Should a larger hard drive take longer to write?

[Mike]

About two months ago, I bought a Toshiba external portable 500 GB hard drive, which I use both for backups and carrying files around. I've been very pleased with the drive.

While at Fry's the other day, figured I'd buy a second drive, and found a Toshiba 640 GB drive. I figured the extra space would be useful, so I bought it.

After getting it home, I deleted the FAT32 partition and reformatted the drive for NTFS. It works, but it seems to be much slower. (I reformatted the older drive as NTFS, too.)

Today, I tried a test. I copied a 1.2GB directory tree (MP3s) to the old drive and to the new drive. Copying to the new drive took 193.5 seconds (3 mins 13 secs). Copying to the old drive took 64.83 seconds (1 min 4 secs). In other words, copying to the new drive took three times as long.

Is there any reason why a 640GB drive should take longer? Or is this a defective drive? ...thanks!

[sidney]

Is there any reason why a 640GB drive should take longer? Or is this a defective drive?

There are quite a few different parts of the specs besides the size that could make a difference. What are the model numbers of the two disks? In Windows XP you can see that in Start Menu | Control Panel | System, select the Hardware tab | Device manager | expand 'disk drives' then right click on the hdd in question. select properties then click the 'details' tab. The from the drop down box select Hardware IDs. I don't know what the equivalent is in later Windows.

I'll look up and the specs of the models and see if I see anything that explains the difference.

[Mike]

I'll look up and the specs of the models and see if I see anything that explains the difference.
Thanks, Sidney. Since I had other errands to do, I went ahead and returned the drive to Fry's. I asked for an even exchange, but no more 640GB drives were available, so I downgraded to a 500GB model (the same as I already have), with a refund of the difference in price. It took a few seconds longer to copy my test folder, but the difference wasn't significant.

Should I encounter this in the future, I'll get the information you suggested and see what you can find. ...thanks!

FWIW, I previously bought WD Passport drives and have been happy with their performance. However, the latest Passport drives require installation of special drivers on the host computer, which isn't always an option for me.

[sidney]

Should I encounter this in the future, I'll get the information you suggested and see what you can find

Probably better to get the information before buying it :)

For future reference, there are three specs that can make a difference in speed (plus the speed of the interface, but since you are comparing USB 2.0 external drives that will be the same). The biggest factor is the rotational speed, standard being 5400 rpm or 7200 rpm, with some expensive high end disks going at 10,000 rpm. The faster the disk spins the faster it can get to the place on the disk that it is going to read or write. Also the faster it can actually read or write a track of data, but it turns out that doesn't matter much given that it will be faster than the USB 2.0 anyway. The second parameter is seek time, the time it takes the disk drive to move the head from one track to another. And the third parameter is how big the disk drive's memory buffer (cache) is.

There is no hard and fast rule about how these three things affect speed because what happens depends on how you are using the disk. For example, if you are copying a large file the seek time won't make much difference if the file on the external disk is contiguous. If it is fragmented, then seek time becomes very important. Or rotational speed might not make much difference for that transfer if there is a big enough buffer so that the data can be stored in there while the seek is happening, then the actual write is done from the buffer. The only way to really know is to test, although all other things being equal a 7200rpm disk will be faster than a 5400 rpm disk, a disk with bigger cache buffer will perform faster than one with a smaller one, etc.

I just looked up a Toshiba external 640GB and 500GB USB 2.0 2.5" external drives and all I found have identical specs:
.
# Rotational Speed: Up to 5400RPM
# Average Seek Time: 12ms
# Cache Buffer: 8MB

Do your drives look like this product range (possibly a different color choice)?
Rocket Red (http://laptops.toshiba.com/accessory/HDDR640E04XR)

Or do you have a 3.5" external drive that requires an external power adaptor? Toshiba does make 7200rpm 3.5" disks.

There is a strange reason why it is possible for a 640GB disk with identical specs to a 500GB disk to perform much more poorly at a particular task. Consider that the 640GB disk is going to pack more data per track (one revolution of the disk) than the 500GB disk. If, when you are performing the copy that was your test, the system was able to process the amount of data on one track of the 500GB in less than 1/5400 of a second, but cannot process the larger amount of data that is on one track of the 640GB disk in that same 1/5400 second, then copying using the 500GB disk might result in reading a track while processing the data and being ready to start reading the next track immediately. On the 640 GB disk that same task might look like reading a track, by the time the data is processed the disk has just passed the start point of the rotation, waiting the extra 1/5400 second for the start of the next track to come around again, etc. That looks like taking twice the time to do the same processing. I'll hand wave how it could be three times longer, but you can imagine how it could happen if there's some other step involved where the timing on the larger disk is off like that.

[Mike]

Probably better to get the information before buying it :)
Very true. However, when buying a device like this, there's a limit to how much homework I will do prior to the purchase. I've learned that after doing the research, often I won't find the "best buys" at Fry's.

Or Best Buy.

...there are three specs that can make a difference in speed...

Thank you for the education!

For example, if you are copying a large file the seek time won't make much difference if the file on the external disk is contiguous. If it is fragmented, then seek time becomes very important.
Understood. That part I recall from my early days writing software to retrieve fragments of files to try to stitch them back together (sort of a precursor to Norton's Undelete, but mine wasn't nearly as sophisticated). I just haven't had to deal with hardware for, ummm, 18 years or so, and (per the other conversation with Judy), I've learned that I just don't have space in my life to keep up. <sigh>

OTOH, since I've had the older drive for several months, I'd expect files written to it to be spread across many non-contiguous clusters, while the new drive should allow writing each file as a series of contiguous clusters.

Do your drives look like this product range (possibly a different color choice)?
Rocket Red (http://laptops.toshiba.com/accessory/HDDR640E04XR)
That's the 640GB drive I tried. The 500GB (both old and new) are Liquid Blue (http://www.toshibadirect.com/td/b2c/adet.to?poid=445558). Neither requires the external power adapter. I'll only consider drives that don't require carrying an extra wall wart (or use of a second USB port).

There is a strange reason why it is possible for a 640GB disk with identical specs to a 500GB disk to perform much more poorly at a particular task.
That makes sense to me. However, I'm not going to do any further testing. I have a drive that suits my needs, and as much as I'd love to have the time and education to investigate fully, it just isn't going to happen. Maybe if we win the lottery and I retire. And I visit all 1000 places.

Thanks again for the suggestions/ideas!