Solidigm

idata · ‎08-18-2010

HardWare

laptop Asus G60Vx (4Gb 800Mhz 2Quad 2.0)

Os - Windows 7 (64bit)

controller ICH9-M - set to Raid 0 strip size of 128 KB

partition - 100Mb market as system by W7 - (no letter) + rest as C with all + pagefile fixed size of 8Gb + always more the 20% free space

(tested also with a C partion only with pagefile and another with Windows files + another formated or not formated - no letter; try it also to format the SSD with HDDErase 3.3 & W7 install or ONLY by W7 install = which is the current one)

the RST drivers & program are the last version (iaStore.sys)~, write cache enabled

the SSD was delivered with the last firmware

superfetch and prefetch disabled, defragment schedule & hibernation & system restore off and others recomanded settings.

the TRIM is working! (in any case W7 reponse as it ok; fsutil behavior query DisableDeleteNotify =>DisableDeleteNotify=0 TRIM is Enabled.)

word noting: I was done several copy's & delete of files from 10Mb to 150Gb. the benchmark was the same after deletion of the files

From the begining (3 months ago), the benchmark on AS SSD was

Seq: read ~ 320-330Mb, write 195-205Mb

Acc time ~ read 0,15-0,25! write 0,09~0,12

because I was done several test and despide the optimization the perf. was similar (except write cache on or off), my simple question is:

Are these perf. the best (max) of ICHR9-M ?

idata · ‎08-18-2010

The fsutil behavior query only tells you that the OS is sending the TRIM command. Unfortunately if you are running raid the command does not get passed through the controller so TRIM will not be working. Intel are working on a driver update that will pass on the TRIM command but to date there is no ETA.

Current performance degradation can only therefore be recovered with a secure erase in a raid configuration. Performance degradation is related to the extent of writes that occur. If you are not constantly writing heavily to the drive you should not notice much degradation.

Edit: There is some good advice on "optimizations" here:

http://blogs.msdn.com/b/e7/archive/2009/05/05/support-and-q-a-for-solid-state-drives-and.aspx

idata · ‎08-18-2010

thanks redux,

the link you have posted was my second reading for SSD.

they are also

http://thessdreview.blogspot.com/p/windows-7-ssd-performance-optimization.html

and more like these on forums.guru~, zikmao.net and so one.

I don't claim "the Trim works on my Raid 0 config", I say only is enabled and sent to raid controller.

I don't claim I have archived the best optimization W7 in town / world.

I say / claim / ask / cry only:

Why [... put all knowing injuries here...] - the bench are the same

read: 325 +/- 4Mb, Acc. time 0,15-0,25

write: 202 +/- 3Mb, Acc. time 0,09-0,15

IS A CONTROLLER LIMITATION ?

note:

several days I was copied DVD iso files or tib files ~ from 4-5Gb up to 150Gb; tested and deleted after

idem, for small files => PortableApps (copy / read / write delete)

the bench are the same like in the first day

except write perf. with write cache enabled (a plus of 20-30Mb)

idata · ‎08-18-2010

Your sequential read speed and access times are below optimal, but I don't know if they are below optimal for your particular set up. I've seen read speeds around 500MB/s and access times ~ 0.09 on an ICH9 workstation.The sequential write speeds look OK. At a guess the read speeds are saturating your set up.

idata · ‎08-18-2010

re-thanks redux (for survey & quick answer)

I was checked the bench for ICHR9, which are between 400-500Mb read & 200-250Mb write with less then 0.10 acc. time.

Why on mine is less? despite all the different setups ?

It's the "M" ? My controller is ICHR9-M (=mobile).

Before to put this query in forum, I was spent several hours in order to find the bench for ICHR9-M ... without luck

I hope to find someone (from Intel?) which know and has the amiability to share his knowledge with me.

note:

In W7 is always 7.9 (max)

Solidigm

strange benchmark/comportment for 160Gb 25-M G2 Raid 0 W7 (64)