6_6_6

NCQ: Best Upgrade For a Power User!

Recommended Posts

FAT Punisher, I cannot vouch for that drive.

I have not tried this drive and no one posted its tests. The one I have is 750GB. The ones czr tested are 500, 750, 1000. 7200.11 yes, but that one... We simply do not have the data.

Why don't you try with a single drive and if you are satisfied, you get the other. Don't you think that would be better?

Next, I'll try two Seagate 7200.11 640 GB. Let's hope they're really as good as you state. ;)

Share this post


Link to post
Share on other sites

Please note the tests done by xerces8 (on another thread).

He verifies that all these drives (even archaic ones) are working properly under linux.

Note that it is not all the disks fault.

For example here is how it behaves under Linux (RIPLinux 6.1 - 64bit kernel 2.6.26):

the HP nw8440 laptop (AHCI mode)

Disk : ST980825AS

1 instance : 42 MB/s

2 instances : 36 MB/s

9 instances : 35 MB/s

Windows values were : 42/11/15

--

HW: P5K-E WiFi

BIOS setting : AHCI

Disk : WD7500AAKS

1 instance : 96 MB/s

2 instances : 78 MB/s

9 instances : 75 MB/s

Disk : Samsung HD642JJ

1 instance : 114 MB/s

2 instances : 92 MB/s

9 instances : 89 MB/s

In Windows it was: 99/90/26 and 116/14/25

The test were instances of this command:

dd if=/dev/sda of=/dev/null bs=256K skip=xxxx # values for offset

measurement with : iostat -c sda

and my response:

xerces8, thanks a bunch. this is very interesting find. I will test my Samsung on linux when i have time.

Conlusion would be: Drives implement NCQ properly...but:

a ) Microsoft/Intel conspired to support only Seagate 7200.11 drives on NCQ front.

b ) Seagate is the only hd manufacturer to find the problem in Intel/Microsoft NCQ implementation and fixed it at their firmware (since 7200.10 does not work properly) or altered their NCQ implementation.

Whatever the scenario / configuration / drivers / controllers are, nobody reported NCQ working properly on windows except 7200.11s thus far. Hence, there is no choice.

I don't have time, but i will test drives under linux in future.

Share this post


Link to post
Share on other sites

LittleJhon,

Please see the linux findings above. If that is so, apparently all these drives implement NCQ fine... Rather the blame lies on Microsoft than the drive makers. So we have been blaming the manufacturers unjustly all along. But kudos must go to Seagate for finding a workaround in the firmware, NCQ implementation etc and making them work.

I will test these drives under linux and freebsd and post the results next month.

Next, I'll try two Seagate 7200.11 640 GB. Let's hope they're really as good as you state. ;)

I'm at the point where I don't care what it costs, as long as I get the performance I want. :D

The parts I don't need will be sold at eBay... <_<

Hi Punisher, I posted this PM to you on the Hexus forum,

I read your thread regarding the 6400AAks, about AHCI and the NCQ performance. You mention that you have had some emails/correspondence from western Digital and with intel. Could you possibly post these on the forum, as I feel I have been ripped off by the manufacturer. I brought the WD on condition that it had good NCQ and AHCI performance.

You have mentioned that intel say that no one else has contacted them, and that WD have simply passed off the problem. They should be finding thier own solution if necessary, because they have our money. I would like to get some more people to write to them with this problem, so please provide the contact details, as I want the same department to start recieving these problem emails to put a little pressure on them to find a solution.

Thanks Punisher

If there is anyone here, who would like to join me in contacting both westerndigital and also intel about this, let me know, and we shall send in a combined message to make them find a soultion, or resolve this issue either way.

Share this post


Link to post
Share on other sites
LittleJhon,

Please see the linux findings above. If that is so, apparently all these drives implement NCQ fine... Rather the blame lies on Microsoft than the drive makers. So we have been blaming the manufacturers unjustly all along. But kudos must go to Seagate for finding a workaround in the firmware, NCQ implementation etc and making them work.

I will test these drives under linux and freebsd and post the results next month.

Next, I'll try two Seagate 7200.11 640 GB. Let's hope they're really as good as you state. ;)

I'm at the point where I don't care what it costs, as long as I get the performance I want. :D

The parts I don't need will be sold at eBay... <_<

Hi Punisher, I posted this PM to you on the Hexus forum,

I read your thread regarding the 6400AAks, about AHCI and the NCQ performance. You mention that you have had some emails/correspondence from western Digital and with intel. Could you possibly post these on the forum, as I feel I have been ripped off by the manufacturer. I brought the WD on condition that it had good NCQ and AHCI performance.

You have mentioned that intel say that no one else has contacted them, and that WD have simply passed off the problem. They should be finding thier own solution if necessary, because they have our money. I would like to get some more people to write to them with this problem, so please provide the contact details, as I want the same department to start recieving these problem emails to put a little pressure on them to find a solution.

Thanks Punisher

If there is anyone here, who would like to join me in contacting both westerndigital and also intel about this, let me know, and we shall send in a combined message to make them find a soultion, or resolve this issue either way.

Thanks 666,

You clearly have a point. But if seagate can do it, western digital can do it. Bottom line. It has been shown possible. Bottom line, we should confront WD and on compling the evidence, confront intel. This is supposed to be a WD flagship model, in 2008. If they expect big enterprises and corporations to fill their profit and loss statement with lots of lovely pluses, from all around the world, they should deliver.

That said, intel might need to be made aware, that forum users can dent thier market share. To FAT punisher, intel said no other users have complained. They are saying - we don't know OR we can be botherd until it has real consequences. Okay then, lets provide them with the information they need to move on this!!!!

I can't wait to see your results on the other OS's, please see my response as a request for help from forum members.

LittleJhon

Share this post


Link to post
Share on other sites

Okay, an old (4+ years i believe) 320GB WD Caviar:

Windows:

63 MB/s: 1 instance

28 MB/s: 2 instances (both at 0% position)

9 MB/s: 2 instances (0% position and 90% position)

15 MB/s: 10 instances (10% gap between each)

Linux:

63 MB/s: 1 instance

63 MB/s: 2 instances (both at 0% position)

45 MB/s: 2 instances (0% position and 90% position)

48 MB/s: 10 instances (10% gap between each)

Linux, stock install of Fedora 9 x64 (2.4.25 kernel).

Linux commands issued:

dd if=/dev/sdb of=/dev/null bs=256K skip=200K (skip was incremented by 200K each instance which corresponds to about 50GB forward in this 320GB drive)

all dd invocations started at the same time (batch).

throughput measured with:

iostat -m 1 /dev/sdb (m shows in megabytes, 1 is the update interval of graph every second)

Windows commands:

dd if=\\.\PHYSICALDRIVE1 of=NUL bs=256K skip=200K (as above)

dd is the same utility both in Linux and Windows. Furthermore, windows results were verified with HD_Speed (they are identical to dd).

This was on jmicron controller. No diff in ICH9R either.

Conclusion: I am not even going to bother to test any other drives. We have been had! Good luck with Crapdows. Only Seagate 7200.11 manages to work properly. They probably found a hack around Windows and implemented it on their firmware.

PS: Please do not mess with 'dd' command. I take no blame if you wipe your drive clean with a single typo. Happenned many times before to people.

Share this post


Link to post
Share on other sites

Here is a PATA 200GB WD Caviar:

Windows:

44 MB/s: 1 instance

30 MB/s: 2 instances (both at 0% position)

12 MB/s: 2 instances (0% position and 90% position)

15 MB/s: 10 instances (10% gap between each)

Linux:

44 MB/s: 1 instance

44 MB/s: 2 instances (both at 0% position)

30 MB/s: 2 instances (0% position and 90% position)

33 MB/s: 10 instances (10% gap between each)

Umm... There is not such huge drops in PATA drives. No wonder I was right when i was saying for a pc assembled in 2008: "Why do I feel that this computer is much slower compared to previous one purchased in 2001 (3.0 GHz Northwood with Hyperthreading, 200GB EIDE Caviar, 2 GB ram, 2003 Advanced Server)?"

Apparently PATA does not take seem to take such huge throughput drops on multitasking... No one seems to notice anything i think everyone just uses their computer to open a browser or play a game. Just spend $5K to open Notepad faster! And type faster perhaps too!

Share this post


Link to post
Share on other sites

Wow, I had 30 MB/s on 2 instances at 0% postion on a 10 year old 200GB PATA drive while I had 20 MB/s for the same on a 2008 500GB SATA drive!

Share this post


Link to post
Share on other sites
Next, I'll try two Seagate 7200.11 640 GB. Let's hope they're really as good as you state. ;)

I'm at the point where I don't care what it costs, as long as I get the performance I want. :D

The parts I don't need will be sold at eBay... <_<

Hi Punisher, I posted this PM to you on the Hexus forum,

I read your thread regarding the 6400AAks, about AHCI and the NCQ performance. You mention that you have had some emails/correspondence from western Digital and with intel. Could you possibly post these on the forum, as I feel I have been ripped off by the manufacturer. I brought the WD on condition that it had good NCQ and AHCI performance.

You have mentioned that intel say that no one else has contacted them, and that WD have simply passed off the problem. They should be finding thier own solution if necessary, because they have our money. I would like to get some more people to write to them with this problem, so please provide the contact details, as I want the same department to start recieving these problem emails to put a little pressure on them to find a solution.

Thanks Punisher

If there is anyone here, who would like to join me in contacting both westerndigital and also intel about this, let me know, and we shall send in a combined message to make them find a soultion, or resolve this issue either way.

I think I may have figured this whole thing out!!!!!!!!!!!

ICH9 does not support AHCI or NCQ, but ICH9R does?????

Has there been a single user who has tested the WD6400AAKs on a ICH10R chipset????

Share this post


Link to post
Share on other sites

I think I may have figured this whole thing out!!!!

Apparently the ICH9 does not support desktop ncq, but the ICH9R does????

Bottom line, has anyone tested the WD6400AAKs on an ICH10 chip. Are all these posters only testing on a ICH9 chipset????

If you post a test, please include the model number of your sata controller, as stated in the manual!!!

Linky:

http://www.overclock.net/intel-motherboard...windows-2k.html

LittleJhon

Share this post


Link to post
Share on other sites

Your assumptions are invalidated by linux tests of xerces8 and me, LittleJhon.

I use ICH9R... and 500 GB Samsung and 320 GB WD did not do NCQ under Windows. But they both worked very well under linux.

This is not a hardware issue (drive, mobo, controller... all checked out fine, otherwise wouldnt be doing NCQ under linux). Purely OS-related.

FINAL CONCLUSIONS:

1. All drives support NCQ as advertised.

2. All drives do NCQ flawlessly under Linux.

3. Only Seagate 7200.11's do NCQ in Windows. No other drive does proper NCQ under Windows.

4. Multitasking throughput of a 7-year old PATA drive is better than almost all new SATA drives (except 7200.11s) under Windows. [personal experience, start point for this whole NCQ deal]

I had my NCQ working nicely since months now. I was just testing around for you guys.

This is my final conclusion. I am done. Good luck all.

Share this post


Link to post
Share on other sites

@6_6_6: I'll buy two 7200.11 640GB, but at first only open and test one of them. If it sucks, I can send back the other one for free.

It's true that nobody has tested the 640GB drive yet, but seriously: if all those other 7200.11s have NCQ working properly, this one will have it, too. ;) At least I'll soon be able to tell you. :lol:

@LittleJhon: after I'll have finished my tests with the 7200.11s, I'll write a summary about everything I've experienced. At the moment, I don't have the time either, sorry.

Here are some things to be noticed:

1. I have two problems with NCQ. 1st: the "normal" problem with NCQ not working properly, which everbody experiences (at least under Windows and with all drives except the 7200.11s). 2nd: the Intel ICH9R behaves exceptionally shitty, because it distributes the available "bandwidth" of the hard drives extremely unevenly. It happens that one (or several) of my applications freeze completely because they can really read or write zero bytes for some time (proven with HD_Speed: e.g. 100 MB/s for 1st instance and 0 B/s for 2nd instance, for a long time). This did not happen with other controllers in the past.

2. I could solve the 2nd problem with an expensive SATA/SAS RAID controller (Dell Perc 5/i, similar to LSI MegaRAID SAS 8408E). Using it, all applications get the same amount of bandwidth (proven by HD_Speed), although it's terribly little. But it feels like with my old nForce 4 chipset, those freezes are gone. Unfortunately, even with the Perc 5/i, NCQ does not work like it should.

Conclusion from 1. and 2.: Intel is not responsible for the "non-working NCQ problem". But at least in my case, it adds to the pain by somehow making everything even worse (much worse). With the 180€ Perc 5/i (used, from eBay), I finally have the same performance I had with my old nForce4 system. Considering that the Perc costs more than the nForce4 mainboard, not speaking of the new Intel chipset mainboard, this is a bad deal... <_<

3. A mate of mine has 12 (!) WD5000YS (RAID edition 2) in a RAID 6 on a 3WARE 9650SE-12ML (yeah, sick, I know). The total speed of 9 HD_Speed instances seems to be like it should be. It's comparable to the throughput of 1 instance (~400 MB/s). So this controller seems to be able to make up for the not working NCQ. (Note: I have two WD2500YD, RAID edition 2 as well, and they show the same non-working NCQ, so it's not because of the RAID edition 2 drives)

It could be this feature of the 3ware:

StreamFusion optimizes I/O accesses to maximize application performance under multiple stream loads

4. Windows Vista shows the same problems with NCQ (tested it), so even the newest Windows will not help us. Therefore, this problem will probably be haunting us for the next years. Considering this, the 7200.1s seem to be the only option.

It is very interesting that Linux does not have this problem though. Incredible. :blink:

Share this post


Link to post
Share on other sites
It is very interesting that Linux does not have this problem though. Incredible. :blink:

Could that be due to more suitable disk I/O scheduler for multitasking? Someone could benchmark also with NCQ disabled (/sys/block/sdX/device/queue_depth=1) and different schedulers (/sys/block/sdX/queue/scheduler).

Share this post


Link to post
Share on other sites
A fanless quad-core system...impressive!

I am using fanless computers for a decade now. I do have a Zalman Reserator XT (fan removed, pump wrapped) this time... It cools my SilentMaxx 500W PSU, Q6600, 8600GT and HD. And Q6600 is o/c'ed to 3.2 GHz. Not a single problem for almost a year now on this setup and it runs 24/7 and temps are within specs.

Yes noise is an issue... and this Seagate... is just not it. I had AAM disabled on my samsung by the way and yet it was inaudible on a fanless setup.

Well... actually. Seagate noise is not an issue no more :) It runs quite cool though... was 35C without water... now with water and insulated/wrapped... at about 50C idle. It is inaudible :)

Hi 6_6_6

I also have a Zalman Reserator XT - however - I have long been wondering if it were possible to remove the (noisy) fan! Could you please tell me how you have modified your Reserator XT - I will be eternally gratefull!!

/Teknisk

Share this post


Link to post
Share on other sites
A fanless quad-core system...impressive!

I am using fanless computers for a decade now. I do have a Zalman Reserator XT (fan removed, pump wrapped) this time... It cools my SilentMaxx 500W PSU, Q6600, 8600GT and HD. And Q6600 is o/c'ed to 3.2 GHz. Not a single problem for almost a year now on this setup and it runs 24/7 and temps are within specs.

Yes noise is an issue... and this Seagate... is just not it. I had AAM disabled on my samsung by the way and yet it was inaudible on a fanless setup.

Well... actually. Seagate noise is not an issue no more :) It runs quite cool though... was 35C without water... now with water and insulated/wrapped... at about 50C idle. It is inaudible :)

Hi 6_6_6

I also have a Zalman Reserator XT - however - I have long been wondering if it were possible to remove the (noisy) fan! Could you please tell me how you have modified your Reserator XT - I will be eternally gratefull!!

/Teknisk

Of course. There are many ways. I have a watercooled PSU and my way is not easy (osiloscopes and stuff).

The easiest (even that is going to require lots of mods):

As you are aware, if you remove the fan connector, the alarms will keep going off. So you can't just unplug the bloody fan.

Do you have a PSU with fan? If you do, does it have a connector to report fan rotation speed to mobo (most high-end PSUs have that)?

Share this post


Link to post
Share on other sites
4. Windows Vista shows the same problems with NCQ (tested it), so even the newest Windows will not help us.

I tested under XP, VISTA, 2003 and (maybe, I don't remember exactly, it was a few months back) 2008.

It was the same every time EXCEPT once.

In one test , with Windows XP SP3(!), it appeared to work without the slowdown.

I described this in the other thread (Slower with NCQ ?, Concurent access).

So if you did not yet, you should test with XP Service Pack 3. Maybe Microsoft fixed their scheduler there !?

Regards,

David

Share this post


Link to post
Share on other sites
4. Windows Vista shows the same problems with NCQ (tested it), so even the newest Windows will not help us.

I tested under XP, VISTA, 2003 and (maybe, I don't remember exactly, it was a few months back) 2008.

It was the same every time EXCEPT once.

In one test , with Windows XP SP3(!), it appeared to work without the slowdown.

I described this in the other thread (Slower with NCQ ?, Concurent access).

So if you did not yet, you should test with XP Service Pack 3. Maybe Microsoft fixed their scheduler there !?

Regards,

David

David,

I had missed that point! Could I ask, if you could please run the steelbytes NCQ test (in the same way - upto 9 instances - as 6 6 6 did) on the sp3 version.

It would add to the information pool when I talk to intel and WD.

thanks David,

LittleJhon.

Edited by LittleJhon

Share this post


Link to post
Share on other sites

David,

You wrote: "There was only one exception to the "AHCI is much slower" result: Windows XP Pro SP2, updated to SP3, then cloned to another partition and finally changed from legacy mode to AHCI (by: force install the intel driver, reboot, change setting in BIOS, boot Windows)

In this case, I did not see the problem."

You did not have solid data to quantify the issue. Besides, clone, partition, force intell drivers, etc.... God knows what happenned where... And you might have mistakenly loaded the programs from memory. So they might have appeared to start faster. Happens to all of us, we might have started them and later on forgot that we started them.

I believe i might have tested under XP SP3 and 2003 Server with latest SP at time with no difference either. I need to find or remember that XP iso i used... which service pack was slipstreamed, i just don't recall.

Share this post


Link to post
Share on other sites
And you might have mistakenly loaded the programs from memory. So they might have appeared to start faster.

No, I did every test with a fresh reboot (you can imagine how much time that took :angry:)

I will retest when I get home.

Share this post


Link to post
Share on other sites

Man I'm in a dilemma...

I bought two Samsung F1 750GB disks a few days ago just to discover one of them had bad sectors and would freeze my computer. The other one was fine but I just couldn't justify the risk of loosing all my data if it should start acting up.

I then looked towards the 7200.11 and found these 7200.11/NCQ wonder threads and got really excited since one of the problems with my older disks were performance during file transfer (8GB-10GB files)

I was almost ready to press the button an order two 750GB drives (not liking the 4 platters in 1TB drives) when I started reading some reviews on Newegg and was immediately discouraged. Click of death and faulty firmware on MANY disks. I just wouldn't be comfortable running clicking disks :(

Convince me that I should take the plunge and order two 7200.11 disks!

Share this post


Link to post
Share on other sites

Umm, nexus, yes that is true. Now i remember it had an irritating clicking sound frequently. But since I had my insulated/wrapped, i hear nothing from this drive. Yes, i guess if i kept hearing that click, it would drive me nuts. Apparently their faulty firmware makes it work with NCQ :)

A disk stress tested for a day prior to commiting any data on it i believe is safe.

Click of death and faulty firmware on MANY disks. I just wouldn't be comfortable running clicking disks :(

Convince me that I should take the plunge and order two 7200.11 disks!

Share this post


Link to post
Share on other sites

czr confirms that ST3640323AS 7200.11 640GB works well with NCQ.

I mentioned this but i will repeat again. These Seagate drives are just a noise-lover's jackpot! They are very loud (compared to Samsung's) and they do have that awful clicking sound frequently... like the click of death... and their seeks are very pronounced. You will need to find alternative ways to quiet these drives, be warned. I had mine wrapped with insulation material. Was too noisy.

Share this post


Link to post
Share on other sites

This WD does NOT support NCQ!!!!!!!!!!!!!!!!

(As reported by low-level disk ATA commands, manufacturer specifications on model number, intel storage manager and misc diagnostic utilities)

And yet, this drive exhibited performance under linux in level with all the other drives where we attributed the results to NCQ.

I am at a loss. This behaviour is not NCQ after all.

Okay, an old (4+ years i believe) 320GB WD Caviar:

Windows:

63 MB/s: 1 instance

28 MB/s: 2 instances (both at 0% position)

9 MB/s: 2 instances (0% position and 90% position)

15 MB/s: 10 instances (10% gap between each)

Linux:

63 MB/s: 1 instance

63 MB/s: 2 instances (both at 0% position)

45 MB/s: 2 instances (0% position and 90% position)

48 MB/s: 10 instances (10% gap between each)

Share this post


Link to post
Share on other sites
I was almost ready to press the button an order two 750GB drives (not liking the 4 platters in 1TB drives)

As far as I know, the ST31000340AS has 4 platters, while the ST31000333AS has only 3 platters! :)

So if you did not yet, you should test with XP Service Pack 3. Maybe Microsoft fixed their scheduler there !?

I had SP3 installed when testing.

SP3 doesn't help, at least not for me.

czr confirms that ST3640323AS 7200.11 640GB works well with NCQ.

Yay! :)

- I can fetch my two 7200.11 640GB from the post office tomorrow. :)

I mentioned this but i will repeat again. These Seagate drives are just a noise-lover's jackpot! They are very loud

Now that is shitty. :(

I hope they won't be too loud for me. :unsure:

But I don't mind seek noise. I only can't stand loud humming noises from the motor.

@Linux: perhaps Linux does something comparable to NCQ itself, like the 3ware RAID controller of my mate. That would explain everything.

Share this post


Link to post
Share on other sites

Hi,

I find somehow absurd that the tremendous advantage provided by a better NCQ support

for the seagates 7200.11 , doesn't appear when you read the incredible amount of reviews available on the web.

I tried to find the benchmark, that would be the most representative of this "NCQ advantage",

and the best one seems all "IOMeter tests".

So when I see theses "cryptical" graphs, it appears clearly that the seagate 7200.11 is a "winner" compared to the other 7200rpm drives:

http://techreport.com/articles.x/13440/10

However NO WAY, you would guess from theses graph that the seagate would outperform other drives to the extent described in this thread.

Also acording to all the graphes, raptor is at the top ,

while this post contradict this.

So I'm very confused because I couldn't find ANYWHERE else , a review, a benchmark , that would confirm what I read on this

thread. Moreover , the seagates have higher failure rates,

especially when you see the reviews at newegg for the ST3320613AS here.

Almost in every forum, when someone ask advice for the seagates, then someone answer something like this:

"you'd rather buy the WD 6400aaks or a samsung for the silence".

Now, I'm fed up, because in fact I'm considering a Velociraptor, and realize I may pay premium price,

and still miss what I'm the more expecting :

ability to launch multiple "hard drive intensive" stuff at the same time.

I'm not sure we could call this "multi tasking",

but let's for instance do a defragmentation in a partion,

and a scan disk in an other.

Doing this with my current samsung hd501LJ

would almost freeze the system, or make it incredibly slow .

Also, I've put the swap file in an other drive,

because it's clear than any concurrent access to my hd501LJ cripple overhaul performance of the system.

So I want a new drive, not necessarily with the best writing / reading speeds (80-90 mb/s is fine),

but that would allow to launch multiple "drive intensive" task at the same time, without

having the feeling to run a Pentium I.

So tell me which drive I should take ?

- The so much praised Velociraptor ,1x150gb platter, 16mb cache ? ( at 175 euros)

- The so much praised WD 6400 AKS , 2* 320 platters , 16mb cache ? (at 77 euros)

- The Seagate 7200.11 ST3500320AS , 2x250 platters, 32 mb cache ? (at 60 euros)

- The Seagate 7200.11 ST3320613AS , 1x320 platters , 16 mb cache ? (at 60 euros)

Note: The seagate with 2x320 platters is still not available in the shop, I plan to make my purchase.

After reading this thread, I'm just lost ! So, Please advise !

Thanks

Ps:

I've also considered SSD, but not enough space,

and the only affordable ones (core series) , have stuttering issues.

Edited by extrabigmehdi

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now