Possible cause of a slow hard drive?

[Jeff7]

My mom's PC is running pretty slow. I'm an optimization freak, so I've tweaked the system for max efficiency already - updated drivers, BIOS, etc.
But the hard drive is painfully slow. It takes about 15s to load Internet Explorer, and 25 to load Media Player. And about a 3 minute boot time. This thing is slow. Specs:
T-bird 750@733 (adjusted for 133FSB).
Soyo K7VTAPro motherboard
512MB PC133
8.4GB ATA66 HDD (WD84AA) - 5400rpm, 2MB buffer, using 80-wire cable.
WinXP SP1

HD Tach won't run on it - it needs the registered version to work on "Windows NT" as it thought the system to be. Sisoft Sandra reported around 3500KB/sec - on par with their level for a 4000rpm HD with a 256KB buffer. Any idea what's causing this?
DMA is enabled, running in Mode 4 according to the Properties; 4-in-1's installed. There's no other HD's in the system, and the CD-drive is on the secondary controller.

Update, 12-9-02 (same as posted below, here for convenience🙂 : This problem is really bizarre now. I thought it was the HD; fine, RMA'd it. ($11 value they said, oh wow.) Got the new hard drive, same problem. Next to go was the motherboard. Got a new one, different brand, and started a clean install of Windows. After it took about 5 minutes to reach 1% of the format, I figured it was still acting up again. Shut down the PC, took the drive out and flipped it over, label side up. Tried again - it formatted the entire drive in a few minutes and installed Windows. It also detected quicker by the BIOS when it was facing this direction; it if was PCB up, the BIOS had trouble seeing it. I've never had a drive that is picky about its physical orientation, especially 2 in a row!
Anyway, it's running normally now, label side up.

Anyone want to venture any theories here? Maybe it's a quantum singularity that compromises the signal when the drive is upside down. Or just evil demons again. 😕

 

[alm4rr]

what brand?
maybe its just a crappy hd?

 

[Jeff7]

Western Digital.

 

[alm4rr]

what is the benchmark breakdown in sandra?

that will be the buffered read, sequential read, random read, buffered write, sequential write, random write, and AVG access time.

 

[Jeff7]

Here's the results:
Cmd queue depth: 4cmds
Test filesize: 511MB
Buffered Read: 44MB/sec
Sequential read: 1175KB/s
Random Read: 751KB/sec
Buffered Write 23MB/sec
Sequential Write: 15MB/sec
Random Write: 5MB/s
access time: 31ms estimated

62% free space
I just defragged it too; no change in speed.
I'm going to take the drive out now and see if it's under warranty. It might be; it's not too old actually.
Ok, it should be under warranty - manufacture date is Dec 10, 1999. However, the warranty verification thing at WD's website is down. Again. I've tried to check serial numbers there a few other times in the past few months; every time, the page says "unavailable for maintenance." Anyone else get this a lot?

Um, the drive just got better now. What'd I do to it? I took it out of the case, flipped it over to see the part and serial numbers, and put it back the way it was. Now I get the following:

11578KB/s overall
Buffered Read: 45MB/sec
Sequential Read: 16MB/sec
Random Read: 3536KB/sec
Buffered Write: 11MB/sec
Sequential Wrtie: 16MB/sec
Random Write: 5MB/sec
Avg access time: 14ms est

 

[mechBgon]

Owie. 🙁 On the bright side, after being launched once, a program should launch from RAM the second time, given an NT-based OS and 512Mb of RAM. edit: and if it doesn't launch faster the second time, maybe it's the RAM?

Is it bothering your mom, or just you? I know my dad's 333MHz Cyrix-based Emachines drives me crazy, but it gets the job done fine for him...

As for the actual issue: I recently had a brand-new 7200rpm Maxtor that was strangely slow too, as if its actuator was sticking. Took forever to do anything 😛 It came up flawless on Maxtor's test utility, however! I would just get another hard drive if I were you.

 

[Jeff7]

Programs do launch faster the second time; the RAM has been tested with Memtest86; all's well.

Ok, got through to the WD warranty thing - the drive IS under warranty for another month. Think I'm just going to do that. I'll see what their utilities say about the drive; probably nothing as usual though - like the Check Engine light on my dad's car. It went on and stayed on for a few days, and as he was finally driving out to have it checked, it went off.

So, any theories as to the drive's miraculous healing? Almost all the numbers increased drastically, especially the sequential and random read speeds. Maybe it knew I was going to test it? I'm serious though, I didn't replace the cable this time - all I did was flip it over, flip it back, and put it in the case again.😕

Edit: Ok, just tried the WD Data Lifeguard v10.0 - Couldn't load some IBMDOS file; system halted. Mmmkay.

 

[Jeff7]

Update finally: this problem is really bizarre now. I thought it was the HD; fine, RMA'd it. ($11 value they said, oh wow.) Got the new hard drive, same problem. Next to go was the motherboard. Got a new one, different brand, and started a clean install of Windows. After it took about 5 minutes to reach 1% of the format, I figured it was still acting up again. Shut down the PC, took the drive out and flipped it over, label side up. Tried again - it formatted the entire drive in a few minutes and installed Windows. It also detected quicker by the BIOS when it was facing this direction; it if was PCB up, the BIOS had trouble seeing it. I've never had a drive that is picky about its physical orientation, especially 2 in a row!
Anyway, it's running normally now, label side up.

Anyone want to venture any theories here? Maybe it's a quantum singularity that compromises the signal when the drive is upside down. Or just evil demons again.😕