My SMR drive - opinions please

[Chrispy_]

Anyone want to argue with this?

The only person who wants to argue is you.

There's no shortage of forums/redit/review evidence highlighting the shortcomings of SMR, and the multiple successful class action lawsuits against HDD manufacturers for failing to disclose the performance issues of SMR are testament to the fact that SMR has issues. How SMR works is no mystery; It is unable to write a small block of data without erasing and re-writing all the other blocks shingled together in the band of shingled tracks. Write amplification goes from 1x to as much as 64x depending on the workload and for a drive that has a maximum write rate of ~200MB/s, reducing that rate by up to 64x is obviously detrimental to performance in certain situations.

My guess that CDM's test size was the reason clearly wasn't right but I'm not interested in exactly what causes synthetic tests to inaccurately reflect real-world usage. That's why they're called synthetic tests rather than real-world tests. It's not my job to convince you that SMR has pitfalls. The HDD industry and international legal systems have extensively investigated and concluded the disadvantages of SMR. Like all technological compromises, there's nothing wrong with it as long as the compromises are clearly labelled and people can make up their own mind about whether the downsides are worth the decrease in cost/TB.

 

[lexluthermiester]

The only person who wants to argue is you.

Not true, the OP wanted input and there has been plenty offering insight.

There's no shortage of forums/redit/review evidence highlighting the shortcomings of SMR

And most of those are dominated by angry nitwits special-snowflaking over what is effectively minor differences in performance. My CMR drive gets about 200MB per second as opposed to the SMR drive at 175MB per second. This is NOT an isolated experience.

Write amplification goes from 1x to as much as 64x depending on the workload and for a drive that has a maximum write rate of ~200MB/s, reducing that rate by up to 64x is obviously detrimental to performance in certain situations.

Given the testing earlier, I have some doubts about how seriously that problem actually effects performance.

My guess that CDM's test size was the reason clearly wasn't right but I'm not interested in exactly what causes synthetic tests to inaccurately reflect real-world usage. That's why they're called synthetic tests rather than real-world tests. It's not my job to convince you that SMR has pitfalls. The HDD industry and international legal systems have extensively investigated and concluded the disadvantages of SMR. Like all technological compromises, there's nothing wrong with it as long as the compromises are clearly labelled and people can make up their own mind about whether the downsides are worth the decrease in cost/TB.

And the ATTO tests? These utilities first create a test file as a part of the testing. Then they use that file to subject the drive to a battery of workloads that closely mimic real world usage. They are "synthetic" only in the fact that they are deliberately inducing mimicking workloads. The work being done by the drive is still "real" in the fact that the drive is performing functions it would normally conduct under non-testing use. The term "synthetic" is NOT a synonym for "fake". To deny this is pure ignorance as demonstrated by the users who voted "Chuck it into the nearest furnace to melt this garbage down".

My drive is subjected to frequent read/write ops that weigh in the hundreds of GB. I've never noticed it slow down or even show signs of doing so.

The proof is in the pudding, and the tests show solid performance, within 12% of their CMR companion drives under ALL testing workload methods.

 

[ThrashZone]

Indeed, it's much higher performance.

Hi,
There are 7200rpm blue's 40.us for 1tb lol only one I have is a 5400 though just for system image storage

https://www.amazon.com/WD-Blue-1TB-Hard-Drive/dp/B0088PUEPK

 

[newtekie1]

While true, it would only slow things down somewhat. It would never cause data corruption problems. The difference is in speed only, not data integrity.

They never said anything about data corruption. The question asked was specifically about performance and the slow down making them less useful for non-incremental backups. Seriously, stop going off on tangents that have nothing to do with what people are discussing.

The proof is in the pudding, and the tests show solid performance, within 12% of their CMR companion drives under ALL testing workload methods.

Fill the drive to 90% then turn it again.

 

[qubit]

Hi,
There are 7200rpm blue's 40.us for 1tb lol only one I have is a 5400 though just for system image storage

https://www.amazon.com/WD-Blue-1TB-Hard-Drive/dp/B0088PUEPK

lol so it is as are a few others up to 2TB. I don't think that there were any 7200 RPM Blues when I bought mine. My one is 5400 RPM and I see that the newer 8TB model is a very odd 5640 RPM. I'd like to know why they made it that speed.

Specs of the whole Blue range are here:

https://www.westerndigital.com/en-gb/products/internal-drives/wd-blue-desktop-sata-hdd

EDIT: I see that the 8TB WD80EAZZ drive (CMR) is about £200 from Scan, so I might treat myself to it. At the time, I wanted an 8TB Blue, but they didn't make them. It would be a pure vanity purchase though lol as I haven't bought an upgrade for my PC in ages. I hope it's suitably quiet though...

Anyway, my 6TB Blue is now over 2 years old and the warranty's expired, so isn't it about time that it failed?

 

[lexluthermiester]

They never said anything about data corruption.

No, but the implication was hinted at, thus the statement to shut that thought down.

Fill the drive to 90% then turn it again.

Perhaps you didn't notice the testing screenshots. It's one of my backup drives which is consistently near full. And the tests above were conducted with the drive at 87% usage. Sooo... your point?

 

[newtekie1]

No, but the implication was hinted at, thus the statement to shut that thought down.

No, it wasn't in any way.

Perhaps you didn't notice the testing screenshots. It's one of my backup drives which is consistently near full. And the tests above were conducted with the drive at 87% usage. Sooo... your point?

You showed a partition on a drive that is 87% full, not that the drive was 87% full. And the Easystore drives are not all using SMR. So you really showed nothing.

 

[Mussels]

Hey lex, can you verify the actual mechanical disk in your external enclosure?
You're making claims about its performance but you didnt seem to actually specify what it is, only that it's a WD drive in an external enclosure

It would be quite a facepalm moment if it turned out to not be the drive you expect, i've had quite a few WD externals throw random disks at me (I bought three 4TB in one transaction, got two greens and a red when i opened them)

 

[lexluthermiester]

No, it wasn't in any way.

That depends on how you read the commentary. Not going argue about it.

You showed a partition on a drive that is 87% full, not that the drive was 87% full. And the Easystore drives are not using SMR. So you really showed nothing.

The other partitions are just as full.

Hey lex, can you verify the actual mechanical disk in your external enclosure?

Looked it up when the SMR drama happened. The drive inside is a white label and when I looked it up the specs said SMR. Not opening it up again right now.

 

[qubit]

Oh sod it, I've bought it! I found myself just wanting the 8TB CMR/PMR version (WD80EAZZ) - enthusiastconomics style lol - as it was what I wanted originally, but didn't exist then, so I've just bought it from Scan and is due tomorrow.

I'll run some comparative benchies and post them here. The differences should be interesting. I'll figure out what I wanna do with the 6TB drive afterwards as it's still in perfect working order. Might eBay it perhaps, we'll see,

https://www.scan.co.uk/products/8tb-wd-blue-wd80eazz-35-hdd-sata-iii-6gb-s-5640rpm-128mb-cache-oem

@Pixel Princess you might want to bookmark this thread.

 

[Pixel Princess]

Oh sod it, I've bought it! I found myself just wanting the 8TB CMR/PMR version (WD80EAZZ) - enthusiastconomics style lol - as it was what I wanted originally, but didn't exist then, so I've just bought it from Scan and is due tomorrow.

I'll run some comparative benchies and post them here. The differences should be interesting. I'll figure out what I wanna do with the 6TB drive afterwards as it's still in perfect working order. Might eBay it perhaps, we'll see,

https://www.scan.co.uk/products/8tb-wd-blue-wd80eazz-35-hdd-sata-iii-6gb-s-5640rpm-128mb-cache-oem

@Pixel Princess you might want to bookmark this thread.

Sooooo.....what are you going to do with 8TB @qubit ?

@qubit I have bookmarked this thread. I'm still reading all of the amazing info shared

 

[lexluthermiester]

Sooooo.....what are you going to do with 8TB @qubit ?

Game installs and backups would be my guess.

 

[qubit]

@Pixel Princess @lexluthermiester

Yup, lex isn't far off. I'm just gonna transfer my 3.24TB of game installs over the weekend to the new drive. That and a few scratchpad items are all that's on it. Note that the copy operation will take more than 24 hours, so I want to do it when I'm not working.

I'll then wipe the old drive and probably eBay it, or give it to a friend maybe.

 

[Pixel Princess]

@lexluthermiester right of course!! (Clever username)

@qubit I had an additional spirited conversation with my daughter this morning concerning saving gaming installs vs. trusting the Steam and/or Origin cloud. She said the main game she'd worry about loosing that she'd consider saving is Valheim.

 

[qubit]

@qubit I had an additional spirited conversation with my daughter this morning concerning saving gaming installs vs. trusting the Steam and/or Origin cloud. She said the main game she'd worry about loosing that she'd consider saving is Valheim.

Personally, I've never lost information to Steam so I don't have any reason to distrust them. Seems other people have though.

 

[Pixel Princess]

Personally, I've never lost information to Steam so I don't have any reason to distrust them. Seems other people have though.

Good to know and I'm sure Steam is exemplary in their member data storage. She just has built a freaking kingdom in Valheim and I can't even imagine the drama if she lost it.

 

[qubit]

Right, so I've just broken the antistatic bag seal, connected the drive to the PC and... woah! This thing vibrates! The platter is totally unbalanced. The vibration is strong when held in the hand and when it touches the table or the PC case, it acts like a sounding board, making for a very unpleasant 94Hz humming noise.

This isn't normal, so has to be either a manufacturing fault, or a design fault, therefore, I've already applied for a return, including shipping, as a faulty product. Note that this strong vibration wasn't disclosed in the product description as I wouldn't have bought it if it had been, so I have a very strong case.

I've never seen anything like it. I've had drives that vibrate a bit, but nothing like this, ever. With most drives, including the ones installed in my PC, the vibration can hardly be felt and that's after years of use too, including the 6TB Blue, and the bearings are quiet too, making for a very quiet environment, which I require. Vibration this bad can't be good for long term reliability, either.

Clearly this thing is built right down to a price and this is one of the things they've compromised on. I wonder if other samples are this bad and suspect they are. I'm not gonna go through the hassle of buying another one just to find out that it's the same and have to return that too. I might buy an 8TB Black at £280, but not right now as this really wasn't an essential purchase and that's a lot of money.

Because of this, I've not bothered formatting it or anything. I just unplugged it and put it to one side.

 

[claes]

I’d avoid WD black if you’re worried about noise/vibrations, so no loss there

 

[qubit]

I’d avoid WD black if you’re worried about noise/vibrations, so no loss there

I've actually had several Blacks over the years and they've all been fine for vibration. Bearing and head access noise can be a bit louder on the older ones, but nothing too bad. In fact, I've got a 4TB Black in there right now, doing great for years. Hence, I have confidence that the next Black I buy will be fine.


EDIT

This was supposed to be a separate post about 12 hours later, but the post merger add-on snaffled it instead.

I've had the RMA number from Scan now - fast service. Glad to say that this shop normally has great customer service.

 

[Chrispy_]

I've actually had several Blacks over the years and they've all been fine for vibration. Bearing and head access noise can be a bit louder on the older ones, but nothing too bad. In fact, I've got a 4TB Black in there right now, doing great for years. Hence, I have confidence that the next Black I buy will be fine.

It's been a while since I bought any serious quantities of consumer-grade spinning rust, but my experience back then across Samsung, Hitachi, WD, Seagate, & Toshiba is that the tier of disk from budget to flaghship made no real difference to the smoothness and balance of the platters. I had cheapo Toshiba drives and OEM Baracudas that were balanced perfectly with near-silent bearings, and WD Gold/Black that sounded rough when new and stayed like that. Drives I bought a lot of were Samsung Spinpoints, WD Blue/Green/Red, and early Seagate Ironwolves back from the times when 4TB was a "big" drive.

Number of platters (capacity) seems to have far more impact on noise levels than anything else. Spin speed only really affects the pitch of the bearing whine.

 

[qubit]

@Chrispy_ While they do vary, I wonder if this drive has simply been damaged in transit, knocking the platter balance out of alignment. At 5640 rpm, it wouldn't take much for it to vibrate like that.

The most balanced drive I've ever had is the WD Raptor X 150GB with the clear window allowing the platter and head to be viewed. Spinning at a whopping 10000 rpm, there's no vibration at all!

It's a helluva drive for its time, although the head seeks we're extremely loud and annoying.

While I retired it years ago, it's still in perfect working order and in perfect condition. I'm keeping it as a collectors item.

 

[Chrispy_]

The most balanced drive I've ever had is the WD Raptor X 150GB with the clear window allowing the platter and head to be viewed. Spinning at a whopping 10000 rpm, there's no vibration at all!

It's a helluva drive for its time, although the head seeks we're extremely loud and annoying.

Before the X25-M I bought, SSDs were too small and too expensive so I "borrowed" (permanently) a couple of 15K Seagate Cheetahs and SAS RAID card for a RAID0 OS+Apps drive from the datacenter. Even those could not make the original release of Windows Vista feel fast, but man were they awesome disks for their day.

Compared to a 7200rpm consumer desktop drive of the time they were truly insane but I donated my stolen goods almost immediately after buying myself an 80GB Intel SSD and never looked back. I do not miss the horrible whine of 15K drives or the jackhammer of "performance above all else" head seeking but I do have a soft spot for them.

 

[lexluthermiester]

I’d avoid WD black if you’re worried about noise/vibrations, so no loss there

Your loss. WD Black drives are among the best made. Full stop.

 

[InVasMani]

The only reason I still keep a mechanical drive in my system is for the added sound acoustics because floppy disks are harder to come by.

 

[lexluthermiester]

The only reason I still keep a mechanical drive in my system is for the added sound acoustics because floppy disks are harder to come by.

That's some hard-core nostalgia going on!