Seagate 7200.10 AAK firmware performance issues!

If you have a Seagate 7200.10 disk, what firmware?

  • EEks! I have one of those AAK disks!

    Votes: 26 52.0%
  • I have an AAE/AAC disk <:)

    Votes: 15 30.0%
  • I have a 7200.10 but with other firmware

    Votes: 3 6.0%
  • I have a 7200.10 but don't know the firmware

    Votes: 6 12.0%

  • Total voters
    50

enlightenment

Distinguished
Mar 9, 2007
111
0
18,680
Hello readers!

I'm writing an article about Seagate's 7200.10 disks with AAK-firmware. Apparently, this firmware has much lower sequential throughput than other firmware like AAE and AAC. I can even say some panic has arisen from the fact that people had one of those AAK disks and were unhappy with its performance. The number of reports on the internet is growing and I'm investigating their claims.

My question to you, readers, is if you have an AAK drive or other firmware? For me it is important to know how widespread the problem is. Only vote if you have a Seagate 7200.10 drive!

Also, i would to ask owners of said disk the following questions:
- Do you have any problems or concerns about the performance of your drive?
- Have you run benchmarks after having bought the drive?
- If so, have you noticed any degraded performance?


Thank you for reading. I will update the topic when i have more information. As of the moment of writing i can confirm the performance issues with AAK-drives! AAK-disks appear to be quite slower than AAE disks. I am currently doing exhaustive and thorough benchmarks and integrating that into an article. I've also written to Seagate asking to respond to the issues.

Some people claim Seagate had offered to take back disks with AAK-firmware even though the disks were error-less. This could mean that they recognize the performance problems with the disk. One user also said he got a reply from Seagate saying that the disks were crafted for a specific OEM customer and that it was not possible to update the firmware of AAK-disks to AAE or other, because of physical difference between the two drives.

Oops! 🙁
 
I have a home server I set up with a RAID 5 array on a 3Ware 9650SE. I currently have 3 Seagate 7200.10 750GB drives, one drive has AAK firmware, the other 2 have AAE firmware.

I have not noticed any performance degradation of the array specifically because I always access the server over the network, and Gigabit Ethernet transfers at only about 30 MB/sec. Local testing with HD Tach shows the array doing almost exactly 90MB/sec reads across the entire array, which seems low to me, as the controller is a PCIe x4 controller in a PCIe x16 slot, on a 945G chipset. I should expect mid to high 100's of MB/sec in this setup.

If you can confirm that Seagate will trade out AAK firmware drives for AAE firmware drives, I'd like to switch mine out and see if performance improves.
 
Hm well performance in RAIDs is trickier, since a lot more factors are involved. But yeah you could see performance improvements with AAE-only. But if you keep using gigabit then it will not matter in practice.

Yet if people bought a 'next generation perpendicular Seagate disk' expecting it to perform admirably, they might feel taken by Seagate. They will get a disk that does not perform as expected in numerous benchmarks over the net. Especially the drop in STR is inexcusable - at least that's what i think.
 
The noises might just be calibrations and could be perfectly normal. You can view your current firmware version:

- on the disk itself, it will say 3.AAE or 3.AAK
- during the POST/BIOS possibly
- with a hardware identify application, like Everest or SiSoftware Sandra and probably a lot more utilities
 
Speedfan will tell you what firmware it uses. Click on the SMART tab, then look just a bit down from the top on the right side. Mine says Firmware:3.AAC. I've never benched this harddrive, it works fine as far as I'm concerned.
 
Question: is the loss in sequential throughput a tradeoff for improved access times, or are those about the same or worse than other firmware revisions? Because I would gladly trade faster access times for less throughput... at least based on my experience with various drives and RAID setups.

-mcg
 
Thanx guys. Turns out I do have the 3.AAK firmware.

Not knowing squat about hard drive performance I couldn't tell you if I'm having any performance issues. It's also my first SATA drive so I can't compare it to anything else. Any programs I can run to test what effect it's having?
 
In XP, simply go to device manager, double-click on your drive to get to the properties page, then the details tab. Under 'device instance id', look after the first long set of underscores and you will see your firmware rev.
 
I have two 400GB 7200.10 rev 3.AAC drives. Burst speed is ~95MB/s and average speed is ~65MB/s. random access time is ~15.8ms

I think this is normal, right?


 
i have 4x 7200.10 drives with the aae firmware, in a raid 0 configuration, my performance is astounding and is never a problem, i don't generally run benchmarks on it myself but the "feel" of the speed is there (bootup, games loading ect)
 
You can see the firmware version from device manager (to bring that up, run devmgmt.msc). Next bring up the property page for the disk drive. From the details tab, select "Device Instance Path". There will be a bunch of underscores then the firmware version. Mine is 3.AAE.

Two questions: (1) is there a small disk drive benchmark program downloadable somewhere? and (2) how do you know if NCQ is available/actually being used? There is a bit of uncertainty whether this drive actually has NCQ.
 
An example HDtune screenshot of an AAK disk:
Seagate_320_AAK.jpg


As you can see, the disk does not perform as expected. It more or less has a straight line (horizontal) and only begins to fall from 80% of the capacity. It almost seems like there is a 'cap' on the disk.

A normal score for the 7200.10 disk is:
Seagate_320_AAE.jpg


Here it begins to slide nicely down, as it should. The AAE disk appears to match the speed of AAK at 80% of the capacity - where the speed drops below 58MB/s.

I first was inclined to believe the HDTune were giving wrong impressions because of lack of read-ahead which the AAK disk may have needed. Yet even with maximum read-ahead i could not move this disk faster. I did real-world copy tests like extracting a WinRAR archive with lot's of small files - not exactly something that needs high sequential performance. That was half a second slower too. The AAK needed 4:02 minutes while the AAE disk needed only 3:35. These benchmarks were run at least 5 times - i did very thorough testing.

There are some cases where the AAK disk manages to get a small lead over the AAE drive, but the occasions are few. Overall i could say the AAK disk appears 15-20% slower than AAE. To me the questions is: WHY? Why would Seagate cripple their own cash cow? It doesn't make any sense to me. Yet i feel customers now buy disks believing it has the performance level they read in the reviews, while actually they get the performance level of a generation backwards. Not really what you paid for.

That said, there appears to be no reliability issues with the disk -- no bad sectors or any sign of reduced lifespan. They are also most likely not rebadged Maxtor disks, since they both feature the exact same size in sectors -- unless Seagate has gone to great lengths trying to obfuscate their operation. I'm still unsure why AAK even exists, and why it's so much slower.
 
Reading your posts it strikes me that maybe the AAK is not actually the same design or possibly not even a Seagate drive. It could be from another disk drive maker with Seagate's name stamped on it or a new design with a different number of platters or something of that nature. That would explain why the firmware is not interchangeable and the performance graphs are so different. Take a look at the two drives physically next to each other; do they have the same identical enclosure & circuit board layout? That could provide a clue.
 
I also thought of that. The disks physically look the same, though i will do another careful inspection. But the sector count is exactly the same. If they are different disks, they should be different. Seagate might have 'capped' the AAK disk to the sector count of AAE disks - that is a possibility.

But all mechanical harddisks should have a falling graph in HDTune - the AAK one just looks capped or crippled. It just does not make any sense to me. The most probable explanation is that AAK is made for a specific purpose - such as databases or any of that. But i haven't been able to find the type of I/O it's really good at. It's more or less.. wanting.
 
I have one of each, and I have confirmed (before ever seeing this post) that they are different. I read about it on Storagereview awhile back.

I have an AAE drive purchased OCT/11/2006, and an AAK purchased JUL/2/2007. The AAE is my boot drive, AAK is a storage drive. Both are SATA drives. Neither has any SMART errors, and they never have. No bad sectors either.

I have had no errors or problems with either drive. I have made no attempt to contact Seagate or to exchange the AAK, as I don't really want to be without its capacity for the time a return would take. As long as there are no problems, I will probably just keep it.

Here are my HDTune results, using the program's default settings.

7200.10AAEfirmwareHDTune.gif

7200.10AAKfirmwareHDTune.gif


 
Thanks for this post. I'm having much better results in my RAID 0 setup after replacing the AAK drives I was using. The low performance was really bugging me and I couldn't figure out what was going on! More details at http://www.tomshardware.com/forum/243893-32-raid-performance-seagate-firmware

Before: RAID 0 w/ 2x Seagate 320GB Barracuda, AAK firmware (ST3320620AS)

(flat graph, average read 86.4 MB/s)

After: RAID 0 w/ 2x Western Digital 500GB RE2 RAID Edition (WD5000YS)

(Normal curved graph, average read 130.5 MB/s)
 
I have two Seagate 500GB 7200.10's with the AAK firmware bought 2 months apart from one another. One I got OEM from newegg and the other from fry's in the older "big" retail box and both were made in thailand. I was curious after reading about this whole firmware issue so I ran some test with a biostar tforce965PT board using the ICH8 controller in sataII "jumpers removed from drives".

I dont know why but I am not seeing this "capped" effect.

HDTune_Benchmark_ST3500630AS-1.png
HDTune_Benchmark_ST3500630AS-2.png


hdtach-1.JPG
hdtach-2.JPG
 
Hrm.. What if the CPU was the bottleneck, then would the hard drives still affect the performance? I'm thinking about a RAID 5. Some good quality used 750gB HDDs would be good. Provided cost + shipping was a good deal as opposed to new drives.

Another question. Is this firmware on all sized drives in the 7200.10 family or only certain sizes?
 
I have two 7200.10s. One with AAE bought December 2006, another AAK bought today. I'm rebuilding my RAID 1 (linux md) right now.

I didn't notice the throughput cap. In fact the resync speed resembles the normal graphs shown above - stared at around 84000K/sec and slowly fell to around 65000K/sec (50% now as I'm typing this on that very box ;-))
 
notice that in the HD Tune graphs all STR-challenged AAK access time samples are capped under ~20ms - could this be tuned for some realtime application or something?

I have (1) non.AAK and (1) AAK, each 500g, but so far too lazy to connect the Thai AAK one. I am over-confident that the AAK one is differently abled - because I bought it from NewEgg, AND I eat my vegetables. That's clearly overwhelming proof! Besides, who's ever heard of a 'chinese fortune hard disk?' I'm thinking about just leaving the AAK sealed in its bag, so it's crunki-badness can't leak into my only good hard drive.

 
ok heres mine one drive drive (the aac) is from newegg and the other (aak) is from wally world.

aak
HDTune_Benchmark_ST3320620AS.png


aac
HDTune_Benchmark_ST3320620AS1.png


So with this my aak is my boot drive with 2 partitons and i've never noticed any issues.

oh i'm running sata 150 with these drives
 
I currently am running a Seagate 7200.10 400GB model #ST3400620AS with firmware 3.AAE. I have the drive set up with Win. Vista Home Premium 32bit. For the most part, the drive seems to perform well, although, I often hear it running alot because of the indexing services in Vista or Windows Defender running when I first boot up. This does slow performance a bit. I did notice a feature for disk drives in Vista for improved performance. To find this go to Device Manger, click on your disk drive, then go under Policies. There you will find a check box under the Enable write caching on the disk, check this box. Also under the aformentioned check box is the Enable advanced performance check box, check this box as well. This second box was not enabled when I loaded Vista, and it is recomended that you have power redundancy in case of power failure, or you could lose data. Good Luck