Raid0 trim ssd, confused

[Somebody_007]

So I'm lost here.

From what I understand trim is not supported in raid0, yet similar cleaning routines are?

If this info is correct what real world performance differences are we talking about here?

Anybody know any good reviews or benchmarks around this subject?

Also can anyone reccomend a raidcard for 2 60gb SSDs in raid0(if this is a good idea depending on the above).

I like the highpoint rocket 620 but they don't sell it on european amazon any replacement suggestions?

 

[RetiredChief]

1) Trim is not passed to raid0 member drive. Must rely on Garbage Collection (CG). While this works, it is not always the best and it works much bettery when used in conjunction with trim.
2) Raid0 is NOT the recommended route for a OS + Program SSD (Andy I LOVED Raid0).
.. Raid0 Improves Sequencial read write performance, This is The LEAST important matrix for a OS + Program drive. Raid0 may make sense if used as a data drive and working with LARGE data structures.
.. Raid0 Does NOT improve access time, Important when a large number of small files are read such as boot time and program loads. This is also why there is little improvement in random 4k read/writes, which IS important for OS load and program Load time.
.. Already covered the loss of TRIM.

.. A larger SSD performs faster than it's smaller brother (ie same family). So a 128 gig M4 SSD is faster than the smaller 64 gig M4 SSD M4 used solely as an example, substute any other model. This then subtracts from any gains that raid0 MAY have.

PS On This forum there are some that swear by SSD, But a higher number that discourge it (some of them have tried it it and gone back to single drive).

Myself I have two SSDs in both desktops and one Laptop. In all three I have set them up as a OS + Program drive and a "work" storage drive. where I keep my most often used files on.

 

[Somebody_007]

So you would not reccomend raid0 from what I hear?(for an OS drive)

So from what you are saying I understand that the performance loss in raid0 will be greater than in normal mode(obviously), but the drive's performance won't get destroyed without it?


I understand that random reads are indeed the most important for loading, however according to these benchmarks and some others I found the improvements are still large.(almost double in fact)

http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/51455-intel-520-240gb-ssd-raid-0-performance-review-10.html

Would I not see these kind of gains(or anywhere near them) when I would go 2x60gb instead of 120gb?

 

[RetiredChief]

Indeed, does look like the raid0 improves the random 4 K (looked at AS SSD review as I do NOT place much stock in ATTO benchmark) will need to look to see if I can find some results using PC Vantage. I know that in mechanical drives, Raid0 does NOT improve access time and only a small gain in random 4K. Benchmartks are one thing, and real-world usage do not always fall hand in hand - reason I prefer PCMark vantage test based on apps that I use. The only use for ATTO is to compare your performance to manuf specs.

Trim and lack off trim seems to effect drives differently, dependent on usage and on How CG is implimented. There have been several post, and some mention in reviews, that you MAY (key word) have to let the drive sit idle for several hours for CG to effectively bring a drive that has reduced it's performance back to Normal performance.

PS you did notice they were comparing a Intel 520 240 Gig drives (one raid0, one not), I doudt that a pair of 60 gig dives will score equally as well.

Added: Trim support for member drive of raid0 (applies only to intel chipsets) may NOt be to far off in the future.
http://news.softpedia.com/news/SSDs-in-RAID-Get-TRIM-Don-t-Lose-Performance-Anymore-235606.shtml
Also caught this "Solid state drives are a rather bad choice for RAID setups"

Another reference:
http://www.xbitlabs.com/articles/storage/display/kigston-hyperx-ssd-raid0_5.html#sect0
Using Pc mark7 there is an advantage to Raid0, Using a Kingston 120 gig. 2 in raid0 yields a wapping 4.2 % increase, and when comparing a 240 gig drive to 2 x 120 in raid0 the diff drops to 2.31 % - Doudt thats enough to notice.
From Conclussion:
Quote
Thus, a RAID0 will only be superior to a single SSD at linear operations whereas random-address operations will expose its weakness. That’s why we can’t prefer the RAID0 solution to a single SSD without reservations. On the other hand, most of our lifelike benchmarks do show the RAID0 to be overall faster. In other words, the RAID0 is better on average, especially as it doesn’t involve any investment: the cost per gigabyte is the same for a RAID0 and an SSD of the same capacity.

There is some inconvenience about running an SSD RAID0 that should also be mentioned. You cannot monitor the health of your SSDs in a RAID0 or update their firmware. A RAID0 will also have lower reliability since a failure of any SSD causes the loss of all data stored on all the SSDs in the array.
End Quote.

 

[Somebody_007]

Yeah I never like artificial benchmarks either. I never understood why really so few hard drive and ssd reviews include very few to no real world benchmarks. They may be less reliable, but at least they mean something tangible.

Anyways thanks for all the info. This build isn't for me personally(raided caviars ) I'm building for a friend, so I guess I'll reccomend the single 120gb seeing as the advantages seem to be smaller than I expected. Especially considering that 60gb raid0 would cost 30-40euros more than the 120gb solution.

The trim for raid0 seems promising though, that might finally make me upgrade myself. I'm the geek in the house and I installed SSDs in all of my family members's PCs yet I'm still running mechanical ^^. Really looking forward to this trim in raid0.

And lastly would you happen to know how Sata III performance is on entry level p67 motherboards? I know that on my ud5 it'S very bad on I'd need an add-in card for decent performance is that still the case with the newer mobos?

Thanks again!

 

[RetiredChief]

As long as you do NOT stick on a Marvel based SATA III controller you should be fine (use the Intel SATA III controller.

A little more info on sata III SSDs. Some SATA III SSDs perform NO better when on a SATA II port than when ona sata III port. Case in point is the Agillity III SSDs. Caught a review comparing the Agillity III on Sata III port vs on a SATA II port. It used AS SSD benchmark and there was almost NO difference. As I had a pair of 128 gig Agility IIIs, i check using mine ans sure enough they only score mid 400s on sata II, about what they score on the Intel SATA II port.