XPost: alt.comp.hardware.pc-homebuilt
Lynn McGuire <
lynnmcguire5@gmail.com> wrote:
On Wednesday, August 19, 2020 at 5:10:16 AM UTC+8, Paul wrote:
The Helium gas is guaranteed to stay inside for 5 years.
Does anybody have a Helium drive more than 5 years old? They arrived
2013 IIRR.
Not yet here. In fact, I am not even sure how tell if a hard drive is
helium filled ?
Lynn
Hydrogen atoms will pass through metal. Helium atoms will not pass
through defect-free metal, but will pass through the defects. I doubt
HDD makers are going to that expense to ensure no defects in each metal
shell half, plus the HDD case is not solid metal which necessarily
mandates the metal shell is not defect-free.
Although users complain their HDDs drive mechanically fail after a few
years, obviously only the negative reports of such are visible. Users
do not proclaim when their HDDs survive 10 years, or more, which is
often the MTBF rating by the manufacturers.
Helium drives, despite getting introduced back in 2013, have not
accumulated enough failure statistics to really know their real MTBF.
They're probably a short-lived fad. Naturally mined helium that took
4.7 billion years to produce is a limited resource not expected to
survive more than another 100 years under current consumption, and using
it HDDs increases consumption, and artificially produced helium under super-high pressures is too expensive. As the helium reserves get used
up, price will go up either due to its increased rarity or through gov't imposed taxation to prod cessation of consumption. Some other
technological advantage will be needed for mechanical drives if they
manage to survive advanced in flash memory (e.g., memresistors) and
other technologies with increased storage density.
I've not heard that 5 years is a statistic at the middle to top of a
bell curve of measured MTBFs for helium drives. I think that's just a guestimate but while figuring HDDs get replaced before that, so the
guestimate is longer than either estimate use-time or the warranty.
https://www.backblaze.com/blog/helium-filled-hard-drive-failure-rates/
May 3, 2018
and:
https://www.zdnet.com/article/do-helium-filled-hard-drives-have-a-lower-failure-rate/
which references the Backblaze data. The conclusion is that
helium-filled HDDs do not outlast air-filled HDDs, but there is some
fudging with analysis of the stats. The market is not to end users, but
to data centers where reduction in power and reduced heat (so less air conditioning load) are important. In the 3-year test by Backblaze,
leakage is not a problem, but no idea how that may degrade with longer
use of the He drive.
Apparently some drives implement SMART attribute 22 to estimate the
remaining helium inside the shell. Perhaps the measure is based on
pressure [change] or a probe that is affected by density of helium
atoms. According to the above Backblaze article, helium has not proven
to lengthen longevity of HDDs, but that wasn't the point of going to
helium which was to up arial density by letting the heads fly closer to
the platters and reduce drag on the platters to reduce power
consumption. The reduction of power is unimportant to end users in a deployment of 1 or a dozen HDDs, but for data centers running thousands
of them. The advantage of increased arial density will be short-lived.
Since the MTBF is not improved, there is little need for helium HDDs by
end users other than to experiment. Oh, helium transfers heat faster
than air, so those drives cool faster; however, end users can implement
better cooling setups or solutions.
To the end user, like you and me, there is no advantage nor disadvantage
to using HE drives -- if you omit the higher price tag for He drives,
but then end users never omit that decision factor. Reduced electrical
and air conditioning load are a factor that data centers consider that
would offset the increased price for He drives, not something that end
users could even measure on such a small scale deployment of HDDs.
The maker and model of the HDD and looking up the specs is how you would identify which are helium filled. There's nothing about the drive's
outward appearance that will identify air versus helium filled. The
stick-on label might sometimes help, like "HelioSeal" or "He" (HGST) on
the label, but many have no such identification on the label.
Air-filled HDD:
https://www.newegg.com/seagate-exos-x16-st10000nm001g-10tb/p/N82E16822184842
Helium-filled HDD:
https://www.newegg.com/seagate-enterprise-capacity-3-5-st10000nm0096-10tb/p/234-000S-00083
He drive is $23 higher. Perhaps not enough for an end user to care
about when buying 1 or 2 of the He drives, but is wasted money for
nebulous advantages in such a deployment scenario. The price increase
is very important to data centers that are purchasing hundreds to
thousands of HDDs unless the added cost of He drives is offset, and
more, by reduced electrical load for the drives and air conditioning,
but there seems little to no longevity to He versus air.
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)