SSDs performance issues,slow speeds,bad transfer rates,solving problems official thread

the write is enabled auto by win 7

 

Does actually disabling ReadyBoot affect the performance with SSD, more maybe harm the SSD less?:

 

Thought this might be worth mentioning regarding C1E power state and SSD 4k write performance. Applies to HP Envy, but maybe in general:
http://forum.notebookreview.com/hp-...-owners-lounge-part-deux-374.html#post6656285

 

Try disabling it. I saw one user getting better scores without. (while normally it should be enabled)

 

With Musuta's issue:
Could it be that TRIM was/is not passed on to the SSD?
If so, probaly only wiping the SSD with Secure Erase or another tool that overwrites all sectors with 0 would help.

 

^ That's a good point.

 

Hi guys
Phil by disabling it will Disasterly reduce performance ..
Enny Trim is on ...
and Secure erase is not supported by my SSDs

Anyway guys I just got bloomin pissed off i made a replacement I purchased better Vaio With RAID Feature and it looks pretty daaamn good ))))
but I need to ask some questions according Intel matrix ..
they say that I should make strip size 16K fro ssd's and I have set 128k

please look on the pictures

 

Anyone know what's causing my varied CDM scores?

I have an Intel 160GB G2 SSDon a dv6tse.
When I changed from "High performance" setting to "HP Recommended," my 4k random read/write scored jumped from 15/23 to 18/30. I restarted my computer again, still on HP Recommended, and now they are back down to 15/23. I don't know what I did.

Later, what I found now was that the first run after changing the power settings is faster, and the second run is slow again. I got a faster results for High Performance now, so I guess that's not what's slowing it down. My runs are all over the place now. For example, I got one run of 20/23, and even as fast as 20/37, but also a 14/22.

 

more pics
one ssd is connected to the Optibay

 

What SSD's do you have in there now?
The stripe size should be usually 64KB (128sectors). Because that would fit all possible cluster sizes in one stripe (512Byte to 64KB).
More important is that you need to align your disk/partitions, easy way is probably to select GPT instead of MBR when you initialize the disk/write the signature.
I'm not sure why they suggest 16KB for SSD's I assume that they think the cluster size will be rather smaller then bigger and having a stripe size that is not much bigger than the cluster size would be better for the SSD. But that is only my theory about that.

@eight35pm when you run benchmarks on the drives from where you started Windows you'll always get different results because the OS and background apps are constantly reading and writing to the drive. If you connect your drive to another computer and run the benchmark then, you'll see other results and more consistent.

Also you should not run the benchmarks that often, especially writing, because it'll harm the SSD-cells life cycle (not from a handful of tests but dozens or hundred of them).

 

will enhance speed with adding these two

 

That are still the same type of the Samsung drives.

 

Why not buy RealSSD C300?

 

Enny
yeah same ssds and look on one of that my pictures where is help from Intel
they saying set stripe to 16k when u using ssds ..

Why not buy RealSSD C300? - man i bought these two one cost me 180 and second 170 so it is very good price
and I neet to buy two more 1.8 SSD's and I will have QUAD SSDs .. and it will be Best performance ever and probably I will beat Vaio Z

 

Ah, thank you very much. Just to clarify, this explains why the first test is faster than all subsequent tests? Or do you mean the tests would be randomly inconsistent?

And thanks for the warning about running too many benchmarks. I was afraid I might be doing too much, so I won't run any more. I wont' be doing tons of writing to the SSD on a daily basis, so I figured it shouldn't really be a problem, but it doesn't hurt to be careful. It should last 3 years, right?

 

@Enny

Also you should not run the benchmarks that often, especially writing, because it'll harm the SSD-cells life cycle (not from a handful of tests but dozens or hundred of them).

Really True ! this

 

@eight35pm it will be randomly inconsistent.
Well running some tests when you get the drive and when you want to optimize something is fine, or once a month to see how it is going when you notice a performance drop of your computer.

I have a RealSSD C300 in my vaio and that is faster than the two Toshiba SSD that came with the notebook.

Keep in mind when you do RAID on the Intel controller, TRIM is not supported.

 

Try running CDM with HyperPI 32M in the background. If your scores go up then it's a power saving issue.
To reduce wear on SSDs you can run 3x 50MB test size in CDM and only run the 4K random tests.

Sandforce SSDs are a bit faster on SATA II connections.

 

My Sandforce is currently underperforming. Is it worth it to perform a secure erase?

If I do, will I have to reformat, or will I simply be able to backup my data to an external HDD and then copy it back?

Can I copy my data to an external HDD, defragment it, and then put it back on the SSD, so it's packed tighter? Or does this not matter?

 

Have you run Atto? What does it say?

You did do a clean install right?

 

I did do a clean install. I also performed the tweaks you suggested. I haven't run Atto yet, but I should be able to by the end of the day What's the difference between it, CDM, and AS SSD?

 

OCZ Tony says ATTO is the better tool for Sandforce SSDs. I think his reasoning it that it uses compressable data.

Anyway, if you don't get normal performance in atto too then it's time to take some action. Asking for assistance on Crucial forums is what I would do.

 

Sure, always use the benchmark tool that shows better results. "Never trust a benchmark result unless you have manipulated it yourself"

 

His CrystalDiskMark sequential speeds were limited at 200 MB/sec. If the same limit shows up in other benchmarks I'd say there is a problem. If it doesn't I'd say there isn't a problem.

 

What settings should I use for ATTO?

 

Default and then compare with other Sandforce reviews.

 

Ok, here it is:



I'm just confused as to why other Sandforce drives are getting much better benchmarks on CDM (especially sequential read/write).

 

Some limit in your ATTO results too. See here:
Corsair Force Series 120GB SandForce SF-1200 SSD Review - ATTO v2.41 & AS SSD Benchmarks - Legit Reviews
Corsair Force 120 (F120) SSD review

If you have another notebook it would be interesting to see if you get the same results. If so I'd ask on the Corsair forums, see what they say.

 

I ran ATTO



hey Phil I inserted picture here and it wont appear .. why ? :-/

 

Ok, I might post on the Corsair forum later. It seems fairly disorganized

Meanwhile, could anyone give me some insight on these?

Will I have to reformat after a secure erase, or will I simply be able to backup my data to an external HDD and then copy it back to a clean partition?

Can I copy my data to an external HDD, defragment it, and then put it back on the SSD, so it's packed tighter? Or does this not matter?

 

After the secure erase you can just start installing Windows 7.

Cloning is not recommend because of the possibility partition allignement problems.

There's no need to defragment your data.

 

Even if I have a single partition?

 

I thought so. + Your current install is not giving the performance you want. So I'd definitely do a clean install.

 

When I got the drive the first thing I did was a clean install, followed by the tweaks you suggested. Unless the secure erase is what will give me extra performance, I'm pretty much out of luck as far as I know.

 

Well to determine whether it's your notebook or the SSD you'd have to test it in another notebook.

Lately I've seen many examples where it was the notebook/controller at fault, while people thought it was the SSD.

 

I'm actually starting to think that it is my notebook's fault. There might be something bad with it besides having the HM55 chipset (which explains the poor random write/read speeds).

 

You might want to ask other Acer 4820 owners to see what kind of SSD performance they're getting.

 

ATTO run test and Crystal and AS SSD

guys if you know more tests please give me the names

thnk you

 

@ KolosoK, I've seen some more results on dutch forums and it seems that your CDM results are quite normal.

If you click in CDM on file, test data and change it to 0 fill or 1 fill your results should improve.

 

Hi Phil
I wanted ask.. how these benchmarks works ? how they are made to measure ?
For example I have SATA II and benchmark softwares and programs showing me these results like over 400 MB/s etc ..
But SATA II is up to 300 MB/s
I would need SATA III to fully use speed above 300 MB/s
How we should understand these tests ?
Thnx

 

I'd guess you are using two SATA II ports? Otherwise it would be bottlenecked, you're right about that afaik.

 

yes two SATA II ports

 

hello Phil
I got another BIG problem AGAN aaaagrrrrhh !!
I just purchased and received two 1.8 samsung SSDs 128GB they are gorgeous !
BUT tere will be always but .. aach .. cant belive it !!
I installed the SSDs and am gutteeed !
I have got these 4 SSDs
and when I got into RAID config i was thinking i wil get heart attack !!
I CANNOT CREADE QUAD RAID WHYYY????? NOOOO (((((((((
I managed to create RAID 2 in windows using matrix

please check pictures ..

 

continue from above
I really cant understand why I cannot make Quad RAID ??????
there says : select two disks why there is not select 4 disks????

 

According to Intel ( Intel® Rapid Storage Technology (Intel® RST) — What features are supported on each I/O controller hub (ICH)?), your ICH9M-E chipset cannot support more than 2 drives in RAID.

 

Can i upgrade it somehow?? or hack it??

 

I don't know. I wouldn't think so, but it may be possible for someone with better knowledge of the inner workings of chipsets than myself. I think most of the people that run more than 2 drives in RAID, though, at least on the desktop level when their chipset doesn't support RAID go in for separate daughter cards that work as full hardware RAID.