• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Question Why is this HDD so slow?

WD Green HDDs have a rep for being slow, but I would like to understand why (at least by looking at the specs); this drive isn't much slower than I expect for a WD Green of its age.

Take for example this drive: WD15EARS.
Trying to find some decent (and hopefully accurate) specs is tricky, for example this site:
I suspect is either AI-fuelled or maybe a bunch of data lazily slurped from different sources because the "failure symptoms" information is either very generic or so generic that it's as good as junk.

The SMART data looks fine, but the write performance of the drive is as poor as I would expect from a WD Green 3.5" drive, so in this case a 1.5TB drive that is probably averaging about 45MB/sec write performance (I saw it peaking at about 65MB/sec so logically it has a USB3 link) via USB3 for large enough files to get its teeth into. If I picked a random 2.5" old 500GB/1TB drive from my collection, I'd expect to write ~100MB/sec via USB3.

Any ideas?
 

WD15EARS is only the first half of the model number before the hyphen, and there are multiple variants. The one I found was 7200RPM so spindle speed wasn't the culprit (5400RPM or lower is often found). 64MB cache is good, so that isn't it, either. I suspect the drive therefore uses SMR (DMSMR) which makes the drive horrendously slower in order to pack more capacity per platter


 
Full model number, copied and pasted direct from Linux: WDC WD15EARS-60MVWB0

I haven't seen any typical indicators that it's SMR (e.g. the drive needs to stop and have a rest after a long write), I think it's older than that.
 
Check that the drive is connected by UASP and not the old BOT protocol.
That gives you NCQ etc which should help.

If not it's because the drive enclosure ctlr chip is too old.
Check out:
MaximumTransferLength
USB-WriteCache V0.2

at the bottem of
https://www.uwe-sieber.de/drivetools_e.html
(Uwe Sieber is the writer of USB drivers for DOS and/or Win95 IIRC)

I'd also disable write cache buffer flushing and do:
copy-paste-(verify)-delete
rather than
cut-paste.

Lastly (not free) PrimoCache with a deferred write cache (and SSD cache) is about as fast as you can make the/any drive.

"But that's fake speed due to RAM cache"
is a shallow way of looking at it:
It cuts out a LOT of writes by updating constantly updated files in RAM and coalescing small writes into large sequential writes which are much faster on any drive.
 
Back
Top