on an ICHxx and wondering what other benefits a dedicated controller like the popular Areca...anyone make the switch? I'd imagine lower cpu usage, but that doesn't bother me. I'm mainly interested in perf and being able to take your arrays w/ from board to board.
 
Yep, Evil is right, currently no there is no other hardware or software offer this hybrid raid implementation except Intel.

Matrix Raid concept isn't new but it was first implemented and introduced at Intel ICH6R South Bridge.

I guess that someone at linux camp will come up with this matrix raid in the future purely on sofware.
 
Last edited:
I added a drive and migrated from Raid 0 to Raid 5 using 3 sata drives now. I like the performance of Raid 5 a lot plus the piece of mind of redundancy is great. Using xp64 too w/ 4gb ram. My temps have gone up a bit on my quad core but I did read that Raid 5 is a bit more demanding on the cpu than other raids. I will be adding a 4th drive probably next week since I have also read that going from a 3 drive raid 5 to a 4 drive raid 5 increases performance by about 25%.
 
Last edited:
DaPoets,

Mind post/share here some benchmark numbers for the before and after migration from 3 to 4 drives ?

It will be a valuable info for all of us. TIA.
 
Thanks, one small thing but important to watch is, try maintain consistent environment as possible for both like the same enabled or disabled background tasks like AV, firewall, services and etc.

I'm interested to see the difference in the CPU utilization.

Also if it's not troubling you too much, mind do the PCMark benchmark as well & post it at ORB ? Since it reflects closer to real life performance rather than those low level benchmarks.


Edit :

Whats those drive's brand & model ? Can't find it in your sig ! :D
 
Last edited:
Back