Friday, July 23rd 2021

Seagate Launching Mass Market 20 TB PMR HDDs in The Coming Months

Seagate has recently announced during their latest earnings call that they are preparing to launch mass-market 20 TB PMR hard drives in H2 2021. Seagate already has several business-focused 20 TB HAMR drives which are available in limited quantities to select partners. These new 20 TB PMR drives will feature two-dimensional magnetic recording and are already sampling with customers. Seagate is also working on SMR 20 TB drives for Hyperscale systems with specialized software. Seagate hopes this growing lineup of 20 TB drives will help them address specific customer needs.
Seagate CEOWe expect to begin shipping 20 terabyte PMR drives in the second half of this calendar year.
Source: SeekingAlpha
Add your own comment

58 Comments on Seagate Launching Mass Market 20 TB PMR HDDs in The Coming Months

#26
R-T-B
Jack1nNo, but I had the important stuff on my Google drive. Everything else is replaceable but it is a chore.
I'll agree that even properly managed data loss is a chore.
Posted on Reply
#27
StaticVapour
Not a fan of drive war, but funnily enough I've been waiting for this drive to finally bite the dust so I could replace it with SSD.. I'm still waiting...

1094 days and 19 hours the time of screenshot, it has some trouble with error rates, but for some reason it's not that much louder than it used to be
Posted on Reply
#28
newtekie1
Semi-Retired Folder
Andy ShiekhAt least its not shingled.
Nothing wrong with SMR drives if they are used in the correct application.
Posted on Reply
#29
thesmokingman
Now you can fit your whole porn collection one drive!
Posted on Reply
#30
Octavean
Some people just buy whatever's on sale and when said cheap drive fails swear off the brand ad infinium.

As has been stated here already, any manufacturer can have a problematic model or line. To that end, all HDD's aren't the same and manufacturers tend to charge more for the higher endurance models (which still can and do fail).

Point of interest though, WD has a tendency to alter drives while keep the naming scheme the same. This is a problem on its face. WD has done this with Green, Blue and Red drives much to the chagrin of end users. So fair point that no one wants one or more SMR Red drives to take down their RAID array. FWIW, WD Red Plus is now what WD Red used to be (CMR). WD Red is no SMR IIRC. Evil, jacked-up AF, sure but it drives home the point of specific models under a brand / manufacturer name being problematic without them all being a problem.
thesmokingmanNow you can fit your whole porn collection one drive!
Ha,...!!!

Mabey you can,......

No way my collection will fit on just one drive,...... :)
Posted on Reply
#31
lexluthermiester
Jack1nJust had a 2TB drive from Seagate fail on me without warning. I doubt I will buy another mechanical drive ever again and if I do, it will not be from Seagate.
Jack1nIt is not just my own personal experience, it well known that Seagate drives have a high failure rate.
R-T-BI've never seen stats to validate that very popular forum rumor, but I have seen stats that sort of go against it, ironically.
Oh? Let's review..
www.backblaze.com/blog/backblaze-hard-drive-stats-q1-2021/
www.backblaze.com/blog/backblaze-hard-drive-stats-for-2020/
Based on those numbers, the most reliable drives made are from HGST and Toshiba. Seagate doesn't fair so well. Especially when you take a closer look at the numbers.

Jack1n's concerns about Seagate are not outside the realm of reality. His concerns about mechanical HDD's however are somewhat flawed. SSD's do fail and they are only marginally more reliable than HDD's.

www.backblaze.com/blog/how-reliable-are-ssds/

Seagate drives do generally tend to fail more often, but in context, we're taking about a difference of .45% failure rate when compared to other brands. Still, that is not an inconsiderable number given the the overall failure rate for Seagate drives is just below 2% while everyone else in the comparison is below 1%(using the annualized failure rate, not just last year).

However, in the scope of the great scheme of things, we're talking about failure rates that fall BELOW 2% on average. Folks, buy the drive that fits your needs and rest assured it is VERY likely to last you a long time. If you're worried about failures(and they do happen) invest in backups(spare drives, external drives, bluray recordables, etc..).
Posted on Reply
#32
TheUn4seen
StaticVapourNot a fan of drive war, but funnily enough I've been waiting for this drive to finally bite the dust so I could replace it with SSD.. I'm still waiting...

1094 days and 19 hours the time of screenshot, it has some trouble with error rates, but for some reason it's not that much louder than it used to be
That's the lottery. Back in 2009 I bought two 500GB WD greens. The same batch, bought together in the same shop. One failed after ~180h, the other, well,
Posted on Reply
#33
newtekie1
Semi-Retired Folder
lexluthermiesterSeagate drives do generally tend to fail more often, but in context, we're taking about a difference of .45% failure rate when compared to other brands. Still, that is not an inconsiderable number given the the overall failure rate for Seagate drives is just below 2% while everyone else in the comparison is below 1%(using the annualized failure rate, not just last year).
And all of that means nothing because the Backblaze failure rates don't apply to consumers using drives in consumer settings. And they also admitted to just leaving certain models entirely out of their stats because every single drive failed.
Posted on Reply
#34
Shrek
lexluthermiesterSeagate drives do generally tend to fail more often
I like Western Digital but wanted a hybrid drive; the solid state cache should mean less strain on the mechanics, so I hope the SeaGate Firecuda's are reliable.
Posted on Reply
#35
lexluthermiester
Andy ShiekhI like Western Digital but wanted a hybrid drive; the solid state cache should mean less strain on the mechanics, so I hope the SeaGate Firecuda's are reliable.
I would love to see some solid stats on those drives too! I sell quite a few of them because they are a great balence between the speed of SSDs and the huge storage of HDDs at a cost that is affordable. I've sold a few hundred of them and not one has come back but that is an admittedly small sample size.

I would love to see Seagate(or anyone else) make a high capacity HDD(12TB+) 7200RPM drive with 32GB or 64GB of MLC NAND cache on it and market it to the enthusiast/prosumer.

The above 20TB PMR drive with 128GB MLC(not TLC) NAND cache for around $500 or so? Yes please!
Posted on Reply
#36
Shrek
I have always wondered if 8GB is enough cache; Windows is 4GB and not all of it gets used all the time. I have 3 such drives (3.5", 2TB) and so far they have all been great; I assumed the 2.5" versions are shingled which would make them work harder.

But given that solid state drives tend to just suddenly stop working I am wondering if the FireCuda has the ability to override the cache should it fail.
Posted on Reply
#37
dragontamer5788
lexluthermiesterI would love to see some solid stats on those drives too! I sell quite a few of them because they are a great balence between the speed of SSDs and the huge storage of HDDs at a cost that is affordable. I've sold a few hundred of them and not one has come back but that is an admittedly small sample size.

I would love to see Seagate(or anyone else) make a high capacity HDD(12TB+) 7200RPM drive with 32GB or 64GB of MLC NAND cache on it and market it to the enthusiast/prosumer.

The above 20TB PMR drive with 128GB MLC(not TLC) NAND cache for around $500 or so? Yes please!
Unfortunately, higher-level filesystems like ZFS need more than just a portion of NAND cache to be effective.

ZFS has multiple caching layers: ARC (RAM/DDR4), L2ARC (SSD-read cache), and SLOG (SSD-write cache). Having the software handle the caching details, with a more complicated machine (IE: using PCIe 4.0 SSDs as your L2ARC / SLOG caches) probably is beneficial.
Posted on Reply
#38
lexluthermiester
dragontamer5788Unfortunately, higher-level filesystems like ZFS need more than just a portion of NAND cache to be effective.
You are thinking of the DRAM cache. On hybrid drives the SSD cache is invisible to the OS and file system. Only the drive controller knows the NAND is present and it can not be directly accessed by the user or system as it sits behind the drive controller. This is by design to eliminate compatibility problems like the one you described.
Posted on Reply
#39
Shrek
Where does Windows stand on booting from a ZFS drive?

I may be way behind the times, so go easy on me.
Posted on Reply
#41
dragontamer5788
Andy ShiekhWhere does Windows stand on booting from a ZFS drive?

I may be way behind the times, so go easy on me.
It can't.

ZFS is pretty good on a NAS: you make a 2nd computer that does everything through ZFS. You access it through Windows by using various technologies: different tech has different advantages and disadvantages. iSCSI and SAMBA are probably my two favorite.

SAMBA is a Windows-share, so the Linux-ZFS storage looks a lot like what Windows expects. However, Windows-shares have peculiar attributes that make them unable to do some common tasks (ex: use as a steam library).

iSCSI makes the remote drives look like a hard drive, except over an ethernet port. As far as Windows is concerned, any iSCSI drive is just another hard drive. However, iSCSI on a NAS looks like a file with no way to "peak" inside of it. SAMBA-shares you can read/write the data on other machines (ex: use your laptop to browse your SAMBA-share). But an iSCSI can only be read/written to from one computer at a time.

-------

So the ZFS plan is: setup your FreeBSD install (probably XigmaNAS) and setup ZFS. Create a iSCSI file that's like 4TB or whatever, then have Windows read/write to that iSCSI target. Windows thinks its reading/writing to a real hard drive, but in actuality, its all behind that fancy ZFS.
Posted on Reply
#42
Shrek
lexluthermiesterdocs.oracle.com/cd/E19253-01/819-5461/ggpco/index.html

That is the only article I could find, but I think there is no OS level support for ZFS for Windows.
I think it takes more than OS support to boot as the OS it not yet loaded, but I imagined some sort of trick where one boots off a small GPT partition before moving across.
Posted on Reply
#43
lexluthermiester
Andy ShiekhI think it takes more than OS support to boot as the OS it not yet loaded, but I imagined some sort of trick where one boots off a small GPT partition before moving across.
Honestly don't think ZFS was engineered to be bootable.

However, we are again off-topic...
Posted on Reply
#44
junglist724
Jack1nIt is not just my own personal experience, it well known that Seagate drives have a high failure rate.
There was a couple years where their consumer models had high failure rates. I saw 50 out of 50 HP desktops at my previous workplace with Seagate 1TB drives ready to fail after only 2 years of use.

On the other hand I have 14 x 8TB Seagate enterprise drives in RAID 10 that have been running 24/7 for over 2 years with not a single failure or pre-failure warning.
Posted on Reply
#45
windwhirl
lexluthermiesterHowever, in the scope of the great scheme of things, we're talking about failure rates that fall BELOW 2% on average. Folks, buy the drive that fits your needs and rest assured it is VERY likely to last you a long time. If you're worried about failures(and they do happen) invest in backups(spare drives, external drives, bluray recordables, etc..).
This. The stats are fine and all, but this is the important thing. Buy what you need (and please do all the necessary research, don't complain about getting a NAS drive and finding it loud as hell) and make your regular backups. Preferentialy, with an off-site extra backup if possible.
Posted on Reply
#46
InVasMani
My only question is where are the 1GB DRAM 2TB to 4TB HDD's with NVME!!? Hell they could put a HBM chip on a 4TB HDD as a cache buffer with NVME and it would make a hell of a block level cache for it. I just don't get it honestly. On a different note M.2 NVME's could easily be done with DDR/HBM and maybe include micro SD backup card and be quicker optane and yes volatile or partially so, but it comes with added security and these days maybe a good thing. I'd kind of like to see a DRAM M.2 with infinity cache even for that matter. How would that be with direct storage?
Posted on Reply
#48
BluesFanUK
There's a huge market out there for SSD manufacturers to just produce high capacity models whilst sacrificing speed. The sooner mechanical drives go the better, not everyone needs a super duper fast NVMe drive...

There's been a serious poor development in this area since those drives became the norm, continually breaking new barriers in read and write speeds but zero thoughts given to consumers who just want a large capacity drive.

It's become even more important with 4K/8K media and games that exceed 100GB.
Posted on Reply
#49
Mr Bill
Jack1nJust had a 2TB drive from Seagate fail on me without warning. I doubt I will buy another mechanical drive ever again and if I do, it will not be from Seagate.
I think in my past 25 years I've had about every mfg. of hard drive fail, I've had 2 Hitachi's deskstar's fail, but IMO, Seagate is one of the best, and I've had to only warranty one, and it was a quick turn around. All hard drives are subject to fail, the best advise I can give you is, get Acronis and back them up regularly, this is what I do. We also have to take into consideration the amount of drives sold, compared to the failures, my guess is Seagate is probably one of the highest sellers over the many years. I've heard Seagate is the largest manufacture of Hard Disk Drives, but I'm guessing WD and Seagate are probably running neck to neck. :D
Posted on Reply
#50
newtekie1
Semi-Retired Folder
BluesFanUKThere's been a serious poor development in this area since those drives became the norm, continually breaking new barriers in read and write speeds but zero thoughts given to consumers who just want a large capacity drive.
There has been advancement in the size while sacrificing speed, it's called QLC. The problem is the write speeds for QLC have now reached HDD levels and everybody wants to whine and cry about it.
Posted on Reply
Add your own comment
Jan 23rd, 2025 21:53 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts