Beware: Seagate dropped the ball...OUCH!!!

Rednroll · 07-15-2008, 02:18 PM

Ok, I recently built my new DAW. I put TWO Seagate 500GB 7200.10 SATA-II hard drives inside the case. Today I was going to buy a 7200.11 Seagate 1TB hard drive to put in an external case and connect it via eSata. While I was shopping I noticed I could get the Seagate Desktop 7200.11 drive for the same price as a Seagate ES.2 hard drive. The ES.2 is their "enterprise" server series of hard drives, while the 7200.11 is their desktop consumer series. So, I started to wonder....what are the differences between the 7200.11 and the ES.2 hard drives?? Since, they're the same price...maybe I can get some benefits from the ES.2? So I downloaded each drive's data sheets from Seagate's website. Their specs where pretty much identical except the ES.2 had a much higher life expectancy. I did some further reading and found the ES.2 additionally was the the same hard drive but had tweaked firmware which should make the drive perform better with more reliability. So from a marketing stand point, it seemed like getting the ES.2 was the better deal.

So, I did some further googling to see how they actually compare on performance and found this excellent comparison review which compared the ES.2 and the 7200.11 to each other as well as other hard drives in their class. I was shocked to find out that the 7200.11 "desktop" version actually outperformed the usually more expensive ES.2 "server series" version of the drive in most every test.

Here's the shoot-out comparison review:
http://techreport.com/articles.x/13732/2

So I decided it looks like I would be better off to go with the 7200.11 over the ES.2 after reading this comparison.

Within that review, it mentioned some problems Seagate has been having with their Firmware....specifically in their 7200.10 Sata drives. Which of course peaked my curiosity, because I had already purchased (2) 7200.10 drives.

So it lead me to this Link, which goes into a lot of the details of the problem.
http://www.fluffles.net/articles/seagate-AAK-firmware

So the next thing I do is go and check the firmware on my current Seagate 7200.10 drives to see if they have the AAK problem firmware. Sure as shit!!! Both of my 7200.10 drives have the AAK firmware in them.

I also found that Seagate has the same problem in all their FreeAgent "Pro" external hard drives, where the eSATA interface connection is only a slight improvement over the USB 2.0 interace connection.
See Link Here: http://blog.noegruts.com/2007/12/sea...-problems.html

Seagate has not posted firmware updates for their drives, but have given some people updates via a private website and some people have reported that the drives performance has increased, but still fall short of other drives in their class, like the Western Digital offerings......but when you look at the posted specs by Seagate and compare them to the other drives by different manufactures....it actually shows that the 7200.11 series drives should be the best in their class.

Just an FYI. Maybe this can be useful to some of you. I know, I wish I had known about this before I bought my (2) 7200.10 drives with some f**ked up performance issues, when I thought I was purchasing some great drives for DAW use.

Rednroll · 07-15-2008, 02:31 PM

You can download "HD Tune" from here, which will allow you to run a bench mark on your drives, as well as get the Firmware version of your drives without having to open up the case.

http://www.hdtune.com/

Bulls Hit · 07-15-2008, 02:42 PM

Yeah that HD Tune reported much better numbers all round for my 320GB WD than my 320GB Seagate

Rednroll · 07-15-2008, 03:31 PM

If you have either a 500GB or 320GB Seagate 7200.10 Hard drive with the AAK firmware version, then I was able to track down and download the firmware updates for both, which Seagate has not posted publicly. It seems they have given some people an AAM firmware update, where most people are reporting that their average throughput has increased 10-20MB/sec from the previous AAK version.

The catch is that your drive has to have a P/N that ends in "-308". So example, my part number on one of my drives is 98J146-308, which means that it can be updated with the AAM firmware. You have to read the part number off the label on your Hard drive. Lucky for me, I wrote down all the part numbers and serial numbers of my drives before I put them in my case. There is a different firmware for the 320GB version and the 500GB version. I have both, so if you need it....then I will post it for you.

Rednroll · 07-15-2008, 03:48 PM

Ok, here's a before and after measurement someone else posted.
I'm going to run some tests on my HDs and see how they do and then I'll venture into upgrading the firmware.

As you can see, he went from a 53MB/Sec sustained transfer rate to 77MB/Sec....which is more like the actual expected specs of the drive.

Rednroll · 07-15-2008, 03:59 PM

This one is scary also. Here was a test that someone ran which compared the USB 2.0 vs. the eSATA connection on the Seagate FreeAgent Pro Hard drives.

USB 2.0

eSATA

Yikes!!!! 32.5 MB/s vs 39.5 MB/S average read for the eSata connection?

BaronStinky · 07-16-2008, 12:57 AM

Heh. I agonized over spending the extra cash to get esata capability. In the end I bought the regular free agent. I'm not exactly thrilled with the speed, but it's what you'd expect from USB 2 and apparently not much worse than I would have gotten for the increased cost of the "pro". I'm learning to avoid that word more and more often...

bullshark · 07-16-2008, 02:37 AM

just tested my eSATA backup drive, a Seagate 320 GBs (with AAK firmware) is in the Vantec case at this time, and this is what I got:

Nothing to write home about, but still way better than my best WD ATA internal drive:

I'll test the two Seagate 7200.11 500GBs SATA next, once my new computer is done.

hourglass · 07-16-2008, 04:07 AM

i'll be extremely interested to see the results of the 7200.11 series... when i was looking at a desktop instead of a laptop the 500GB model was on my shopping list.

Rednroll · 07-16-2008, 09:36 AM

Quote:

Originally Posted by hourglass

i'll be extremely interested to see the results of the 7200.11 series... when i was looking at a desktop instead of a laptop the 500GB model was on my shopping list.

Yep, I did some googling yesterday on the 7200.11 series, since that's what I was going to purchase. It seems they had a similar problem, but it seems there is a publicly available firmware release available for that series, which fixes a lack of performance problem. So I may go for the 7200.11 1TB drive yet.

Read more here:
http://www.driverheaven.net/hardware...must-read.html

bluzkat · 07-16-2008, 11:51 AM

Here are a couple very good 'hardware' tech sites...

http://www.tomshardware.com/us/#redir

http://www.anandtech.com/

Hope you find these helpful.

ted01 · 07-16-2008, 12:43 PM

Let's keep in mind that at 16/44.1 you generate about 5 MB per minute of data per track. So, that means that if your drive can do 30 MB per second, you could theoretically record/playback 360 tracks. Not bad...

Rednroll · 07-16-2008, 12:53 PM

Quote:

Originally Posted by bullshark

just tested my eSATA backup drive, a Seagate 320 GBs (with AAK firmware) is in the Vantec case at this time, and this is what I got:

That's interesting because your drive with the AAK firmware seems to do what it is suppose to do. I think this may have to do with one of those Chinese manufacturing mistakes. That's a rumor I heard anyways, that the drives that have seen the problem have been built in China and the ones that work properly are built in Taiwan. I highly believe that rumor, since I have a lot of personal experience with Chinese manufacturing and some of the sh*t they do.

Bullshark,
Let me know if you want me to post the Firmware for the 320GB drive for you, or maybe you've been able to track it down yourself. One thing I noticed from other people who have done the Firmware update which is similar in your results is that the AAK firmware seems to always show a 3-4% CPU usage hit like yours. Then after the AAM firmware update that usually goes down to 0% CPU.

I still need to run my tests on the 2 AAK firmware drives that I have. I'll post the results once I have a chance to run the test.

fester2000 · 07-16-2008, 02:02 PM

Thanks for starting this thread. It made me discover (1) there's something really weird about my hard drive performance -- I had lots of negative spikes down below 10mb/s, though the average was well up in the 40mb/s range, and (2) the seagate external firewire drive I bought six months ago and set up to auto-backup hasn't done sh*t since March. So I did an old school backup of the main drive.

Cheers,

Fester2k

bullshark · 07-16-2008, 02:19 PM

Quote:

Originally Posted by Rednroll

Bullshark,
Let me know if you want me to post the Firmware for the 320GB drive for you, or maybe you've been able to track it down yourself. One thing I noticed from other people who have done the Firmware update which is similar in your results is that the AAK firmware seems to always show a 3-4% CPU usage hit like yours. Then after the AAM firmware update that usually goes down to 0% CPU.

Well, this is one of the drive I use in my backup enclosure so I'm not too concerned with it's performance, usually backups are done overnight by Cobian, which is even nice enough to shut down the machine when it's done, so I'd rather not try and mess with the firmware. Beside, it's already twice as fast as what the best USB2 drive have to offer...

My two 7200.11 are a different story, I just bought them for my new machine, and those are better be fast or there will be hell to pay. I don't mind trying firmware on those if they aren't as fast as they're supposed to be since there won't be any data to loose should the firmware thingy fail. But I'll do a quick test first once the machine is built (should be this week-end).

Thank's.

synth · 07-16-2008, 03:51 PM

Just get an old skool UDMA-6 (ATA) internal drive.No joking here.A well-made single platter ATA drive with 8MB cache beats the shit out of a 'badass' Serial ATA 2 drive with 32MB cache.It's not all in the cache or interface.
Old ATA drives are still the best performing drives overall.They're a bit more expensive than their Serial ATA 300 'brothers' for a reason.It's funny to see how people avoid them and go for the Serial ATA ones.LOL

Better stick to old,proven technology.

bullshark · 07-16-2008, 04:05 PM

Quote:

Originally Posted by synth

Just get an old skool UDMA-6 (ATA) internal drive.No joking here.A well-made single platter ATA drive with 8MB cache beats the shit out of a 'badass' Serial ATA 2 drive with 32MB cache.It's not all in the cache or interface.
Old ATA drives are still the best performing drives overall.They're a bit more expensive than their Serial ATA 300 'brothers' for a reason.It's funny to see how people avoid them and go for the Serial ATA ones.LOL

Better stick to old,proven technology.

Still, the test I've done, which is posted above, says otherwise. And that's the best out of five older ATA drive I've tested against a single (and reportedly not that fast either by today's standard) SATA drive. Now, there may be circumstances where what you say could be true, but if there is, I'd like to know what those are since testing a drive for speed seems pretty straightforward to me.

synth · 07-16-2008, 04:13 PM

Quote:

Originally Posted by bullshark

Still, the test I've done, which is posted above, says otherwise. And that's the best out of five older ATA drive I've tested against a single (and reportedly not that fast either by today's standard) SATA drive. Now, there may be circumstances where what you say could be true, but if there is, I'd like to know what those are since testing a drive for speed seems pretty straightforward to me.

I'll post some of my results then

synth · 07-16-2008, 04:16 PM

Quote:

Originally Posted by bullshark

just tested my eSATA backup drive, a Seagate 320 GBs (with AAK firmware) is in the Vantec case at this time, and this is what I got:

Nothing to write home about, but still way better than my best WD ATA internal drive:

I'll test the two Seagate 7200.11 500GBs SATA next, once my new computer is done.

52*C?! Wow,that's really hot!

bullshark · 07-16-2008, 04:23 PM

Yeah, don't know how accurate those readings are. And the internal started at around 40 before the test and ended at 52 after, guess the test work those drives pretty hard.

synth · 07-16-2008, 06:23 PM

Let's have some fun first.Here are some performance graphs of a POS (by today's standards) 6-yr old ATA drive running inside a crappy 6yr-old 2GHz single core box.It has a 'whopping' 1.7MB of cache,LOL

Impressive,isn't it?
Blazingly fast burst rate,quick access times,very low temps and not so shabby read speed averages make this one a winner. Much faster than that eSerialPenis Seagate drive

In the next post,I'll show you what a modern ATA drive is capable of.

Rednroll · 07-16-2008, 08:27 PM

Well, surprisingly enough I didn't seem to see what the others have seen with my 2 7200.10 drives with the AAK firmware.

Drive1: (AAK Firmware)

Drive2: (AAK Firmware)

Bulls Hit · 07-16-2008, 08:49 PM

WD is quite a bit faster on the face of it.

Mind you the Seagate results are better than yesterday's after I did a defrag

synth · 07-17-2008, 07:15 AM

Quote:

Originally Posted by Rednroll

Well, surprisingly enough I didn't seem to see what the others have seen with my 2 7200.10 drives with the AAK firmware.

Drive1: (AAK Firmware)

Drive2: (AAK Firmware)

Now that's what I expect to see from a SerialATA 300 drive
(at least)

Not bad for a 500GB drive.
The temps are great

synth · 07-17-2008, 08:22 AM

[EDIT]

[EDIT]

bullshark · 07-18-2008, 12:11 AM

And here's the result for the 7200.11 using firmware SD15:

Rednroll · 07-29-2008, 10:22 PM

Well, I think Seagate might have gotten a unfounded bad wrap for this one. I'm leaning towards that there might have been a problem with the v2.53 of HD Tune where everyone was reporting the bad results on the 7200.10 drives with the AAK firmware.

Here's my A/B test results of AAK vs. AAM firmware. A slight cap seemed to be lifted and I gained about 4Meg/Sec. Both drives had pretty much identical results.

I should be getting my 7200.11 1TB drive with SD15 firmware in the mail tomorrow. I'll run that test tomorrow to see how it compares.

Disk Info Before/After:

AAK vs. AAM Read (64K Block Size) Bench mark comparison

AAK vs. AAM Read/Write Bench mark Test

Rednroll · 07-30-2008, 05:25 PM

Here's how my 1TB Seagate 7200.11 drive stacks up connected via eSATA.

BenchMark (Read) 64K Block size

Rednroll · 07-30-2008, 05:40 PM

Hmmmm??? I found something interesting. For some reason HD Tune is not showing that this 7200.11 drive has a Buffer.

Bullshark....can if you're still reading this thread can you check your 7200.11 drive with the SD15 firmware if HD Tune is showing your correct 32MB buffer size? It seems this might be another Seagate firmware problem.

Here's what mine is showing....notice no buffer size in the bottom left.

bullshark · 07-30-2008, 05:53 PM

It did show the correct buffer, but I was using a different (older, and non-pro) version of HDtune. Probably a quirk of this version.

Rednroll · 07-30-2008, 06:59 PM

Quote:

Originally Posted by bullshark

It did show the correct buffer, but I was using a different (older, and non-pro) version of HDtune. Probably a quirk of this version.

Hmmmm??? I just downloaded another program called AIDA32 and it reports an unknown buffer size also, while both of my 7200.10 drives both report 16MB. This sh*t is starting to drive me nutz.

bullshark · 07-30-2008, 07:50 PM

Just had a quick look, it now report n/a just like your's. I swear it did report the correct 32 MB buffer before, but I since changed motherboard and also protocol from IDE to AHCI, maybe that's the cause?

In any case, those drive smoke so I wouldn't worry about it. I just did a complete ghost image of my boot drive in under 2 minutes while it would take 5-6 minutes with my old drives; the only drawback I can see with those drives is that I don't have time to make coffee anymore when ghosting or moving large amount of data between them.

Rednroll · 07-31-2008, 06:44 AM

Quote:

Originally Posted by bullshark

Just had a quick look, it now report n/a just like your's. I swear it did report the correct 32 MB buffer before, but I since changed motherboard and also protocol from IDE to AHCI, maybe that's the cause?

In any case, those drive smoke so I wouldn't worry about it. I just did a complete ghost image of my boot drive in under 2 minutes while it would take 5-6 minutes with my old drives; the only drawback I can see with those drives is that I don't have time to make coffee anymore when ghosting or moving large amount of data between them.

Thanks....it's at least good to know that you're seeing something similar. I just had a few concerns. My 1st concern was that this could be a Seagate problem again with the firmware. I had read that SD04 and SD14 firmware 7200.11 drives where not reporting the correct buffer sizes and Seagate released a firmware update for those drives....but nothing was ever mentioned about SD15 firmware drives. My next concern was that I had just bought this drive off of someone on Ebay and I felt that maybe I was sold a defective drive, where 99% of the people would never notice that the Buffer was missing. My last concern was that I saw that my drive had a slight lower performance than yours, which made me believe a possible reason I had a slight lower performance was because my drive was not reporting the 32MB buffer.

So I'm not feeling fully comfortable yet....I would feel 100% better if I knew that this drive was correctly reporting to my bios and OS the 32MB buffer.....afterall, one of the main attractions for buying this drive was the larger buffer size and if the drive is not reporting it correctly, then that means it won't be utilized by your system.

I'm going to take this drive out of the external enclosure and put it inside my case to elliminate the enclosure as a possible source of the problem, and then I'm also going to connect with the USB connection of the case to see if I see different results.

Then I'm going to go cry to Seagate's tech support if all else fails.

earlabs · 07-31-2008, 06:55 AM

I'd like to ask a few questions, that may be rather stupid to most of you guys battling over this.

What problems did you encounter that made you do these tests?
How big are the projects / files that you use that made you do these tests?
How many hours did you spend on these tests and what did you gain from them? Is that any use to people like myself? I have not had any problems in recent years with hdd/s.

Rednroll · 07-31-2008, 08:21 AM

Quote:

Originally Posted by earlabs

I'd like to ask a few questions, that may be rather stupid to most of you guys battling over this.

What problems did you encounter that made you do these tests?
How big are the projects / files that you use that made you do these tests?
How many hours did you spend on these tests and what did you gain from them? Is that any use to people like myself? I have not had any problems in recent years with hdd/s.

My system is relatively new. Alls these tests are doing are just making sure we got what we paid for. If you read my 1st post it pretty much outlines how this all came about. I've been using Seagate drives for many years because they have had great performance and reliability. I've only had one Seagate drive ever fail on me in about the 20 that I own. Even with the drive that failed Seagate would have exchanged it for me.....but the drive was a 60 Gig EIDE drive and it was pretty much outdated, so I didn't even bother to send it in for an exchange.

The point is that we try to buy a great performance drive for our audio work, and we just want to make sure the drive is living up to it's posted specifications. In my current case, I paid extra for a drive that currently has the largest available buffer available....but the drive does not seem to be telling my system that it has a 32MB buffer, so my system will not take advantage of the buffer and the drive will have reduced performance.....better performance means a faster DAW overall.....and who doesn't want to get the best out of their DAW?

A DAW is constantly reading and writing data to a hard drive. The speed of the hard drive is one of the primary bottle necks of your DAW's performance.

If you want to experience exactly what I'm talking about then do this on your system.

GOTO CONTROL PANEL>SYSTEM>HARDWARE>DEVICE MANAGER

Click on the IDE ATA/ATAPI CONTROLLERS icon to expand it

Right click on each of the listed controllers and select properties. Goto the Advanced settings tab and in the TRANSFER MODE selection box choose "PIO ONLY" and click OK.

Now Goto SCSI AND RAID CONTROLLERS and do the same thing for all your listed SATA controllers.

At this point you just reduced the transfer speeds of all your hard drives. Now go and record some tracks in Reaper and see how yoour performance compares. The only thing you changed was the transfer speeds of your hsard drives but now your PC will be 100 times slower.......so transfer speeds of your HDD is pretty important especially while running a DAW......so you better bet we want to get the fastest transfer speed available.

On one of my older DAWs I had a 700Mhz Athlon processor. I was able to record or playback up to 60 Tracks at a time, while others with faster systems where coming in around 30-40 Tracks. The difference was that I had a Seagate 10K RPM SCSI hard drive compared to their 5400 or 7200 RPM IDE drives. So yes...having a fast transfer rate drive is a major performance boost for a DAW.

So what could you gain from this knowledge? Well, you could probably save yourself a lot of money in the long run by realizing you don't always need to buy a new and faster PC because you're having performance issues while running your DAW. Sometimes alls it takes is a faster hard drive to get near double the performance.

bullshark · 07-31-2008, 08:29 AM

Quote:

Originally Posted by earlabs

I'd like to ask a few questions, that may be rather stupid to most of you guys battling over this.

What problems did you encounter that made you do these tests?
How big are the projects / files that you use that made you do these tests?
How many hours did you spend on these tests and what did you gain from them? Is that any use to people like myself? I have not had any problems in recent years with hdd/s.

The HD test itself takes a whole of 5 minutes, if that, to do. Can't speak for others, but in my case it is just one of a series of stress test I subject the new machine I just built to.

Why? Well, considering it's a new built with new parts, I want to make sure everything is working as it should before releasing the machine into service, especially since now I can exchange components if needed; a month down the road it will be too late for that.

The whole routine of testing takes me about 2 hours to do. This time it took longer because I discovered a faulty motherboard in the process and had to change it; better now than later thought. Considering I've had the last machine I tested this way in use for almost 8 trooblefree years, I think it's time well spent.

Rednroll · 07-31-2008, 11:26 AM

Quote:

Originally Posted by bullshark

The HD test itself takes a whole of 5 minutes, if that, to do. Can't speak for others, but in my case it is just one of a series of stress test I subject the new machine I just built to.

Why? Well, considering it's a new built with new parts, I want to make sure everything is working as it should before releasing the machine into service, especially since now I can exchange components if needed; a month down the road it will be too late for that.

The whole routine of testing takes me about 2 hours to do. This time it took longer because I discovered a faulty motherboard in the process and had to change it; better now than later thought. Considering I've had the last machine I tested this way in use for almost 8 trooblefree years, I think it's time well spent.

Exactly!!! My last DAW lasted me 7 years, while I hear a lot of people saying you have to upgrade your PC every 2 years. Well, if you don't know how to maintain your PC and make sure it's working properly when you 1st build it....then you probably need to buy a new PC every 2 years. I usually plan on getting at least 5 years out of my DAWs when I build them.

After I make sure everything is running and functioning as it should, then I start running the real world stress tests like seeing how many audio tracks can be played back simultaneously before it spudders. How many simultaneous tracks I can record before I see a dropped sample. How many simultaneous FXs I can run. Just things so I know the limitations of my DAW before doing real work.

For me, it's actually pretty fun stuff when you build your own DAW, because you are rewarded with a sense of accomplishment once you see what the system you just built is capable of doing. It also gives you a sense of when things are going wrong later in the life of your DAW, when you start noticing it can no longer do what it did when you 1st finished building it.

earlabs · 07-31-2008, 02:57 PM

Quote:

Originally Posted by bullshark

The HD test itself takes a whole of 5 minutes, if that, to do. Can't speak for others, but in my case it is just one of a series of stress test I subject the new machine I just built to.

Why? Well, considering it's a new built with new parts, I want to make sure everything is working as it should before releasing the machine into service, especially since now I can exchange components if needed; a month down the road it will be too late for that.

The whole routine of testing takes me about 2 hours to do. This time it took longer because I discovered a faulty motherboard in the process and had to change it; better now than later thought. Considering I've had the last machine I tested this way in use for almost 8 trooblefree years, I think it's time well spent.

got it... I just thought it'd be a nerdfreak thing. But perhaps you're right. You'd want to know if you sold what was promised to you.

Rednroll · 07-31-2008, 03:12 PM

Quote:

Originally Posted by earlabs

got it... I just thought it'd be a nerdfreak thing.

You mean kind of like buying a Reaper T-shirt and then spending time to take your picture wearing the shirt and uploading it as your Avatar? Yeah, it's kind of like that, but a little more useful.

Rednroll · 08-01-2008, 08:24 AM

I ran the test on the 7200.11 drive with the USB connection. When you run a test like this, it makes me really wonder how they ever come up with that spec. that says USB has a max data rate of 480MB/sec? It's pretty obvious that it's pretty much capped at 33MB/sec and this was the only device I had connected on the USB bus. This same drive with the SATA connection did an average of 90MB/Sec....you would think with a bus speed of up to 480MB/Sec, that the 90MB/Sec would be no problem??? It really makes you question these numbers they post about a bus speed and how they ever come up with them?

7200.11 with USB 2.0 interface connection