AHCI causing more HDD vibration and drive wear?

[Immor32]

AHCI causing more HDD vibration and drive wear?
The NCQ feature of AHCI seems to mean the drive head moves around faster, which would cause more vibration and more wear and tear on the drive.
Like shown here:
https://expertester.wordpress.com/2008/07/24/ahci-vs-ide-–-benchmark-advantage/
The non NCQ drive process is a very gentle movement of the head, while NCQ means the head has to jerk right, then left, then right.

This would match my own experience with using AHCI on a Seagate drive that is getting a lot of bad sectors after barely two years.
Is this idea of mine making sense or am I missing something?
Because I could care less about the few percent speed NCQ gives, if it means my drives die like flies.

 

[Someone Somewhere]

NCQ should mean that the drive returns to idle faster, though.

Also, that image is created to show a pretty ideal case for a random read/write; they're all in a smooth line travelling outwards. It's entirely possible for it to be just as choppy as the NCQ image, or for the NCQ image too be much smoother.

Sample size of one is less than worthless for things like HDDs with infrequent failures.

 

[Leaps-from-Shadows]

NCQ is Native Command Queuing. It automatically re-orders the commands to make the heads move more efficiently across the platters. This should mean the heads move less, not more.

NCQ has been enabled in drives for years. If it had a serious negative impact on longevity, it would have been discovered by now.

In other words, if your drive has bad sectors it was going to have bad sectors whether or not NCQ was active.

 

[Immor32]

Good point Someone, although I get the feeling NCQ has been designed with only speed in mind, as that is all the average customer looks for. If the HDD fails faster it just means another unit can be sold.
On an unfragmented disk and for larger files, I don't see NCQ doing anything but harm. Not that I could be sure, but I'm going back to IDE anyways. That will also allow "HDD Regenerator" to do its magic.

 

[Someone Somewhere]

Linux's hdparm tool claims that most drives have the ability to set the acceleration the head is under:

-M Get/set Automatic Acoustic Management (AAM) setting. Most modern harddisk drives have the ability to speed down the head movements to reduce their noise output. The possible values are between 0 and 254. 128 is the most quiet (and therefore slowest) setting and 254 the fastest (and loudest). Some drives have only two levels (quiet / fast), while others may have different levels between 128 and 254. At the moment, most drives only support 3 options, off, quiet, and fast. These have been assigned the values 0, 128, and 254 at present, respectively, but integer space has been incorporated for future expansion, should this change.

I've heard negative things about the effectiveness of HDD Regenerator.

On an unfragmented disk and for larger files, I don't see NCQ doing anything but harm.

Pretty sure NCQ will have zero effect there. It only makes a difference when the drive has multiple non-sequential items in its queue.

NCQ can significantly reduce head movement under some circumstances, by re-ordering the queue so that the sectors accessed are near-sequential.

 

[Immor32]

I would have lost a lot of important data if not for HDD Regenerator, not to mention have had to buy more drives. It never let me down.

So I switched to IDE. It was easier than I thought. Boot time is noticeably reduced and so is start time for some applications. NCQ is the worst. This sums it up:
"I'd gather that NCQ in most normal applications, like games, and with defragmented (if applicable) drive is infact slower because of the overhead required by the intelligent ordering increaseas the latency of the IOs. Not to mention the fact that NCQ always brings in variability to performance, even on linear reads, because of the re-ordering schemes."

Setting AAM sounds like a good idea. It seems my Seagate drive can't do it though. I already ordered a Hitachi drive, which should also live longer.