Yet another Win2KPro ATA100 Problem...
I'm hoping this is easier to solve than I think. My problem is that Sandra is reporting that my Maxtor ATA100 reports at about 8400 compared to a typical ATA100 at 24000 - in otherwords, it seems to be running PIO speed. I've been through this before, so here's what I've done to no avail:
Firstly, I just built a KR7A-RAID machine and just loaded Win2KPro SP2. As I understand it, the Microsoft Win2K ATA100 fix is included in SP2, so that's not the problem. I went into device mangler, and that is set to automagically see UDMA, and reports that the IDE ports are infact UDMA. I went into the BIOS and that too is reporting IDE ports on AUTO. I have installed the VIA 4in1 4.38 drivers, but still it shows as slow as molasses.
Now, I heard somewhere that if you have a ATA100 drive and a ATA33/66 drive on the same cable, then that is not good...is that true and would it cause this kinda problem? Right now, my machine is incredibly generic ... not much customization and everything is pretty much out of the box. What would be the fix for this?
A pizza for anyone with the answer (that works) Thanks Much!
You can run no faster than the slowest device on a controller.
actually u can mix and match the devices on a drive .. (I'll post the reference if i find it .. i lost it it was on either maxtors FAQ or maybe it was seagate... )
as long as they are UltraATA compatible . u should ofcourse be useing an ultraATA cable .... if ure having problems with the speed u might try to reinstall the drivers for the controller chip .. or an updated driver might help
i dunno about Via ... but Intel's application accelarator helped me on that issue a couple of times(i,e reinstalled that and the drive's speed increased)
oh i presume u have everything set from the Bios on manual. like UDMA 5 or PIO etc ect . the BIOS might not set it right properly if set on Autodetect (althought it is the reccomended method)
Last edited by philocipher; 04-21-2002 at 08:16 PM.
Well, put my ATA100 alone on the cable, tried swapping with a new cable (and yes, it's 80pin), My bios only disables UDMA and that only made it worse, and there seems to be no way to set auto to none in device manager for this port because it's grayed out. This is really a bummer...I must be missing something obvious. I guess the next thing to check is the hard drive itself? What could be causing this? Maybe I should use something besides Sandra? I'm running out of ideas. Any solutions you may have tried successfully would be much appreciated.
Another Update...going to the absurd now. Tried applying EnableUDMA66 to registry via Xsetup 6.1 ... no avail. Tried jumping the drive as master instead of cable select ... still no avail.
Then, tried benchmarking the other partition of this drive. Yes, it is a Maxtor ATA100 30 gig 7200rpm drive, partitioned as NTFS on both primary (where the OS is) and the other partition (for data). I mention this because I just benchmarked the other partition as 19880! Sandra say's that the average UDMA100 7200rpm drive is benched at 24000. At this point that is close enough to making me happy. However, the primary partition of the same harddrive is benched at 8450. Can anyone explain what is happening? How can the same physical device report at two different speeds? Thanks much
I don't know that one.It is odd.
But I have a suggestion:
Why don't you just try some other benchmark utilities to get an idea about how anything to do with OS and stuff primary part is running.Then you'd get an idea if it really is at 8450 or not.
You could try some of these:
OR even a Cd Burning prog to test the speed.
I would think that that 24000 would be the fastest ... as the hard drive spins at a constant speed
I mention this because I just benchmarked the other partition as 19880! Sandra say's that the average UDMA100 7200rpm drive is benched at 24000. At this point that is close enough to making me happy. However, the primary partition of the same harddrive is benched at 8450.
I would also think that the inner tracks would be slower than the outer track. (same as a Cd) i might be wrong ..
another factor is the cluster size. smaller sizez make for slower speeds but more efficient space usuage...and so one and hence forth
u said its a new drive.... according to Maxtors site . during the first 10 power cuycles the 'write verify ' is turned on so the drive performance is reduced
but i guess ure past that by now...
oh try the maxtors site. there might be a firmware update to update your ata66 to ata 100 instead of doing the reverse!.. if its not too old that is
Update: just got off the phone with MAXTOR, and they claimed this problem was out of their rhelm ... which is probably true. However, I was very tenacious and pressed them to at lease listen to my issue and he had an interesting twist. He said, and he did admit that this was a guess, that the reason for the primary partition being slower than the secondary partition running on the same port, same device, is because the OS is competing with the benchmark software. BTW: the device in question is a MAXTOR DiamondMax 60 Plus 5T030H3 (30gig ATA100).
Sounds strange, but it fits the puzzle, since I bench C: (10 gigs of the 30 gig device) as ~8350, yet the D: partition (20 gigs of the same device) is benched at 19880, as mentioned in previous posts. As desperate as I am for an answer, does this sound right? Can the OS really effect the transfer rate that much? What benchmarking alternatives out there give a good number that assures me that this partition on this drive is running mode 5 at it's maximum?
Thanks Much...this is really weird
OK, for all those who think that one can't mix DMA devices, here is an intersting post:
As far as my problem, I am going to sum it all up here:
1. Tried all the remedies short of salt over my shoulder
2. Downloaded 'Active S.M.A.R.T.' to verify the device is UDMA Mode 5 and it is
3. Downloaded Diskspeed32 (HDTach is not free for Win2K) to verify the speed is not what Sandra is reporting
4. Came across a utility from MAXTOR that varies the speed of the drive to make the acoustics of the drive quieter ... quieter, but at the expense of speed (I will look into that later)
1. Although I'm only 90% convinced, I think I'm OK
2. This was most likely a non problem, a wild goose chase
3. Sandra is not adequate for all disk benchmarks
4. If I'm looking for speed, then I need to go SCSI
5. There are as many opinions and remedies for this particular problem as there are for hic-cups
So, no one won the pizza, but as usual all your opinions were very intertesting and this was quite a learning experience. Maybe later I will post the graphs/results of this debacle.
Last edited by coyote; 04-23-2002 at 12:37 AM.
someone posted this
Again, today's chipsets all pretty much support independent (device) timing, so this is less of an issue than it once was.
!!DUH!! thats like ever since they started being IDE
I D E stands for . err u go figure it out . they have thier own controller on boa4rd ever since. and the ATA cable is just a data channel... so .. u cabn send data at 100 Mhz .. or 33 Mhz depending on which device is sending.. OR being sent to .