Jump to content
Sign in to follow this  
NBouc

320 GB SATA2 drive slower than a 160 GB SATA2 drive ?

Recommended Posts

Hi, I am puzzled by the poor performance of my new SATA2 drive compared to my previous ones (that I still have and use)that is affecting FSX performance.The new drive is a Seagate Barracuda ST3320620AS (SATA2) 320 GB 16MB buffer, partitionned into C (60GB) and D (238 GB). FSX is installed on the D drive.My previous drives, both WD1600JS SATA2 7200 RPM 8MB buffer, now holds data on a single 160 GB partition each (Drive F and G).Since FSX was running with no stutter on the previous system (P4 3.6GHz CPU and WD1600JS 160 GB), I was a bit surprised to see the introduction of stutters on the system (C2D 8400 at 3GHz with a Seagate Barracuda ST3320620AS). Since my old WD drives are now installed on the new system as additional drives, I compared the two drive model with a very basic file copy and saw a big difference. Copying 200 MB of files (150 files) on the WD1600JS (from and to that same drive/partition)took less than 10 seconds, but the same files copied from the WD1600Js to the Seagate drive took up to 30 seconds. Those same files copied from and to the Seagate drive took up to 30 seconds to complete. How could that be ?I am running Windows XP SP2, on an Asus P5E board (X38) with the SATA drives as Standard IDE, connected as follows:- SATA1: Seagate 320 GB- SATA2: WD1600JS 160 GB- SATA3: WD1600JS 160 GB- SATA4: empty- SATA5: DVD Drive (LG GH20NS10 SATA)- SATA6: emptyThe average latency for both models are almost the same (4.20 vs 4.16 ms). Is it normal for the Seagate drive to take 3 times longer for the same simple file copy operation ?Thank's in advance for any advice or explanation one may have.


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites
Guest baksteen33

Hi Nbouc, Was the old drive also partitioned? Assuming the OS is on C, D would be the worst place to have FSX to start, particularly because it's the same drive. It's the first successful step to annihilate HD performance sorry to say. Other than that, is everything running in the right modes, drivers, controllers OK? The Seagate should be about just as quick. Good luck! Kind regards Jaap

Share this post


Link to post
Share on other sites

Thank's JAAP for your input,The OS is effectively on C. The old WD drive, when it was on my previous system, was also partitionned as C and D (OS on C, FSX on D) and it ran faster than the Seagate on my current system (and is faster that the Seagate on my new system too).The system itself seems to be OK. In Windows XP Device Manager, I see the following under IDE/ATAPI Controller (I have 6 SATA connectors on the Motherboard - set as Standard IDE):- Intel® ICH9 2 port Serial ATA Storage Controller 2 - 2926- Intel® ICH9 4 port Serial ATA Storage Controller 1 - 2920- Primary IDE Channel- Primary IDE Channel- Primary IDE Channel- Secondary IDE Channel- Secondary IDE Channel- Secondary IDE Channel- Standard Dual Channel PCI IDE ControllerAll devices are working properly according to Device Manager.Are you saying that FS should run better if it was on a separate physical drive than the OS ?I've always installed FS (since FS98) on the second partition of the OS Drive (OS on C, FS on D) with good results. When I have the time, I will try and install FSX on the WD drive (currently drive F) to see if there is any improvement with FSX. It is a separate physical drive, but that will certainly not make the Seagate operate faster :-/Best regards,


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites

Hi, Nbouc.We will need a lot more info on the setup to give you better feedback, like is it a new installation of FS on the new drives, or is it copied from the old drives? Are you running the same Fsconfig file, the same background Tasks etc.?In the mean time you can try to Benchmark your drives by using HD Tach, or HD Tune, both are free utils, to compare just the Drive transfer rate without other things running that may affect performance. Also post your Logical drive config. From My Computer, A,B C, D size etc.Many things can affect IO transfer, but keep your Partitioning, that is a good thing, and if you plan it right, select a size for OS only around 10GB you can easily make backups and restores, if your OS / C: Logical drive goes bad you just loose your OS and so installed programs, not all your other Data, assuming you save it on D:.You may want to look at Your BIOS and see if your drives are configured properly? The most likely cause is a BIOS or some hardware configuration problem. Post here again after you run your Benchmarks. TV

Share this post


Link to post
Share on other sites

Hi Avcomware,I'll get HD Tach & HD Tune and post back with more specific system description after having run the benchmark.Much appreciated.


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites

Hi Avcomware,Here's more info on my system and the HD Tach benchmark results for the SATA drives.System description:- Asus P5E Motherboard (X38 Chipset ICH9R) BIOS 0601- Intel C2D E8400 (3.0GHz 1333 FSB)- 4GB Crucial Ballistix DDR2-1066 (running at DDR2 800 4-4-4-12 for stability)- NVidia GeForce 8800GTX 768MB PCIe x16 (Asus)- 3 SATA2 drives (300 MB/sec) + 1 SATA DVD Burner- Windows XP Professional SP2 (32-bit)This is a brand new system that came with brand new Seagate drive. I partitionned and formated the Seagate SATA drive myself (using the Windows XP Pro Install CD) and did a fresh install of Windows XP Pro SP2 in the following sequence:- Made sure that SATA was set as IDE in the BIOS.- Partition and format Seagate 320 GB Drive as C (60GB) and D (238GB, the remaining available space).- Installed XP Pro SP2 on C: drive- Installed the latest Intel INF- Installed all Motherboard Drivers from Asus CD (Sound, LAN, etc...)- Installed required drivers for additional hardware (Latest NVidia, DX9.C)After all was done, I then connected the two remaining Western Digital SATA drives that I use for Data storage (ported over from my previous system). FSX is a fresh install (FSX + Accelleration) on the D: drive. I did not copy any CFG file from my previous system. I like it clean. The computer runs exactly the same background tasks as my previous system (I reinstalled the same softwares). I don't want to fill this post with lots of screenshots, so I will simply type in what I see of interest.In the BIOS, all SATA drives are correctly listed: SATA 1: ST3320620AS SATA 2: WDC WD1600JS-55MHB SATA 3: WDC WD1600JS-58NCB SATA 4: not detected SATA 5: HL-DT-ST DVDRAM GH SATA 6: not detectedBIOS SATA Configuration: Enhanced (choices are Disable, Compatible, Enhanced)Configure SATA as: IDE (choices are IDE, RAID, AHCI)Hard Disk write Protect: DisabledSATA detect time out: 35 secWindows Device manager listed SATA Controllers:- Intel® ICH9 2 port Serial ATA Storage Controller 2 - 2626.- Intel® ICH9 4 port Serial ATA Storage Controller 1 - 2620.Connected SATA Drives description:- SATA port 1: Seagate Barracuda ST3320620AS (320 GB/7200 RPM/16MB buffer) with 2 partitions, C: 60 GB, 45.4GB free (Boot partition holding OS and Software) D: 238 GB, 205 GB free (FSX)- SATA port 2: Western Digital WD1600JS (160 GB/7200 RPM/8MB buffer) with 2 partitions, F: 74.5 GB, 22.2 GB free (Data Storage) G: 74.5 GB, 40.4 GB free (Data Storage)- SATA port 3: Western Digital WD1600JS (160 GB/7200 RPM/8MB buffer) with 1 partition, H: 149 GB, 58.9 GB free (Backup & Download Storage)- SATA port 4: empty- SATA port 5: HL-DT-ST DVDRAM GH20NS10 (LG DVD/CD-ROM) as Drive E:- SATA port 6: emptyTHE BENCHMARK using HD Tach analysis shows the poor performance of the Seagate drive on my system):Seagate Barracuda ST3320620AS on SATA port 1, quick bench: 4.3 MB/s Burst Speed sequencial read test - random access 13.4ms - CPU utilisation 26% - average read 4.2 MB/sSeagate Barracuda ST3320620AS on SATA port 1, long bench: 4.3 MB/s Burst Speed sequencial read test - random access 13.3ms - CPU utilisation 21% - average read 4.2 MB/sWestern Digital WD1600JS on SATA port 2, quick bench: 168.6 MB/s Burst Speed sequencial read test - random access 13.5ms - CPU utilisation 0% - average read 55.2 MB/sWestern Digital WD1600JS on SATA 2 port, long bench: 168.8 MB/s Burst Speed sequencial read test - random access 13.5ms - CPU utilisation 0% - average read 55.0 MB/sWestern Digital WD1600JS on SATA port 3, quick bench: 202.3 MB/s Burst Speed sequencial read test - random access 13.3ms - CPU utilisation 0% - average read 52.6 MB/sWestern Digital WD1600JS on SATA port 3, long bench: 202.1 MB/s Burst Speed sequencial read test - random access 13.4ms - CPU utilisation 3% - average read 52.5 MB/sEven though the randon access time is normal on the Seagate, the average read time is extremely low.I am a power user, but not an expert. Any idea What is going on ?


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites
Guest baksteen33

Hi Nbouc, No worries! :-) With previous versions of FS, only a couple of 100MB had to be loaded, with FSX, GBs are loaded depending on a couple of factors. So, it's a good time to say goodbye to technically less ideal solutions. ;-) If you have OS and pagefile on C and FSX od D, the HD heads will always be travelling in between the partitions and during this time they won't read or write. With modern drives, close to 1MB of data could travel instead of the heads travelling. If you like, search for 'HD head thrashing'. It's also good to have large partitions with heaps of free space, OSs calculate a couple of dependencies based upon the free partition space. At least you gave the OS 50GB. ;-) Good. OTOH, FS(X) starts in a even slower place to boot, secondary partitions are always slower than primary partitions. 50/~270 already takes ca 10-15% off your max HD throughput. Of course there are reasons to do as some still do, but we should maybe leave our personal organisational wishes aside and do what's opportune for the OS? FWIW, I'm confronted with these kinds of situations frequently - supporting an addon which depends on good I/O - if it's not sub-ideal amounts of RAM, it's very often partitioned drives. All I usually do to successfully help with the latter, is, explain what I just did and then 'leave it to Beaver' whether the advise is accepted or not. You might not be surprised to hear that the vast majority of I/O probs are solved by either plugging more RAM or setting up the OS/FSX properly. It's a very delicate task to bring this across sometimes. ;-) Anyhow, this is only practical stuff and doesn't offer an explanation about the 320GB's slowness. AIM, if the drive isn't broken it should be just as quick. 'Standard IDE' doesn't look right for example. It should read 'something Intel', i.e. 'Intel ICHx SATA xxx controller.' You should for example install winXP with a F6 (driver) floppy, otherwise no AHCI (or Raid). Hope you get it sorted, kind regards Jaap

Share this post


Link to post
Share on other sites

Thank you Jaap and Avcomware for your help. After reading your comments, I tried something just to see what would happen and it solved the problem. HD Tach software has been a very good testing aid in that matter.The Seagate Barracuda drive now bursts at 225 MB/sec with 67.7 MB/sec average seek. I solved the problem in 4 steps:1) I re-arranged the SATA connectors to only use port 3 to 6 (the goal being to isolate the specific controller of ports 1 and 2). FIRST GAIN: The Seagate drive performance jumped from 4.3 MB/sec to 158 MB/sec using the SATA port 3. Still not good enough, looked like a SATA1 drive.2) I checked the jumper on the Seagate drive and to my surprise, it was set to limit the drive to SATA1 performance (150 MB/sec). This should never come like this from factory, but it did. I set the jumper to allow SATA2 (300 MB/sec) and SECOND GAIN: the drive's performance jumped to 225 MB/sec.3) Just to be sure, I reconnected the drive to its previous SATA port 1 and was back at 4.3 MB/sec. 4) Returning it to SATA port 3 resulted again in a 225 MB/sec burst transfer rate. No need to say that it will remain on that port. I am not moving it again.The overall performance of Windows and all my programs have increased a lot. Programs start faster, files opene faster, file copying is now super fast on the Seagate (within itself as well as between drives). The other SATA drives are still performing well (they always did).FSX is performing a lot better. No more stutters, textures load faster, FSX starts faster, menu access is way faster (no more delay after clicking the mouse).I am now a happy camper.It seems that either one of my SATA Controller or the SATA port #1 is at fault. They both use the same driver (Intel 8.3.0.1011). I'll check with my vendor for a Motherboard replacement. I suspect the problem comes from the SATA port #1 since my Western Digital drive performed well when it was connected to SATA port 2 (same Controller for ports 1 and 2). I guess your input helped me look in the right direction.Have a wonderfull week !(I sure will ... going FSXing now)


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites

Hi Jaap,I solved my problem. I posted the solution under my initial message so that both you and Avcomware can see it easily.Thank you for those very interesting facts regarding modern drive usage. I must admit that my habit of installing FS on the second partition of the main drive comes from many years ago when having a 20 GB hard drive was insane. I might rethink my way of doing it and eventualy install FSX on a separate dedicated physical drive.On my previous system (P4 3.6 GHz) I ran the SATA drives in IDE mode for a year a so, then, when I had to re-install Windows, I took the opportunity to set it up in AHCI mode (F6 disk method). I did not notice any significative improvement in SATA drive performance, which lead me to use easier to setup IDE mode again on the new system (I don't use nor need RAID for the moment).Being over 50 and using computers since I was 20 makes me a knowledgable person, I guess, but it also makes me an individual that sticks to hold habits simply because they make me feel comfortable and secure LoLBest regards,


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites

Hi Nbouc.I am glad you solved your problem. Isolating, as much as possible, narrows the items you have to deal with. It sounds like you know what you are doing, it

Share this post


Link to post
Share on other sites
Guest jahman

>A logical drive,>Partitioning, has very little overhead, just a lookup table,>we are talking a few CPU cycles. Compare that to the Seek />Access time which is in milliseconds and you can see that this>is negligible. Um, actually you are forgetting about what happens if the HDD is being accessed simultaneously by 2 different pieces of software, in this case the Win operating system (DLLs and swap file) and FSX (program and terrain files).What 2 logical drives on one single physical drive does is make sure the files are physically *very far* from each other on the disk platters. When both logical drives are being accessed simultaneously (as would be the case when running FSX) the disk heads would be thrashing back and forth from one logical drive to the other and back and the time wasted could be in the order of 25 mSec for a full back-and-forth (compare that with the time average rotational latency of a Western Digital Raptor 10,000 rpm HDD of about 1/20th of 1 mSec, i.e. 500x faster!)Ideally get on physical drive for your operating system, another for the swap file, a third one for FSX and a fourth one for the terrain files.generally try to use a larger number of smaller drives than a smaller number of larger drives. Use larger drives for backup only.And get the 10,000 rpm Raptors if you can!Cheers,-jahman.

Share this post


Link to post
Share on other sites
Guest dswo

Thank you, Nbouc, for taking the time to document your solution in detail. It's very tempting, when you've worn yourself working on a problem, to NOT follow-up like this, but by making a record of what worked, you will have helped a lot of people you'll never meet.

Share this post


Link to post
Share on other sites

Hey, good advice. That's one thing I haven't thought of checking. They're all using UDMA-5 (ATA100), which looks OK to me since I am running the SATA drives in IDE mode.


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites

Hi David,You are so right. I've seen quite a few posts missing the end story or simply stating "solved" without any description of the solution, and I find that disrespectful to those trying to help. Furthermore, other people's solution has sometimes helped me fix my own problem, saving me the trouble of posting for help. So, if I can do the same for others, why not ;-)Take care.


 

Normand

Intel i7 9700K @ 4.9 GHz / Asus Prime Z390-A / 32GB DDR4 3200 MHz / MSI RTX 4080 / PSU 750 Watt / Microsoft Flight Simulator / Windows 10 Pro x64

Share this post


Link to post
Share on other sites

Hi, jahman.>What 2 logical drives on one single physical drive does is make sure the files are physically *very far* from each other on the disk platters. When both logical drives are being accessed simultaneously (as would be the case when running FSX) the disk heads would be thrashing back and forth from one logical drive to the other and back and the time wasted could be in the order of 25 mSec for a full back-and-forth (compare that with the time average rotational latency of a Western Digital Raptor 10,000 rpm HDD of about 1/20th of 1 mSec, i.e. 500x faster!)

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
Sign in to follow this  

  • Tom Allensworth,
    Founder of AVSIM Online


  • Flight Simulation's Premier Resource!

    AVSIM is a free service to the flight simulation community. AVSIM is staffed completely by volunteers and all funds donated to AVSIM go directly back to supporting the community. Your donation here helps to pay our bandwidth costs, emergency funding, and other general costs that crop up from time to time. Thank you for your support!

    Click here for more information and to see all donations year to date.
×
×
  • Create New...