Corsair Force GT SATA 3 RAID 0 Performance

[Speedway]

Well SSD space on my 120GB Force GT was starting to get low, so instead of reinstalling games or large programs onto my HDD, I did what any SSD speed addict would do and grabbed another 120GB Force GT I figured I would get a performance boost, to go along with doubling the space, which was of course the main reason for adding the 2nd SSD! But, I really didn't think I would see this kind of jump in CrystalDiskMark. I know, I know...it's just a benchmark, but I'm a benchmark junkie

 

Single 120GB Corsair Force GT with 13GB of space remaining and Windows installed:

 

Single 120GB Corsair Force GT - Completely Empty:

 

(2) 120GB Corsair Force GT SSDs in RAID 0 with everything copied from my original using Acronis:

[Waco]

Is that with random data or 0/1 fill? Either way, nice improvement from the single drive! I did the same thing with my drives.

 

Here's what my Vertex 3s do with a 128 KB stripe (random data, of course). I only have it on a single iteration but I've never seen the numbers change going from 1 to 5 (and I don't want to waste cycles by benchmarking all the time ).

[Speedway]

I always just use random. Yea, I will never complain with close to a 50% increase in performance when upgrading anything for my PC I actually forgot to change the iterations and didn't even realize it till you sd something I don't like wasting the cycles either

[ccokeman]

Nice speeds

[MJCRO]

I'm jelly.

[PremiumAcc]

Thanks for the benchmarks. This further supports what I have been saying about SSDs in RAID, that it does not improve random 4k performance, which is the most important to daily users who are not reading/writing large blocks of data.

Edited February 15, 2012 by PremiumAcc

[Muchoman1]

Thanks for the benchmarks. This further supports what I have been saying about SSDs in RAID, that it does not improve random 4k performance, which is the most important to daily users who are not reading/writing large blocks of data.

 

this is true, but remember that (usually) dual 120gb are cheaper than the same model 240gb, and you get more speed with raid 0 than with the higher capacity drive. the only main drawback is that there is no TRIM, but thats getting fixed soon.

 

i got dual force 3 (non gt ) here, got them for 150 each, so now i got 240gb of fast ssd for less than the price of a 120gb vertex 3 at launch

[PremiumAcc]

this is true, but remember that (usually) dual 120gb are cheaper than the same model 240gb, and you get more speed with raid 0 than with the higher capacity drive. the only main drawback is that there is no TRIM, but thats getting fixed soon.

 

i got dual force 3 (non gt ) here, got them for 150 each, so now i got 240gb of fast ssd for less than the price of a 120gb vertex 3 at launch

 

True, it usually is cheaper by ~$10 - $20. But not enough to justify the loss of TRIM and the reduced reliability.

Prices at launch for any piece of hardware are generally higher.

 

Edit: Also, the Force 3 uses cheaper NAND--the asynchronous NAND--which also contributes to the price difference.

Edited February 15, 2012 by PremiumAcc

[Muchoman1]

True, it usually is cheaper by ~$10 - $20. But not enough to justify the loss of TRIM and the reduced reliability.

Prices at launch for any piece of hardware are generally higher.

 

Edit: Also, the Force 3 uses cheaper NAND--the asynchronous NAND--which also contributes to the price difference.

 

im aware of all those points, but intel is releasing trim for raid 0 soon, i know the force 3 uses cheaper NAND, and im aware that prices go down, but my point is, performance for your dollar, raid 0 is cheaper than the larger drive

[PremiumAcc]

im aware of all those points, but intel is releasing trim for raid 0 soon, i know the force 3 uses cheaper NAND, and im aware that prices go down, but my point is, performance for your dollar, raid 0 is cheaper than the larger drive

 

True. But not by much.

What I want to know is why not use them as two separate drives? Most of the time, users won't be doing large file transfers, therefore there is NO performance increase. Furthermore, I do not see this as being the most practical use of the SSDs since there is no "real" performance increase, the loss of TRIM, and the reduction in reliability.

[Waco]

Thanks for the benchmarks. This further supports what I have been saying about SSDs in RAID, that it does not improve random 4k performance, which is the most important to daily users who are not reading/writing large blocks of data.

You must be looking at different numbers than me...the 4 KB results for a pair of drives is essentially double at any queue depth higher than 1.

[Muchoman1]

True. But not by much.

What I want to know is why not use them as two separate drives? Most of the time, users won't be doing large file transfers, therefore there is NO performance increase. Furthermore, I do not see this as being the most practical use of the SSDs since there is no "real" performance increase, the loss of TRIM, and the reduction in reliability.

 

so i have 240gb of c drive.

 

trim is coming to raid 0