Buying my first SATA drives

M

MrFace

2[H]4U
Joined
Feb 23, 2003
Messages
2,716
My IDE data drive is dying, so I decided to get with the times. Will I see a speed difference with a Sata 3gb drive compared to IDE ATA 100?
 
Will you see it, probably not.
Will benchmarks show a slight imporvement? Probably so.
 
Well, you'll probably see a speed boost if only because the SATA drives are probably much newer than your IDE drives, so they will have better performance because of new technologies being used in production and in systems (like NCQ for multi-user, multi-I/O environments).
 
good luck! I noticed a difference jumping from a 30gb Hitachi DeskStar 5400 RPM IDE drive to a Seagate Barracuda 160gb 7200 RPM sata, heh. not sure if your jump was as drastic, but i could definitely see the difference in my upgrade.
 
Only thing that will be faster on the new drives is the 16mb cache. My current drives are only 8mb. What is the point of SATA if it isn't any faster?
 
MrFace said:
Only thing that will be faster on the new drives is the 16mb cache. My current drives are only 8mb. What is the point of SATA if it isn't any faster?
Click to expand...

The SATA standards support other cool things like hot swapping and port multiplexing (you can plug in multiple drives to just one port), which would be a big reason why there is a 3 Gbps standard (and soon a 6 Gbps standard). Also, in SATA, all the ports are independent of the others, there is no sharing of the bandwidth between ports (like in Master-Slave IDE configurations). Additionally, SATA interface cables are much thinner than the wide ribbons used for PATA. Features like NCQ (only some SATA drives have this) help with performance in multitasking/multiuser environments. Not to mention, there is also a standard for eSATA (external SATA) which provides a fast link to external hard drives (and I think even makes the system recognize the drive as an internal drive, so you can do stuff like install and boot Windows off the drive).

And how old is your drive? Newer hard drives have higher density platters, may use technology like parallel recording, and might seen other mechanical or logical improvements that will probably help with performance. 16 MB cache is not as awesome, I don't think.

The interface is really meant to be extensible for the future, but 3 Gbps is expected to be sufficient for drives for a while (at least until fast solid state drives are plentiful I'd think).
 
MrFace said:
Only thing that will be faster on the new drives is the 16mb cache. My current drives are only 8mb. What is the point of SATA if it isn't any faster?
Click to expand...


the point is its got a cool looking colored cable...cha ching!
 
i got 2 sata drives..ones 300 gigs programs the other is 500 gigs media...one day i might put in a second 500 gig drive and run raid, cause its built in....anyway...get sata!
 
MrFace said:
What is the point of SATA if it isn't any faster?
Click to expand...
Smaller cables are nice.
SATA is faster, since it is a nice marketing tool. Also, if port multipliers were common then there'd be a reason to go with 300MBps SATA.
 
bashPenguin said:
The SATA standards support other cool things like hot swapping and port multiplexing (you can plug in multiple drives to just one port), which would be a big reason why there is a 3 Gbps standard (and soon a 6 Gbps standard). Also, in SATA, all the ports are independent of the others, there is no sharing of the bandwidth between ports (like in Master-Slave IDE configurations). Additionally, SATA interface cables are much thinner than the wide ribbons used for PATA. Features like NCQ (only some SATA drives have this) help with performance in multitasking/multiuser environments. Not to mention, there is also a standard for eSATA (external SATA) which provides a fast link to external hard drives (and I think even makes the system recognize the drive as an internal drive, so you can do stuff like install and boot Windows off the drive).

And how old is your drive? Newer hard drives have higher density platters, may use technology like parallel recording, and might seen other mechanical or logical improvements that will probably help with performance. 16 MB cache is not as awesome, I don't think.

The interface is really meant to be extensible for the future, but 3 Gbps is expected to be sufficient for drives for a while (at least until fast solid state drives are plentiful I'd think).
Click to expand...

Thank you. Looks like the only useful thing is the fact they don't share ports. Oh well :)
 
All of this is b ullshit. First I would like to say, opening internet explorer will be a jffy. SOme of the fastest maxtor 133 drives don't even touch my new wd aajs. It's simply amazing, and going w/raid inst' gonna help much. Just a heads up
 
EnderW said:
DRIVE SPEED IS INDEPENDENT OF INTERFACE
Click to expand...

scsi is an interface right - and there is a correlation in performance. the reasons may not be because of the interface, but the two things are not independent.
 
sovs said:
scsi is an interface right - and there is a correlation in performance. the reasons may not be because of the interface, but the two things are not independent.
Click to expand...

The speed of the drive is constant. It will spin at the rated speed it is made for. How fast the data is accessed is a whole 'nother story (interface).
 
No, but 15K Serial attached SCSI (which uses a SATA port) do. But what is the argument in that? 15K, 7200, 10K are all mechanical properties of the DRIVE, not a property inherent in the interface the drive uses.
 
bashPenguin said:
No, but 15K Serial attached SCSI (which uses a SATA port) do. But what is the argument in that? 15K, 7200, 10K are all mechanical properties of the DRIVE, not a property inherent in the interface the drive uses.
Click to expand...

sovs said:
that doesn't sound very independent to me.
Click to expand...

Surely, the interface can play a critical role in drive performance. For example, when running a recent generation IDE drive on a PIO mode-1 controller, the performance would be severely limited. At the same time, the reasons for 15k rpm drives to be limited to the SCSI market are more associated with the margins with that segment, than technical in nature.

It appeared to me that bashPenguin wanted to educate the `average user' about the negligible effect on performance exhibited by moving from a 150MiB/s to a 300MiB/s SATA interface. I think that most regular posters on here are painfully aware of the excellent job that marketing departments have done in convincing users that SATA-2 (lmao at that) offers incredible benefits over the `old' interface. Statements such as ``DRIVE SPEED IS INDEPENDENT OF INTERFACE'' are technically incorrect. However, if bash was extremely detailed in his expression, he would not have cleared up the confusion about drive speed, but rather added to it.
 
You must log in or register to reply here.
Back
Top