Code: Select all
MB: Asus A7N8X-E Deluxe (with Sata controller Silicon Image 3112)
HDD: WD 1600JS
Shouldn't a Sata drive be a bit faster than an IDE drive?
Sata HDD on SiI 3112
Message
Sata HDD on SiI 3112
I have a bit of a problem. I have this configuration:
Can anybody tell me some performance values for this configuration? With kernel 2.6.20 I get values round 200MB for cached reads an 40MB for buffered disk reads, while with kernel 2.6.17 the first value goes up to 400MB.
Shouldn't a Sata drive be a bit faster than an IDE drive?
Shouldn't a Sata drive be a bit faster than an IDE drive?
- NeddySeagoon
- Administrator
- Posts: 56108
- Joined: Sat Jul 05, 2003 9:37 am
- Location: 56N 3W
darken,
How are you measuring the data transfer speeds ?
If you use hdparm, what version do you use and is it the same version in both cases ?
The high number is the data transfer rate from the cache RAM on the drive. The lower number in each case is the sustained data transfer rate, which is normally limited by the head/platter data rate.
40 Mb/sec is a little on the low side but it depends where on the disk the measurement it made. Zoned drives have more sectors per track at the edge of the platter than near the spindle. You can see this when you speed test all your partitions.
hdparm had a speed reporting bug for buffered disk reads by about a factor of 2. Its been fixed.
From the above, you can see that SATA drives are not really faster that PATA drives. Indeed, the head and platter assemblies can be identical, with only the electrical interface being changed.
How are you measuring the data transfer speeds ?
If you use hdparm, what version do you use and is it the same version in both cases ?
The high number is the data transfer rate from the cache RAM on the drive. The lower number in each case is the sustained data transfer rate, which is normally limited by the head/platter data rate.
40 Mb/sec is a little on the low side but it depends where on the disk the measurement it made. Zoned drives have more sectors per track at the edge of the platter than near the spindle. You can see this when you speed test all your partitions.
hdparm had a speed reporting bug for buffered disk reads by about a factor of 2. Its been fixed.
From the above, you can see that SATA drives are not really faster that PATA drives. Indeed, the head and platter assemblies can be identical, with only the electrical interface being changed.
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.