Western Digital is trying to redefine the word “RPM”

Why would merchants accept this?
They often copy spec sheets, and are just as likely to transcribe "5400 RPM class" as "5400 RPM" as consumers would be.

If a consumer bough one from a deceived merchant, because they had gone by the merchant's published numbers then that could be a cause for a full refund, but it is unlikely that WD would cover the merchant's cost.

The merchants don't understand, and assume the manufacturer does.
 
Upvote
9 (9 / 0)

jnv11

Ars Scholae Palatinae
686
I do not like any of the HDD manufacturers that much. Western Digital keeps getting caught with false or misleading ads. Seagate moved its legal domicile to Dublin, Ireland while keeping its operational headquarters to use a legal loophole to essentially cheat on its taxes. Toshiba has the Toshiba-Kongsberg scandal as seen in https://en.wikipedia.org/wiki/Toshiba-Kongsberg_scandal in its history, and does not offer any models with 5 year warranties unlike the high end consumer drives from Seagate and Western Digital.
 
Upvote
4 (7 / -3)

zogus

Ars Tribunus Angusticlavius
7,269
Enough with the loot box mechanics! Just label the actual contents of the package!
Okay, this got to me. "Just pay us $120, and you'll get a brown, unlabeled box which might contain a sticker, a LED keyboard, an SSD with a capacity between 64gb-4tb and may use SATA or PCIe, or a HD with a capacity of 256gb-14tb may have cache, use 5-50 watts of power, have a spindle speed of 2,500-10,000rpm, something weird we found in the lab, or farts. Distributed randomly. Good luck!"
I for one am NOT going to blow $120 on such a box, unless it may also possibly contain a playable character wielding a 7200rpm Western Digital chainsaw!
 
Upvote
5 (6 / -1)

AmanoJyaku

Ars Tribunus Angusticlavius
16,197
I do not like any of the HDD manufacturers that much. Western Digital keeps getting caught with false or misleading ads. Seagate moved its legal domicile to Dublin, Ireland while keeping its operational headquarters to use a legal loophole to essentially cheat on its taxes. Toshiba has the Toshiba-Kongsberg scandal as seen in https://en.wikipedia.org/wiki/Toshiba-Kongsberg_scandal in its history, and does not offer any models with 5 year warranties unlike the high end consumer drives from Seagate and Western Digital.
Yet another reminder that no business is your friend. If they do something that benefits you, it's likely a side effect of something that benefits them. And their practices may benefit you today, but can easily switch to something harmful tomorrow.
 
Upvote
8 (11 / -3)
Post content hidden for low score. Show…

Wolvenmoon

Ars Tribunus Militum
1,691
Holy crap, I'm glad I didn't buy into WD products for my homelab. I'm operating on a very tight power budget because my workstation and 40-core 676 gigs of RAM+72+29 3.5"-bay + a few USB disk homelab are on the same 15A circuit.

I was looking at 5400RPM spinners specifically to conserve power, reduce heat dissipation, and manage noise.
 
Upvote
3 (4 / -1)
Post content hidden for low score. Show…
Post content hidden for low score. Show…

unconcerned

Ars Scholae Palatinae
1,071
If you cannot tell the truth on a spec sheet I see no need to even consider your products. Accurate labeling is not optional.
I think it's also not legal. Anyone want to file a class action lawsuit and get $5?
AFAIK no law requires that a HD manufacturer provide any particular set of specifications. In this case WD appears to be truthful with what they say, in the sense that it's *at least as good* as the "class" specification. It's not uncommon to see things overprovisioned to make sure they meet a spec. As the article notes, they haven't said what they mean by "5400rpm class." So 1) are they lying? and 2) does it matter?

It would still be good if they were transparent about what's in the box and how it performs. Not being so enhances suspicion, considering other recent faux pas. It's interesting. But I wonder if there's a better hill to die on.

Here unlike SMR debacle I would say they don't mean bad. From performance point of view 7.2 k rom is better than 5400. However, that changes the power drawn, hear dissipation and probably the disk failure distribution. So not an intentional deception but rather they wanted to have the freedom to swap 5400 and 7200 rpm drivers depending on their supply chain. Still, I would be pissed if the performance of the better same drive model changed from batch to batch
 
Upvote
8 (10 / -2)

real mikeb_60

Ars Legatus Legionis
13,170
Subscriptor
This is a storm in a tea cup. WD isn't listing the rotational speed as 5400 RPM, they're listing the "performance class" of the drive as "5400 RPM-class". Frankly, the word "class" alone appended "5400 RPM" should be enough to tip folks off that what is being described is not the actual spindle speed, but even if not -- they never claimed it was!

I look forward to Ars doing a deep dive into motorcycle manufacturers and their "deceptive practice" of using model numbers that do not actually align with the engine capacity in cubic centimeters, even though those manufacturers never claimed that they did.
How about several carmakers (Germans seem to have invented this) that use model numbers that once indicated the engine displacement. Of a normally aspirated engine. When they started using little turbo 4 cylinders, their model numbers seem to have morphed to providing some kind of equivalent normally aspirated engine size that would theoretically produce the HP claimed from the turbo. The turbo, unfortunately, doesn't work the same way so while the peak numbers may be equivalent (or even better) the driving experience can suffer.
 
Upvote
-7 (5 / -12)
If you'll pardon an automotive analogy, this doesn't strike us as much different than a window sticker on a V6 sedan which declares it to be "four cylinder class," with no indication of what's actually under the hood.
With the V6 detuned to produce the limited horsepower and torque of a 4-cylinder.
And the fuel consumption of the V6.
 
Upvote
16 (16 / 0)

sarusa

Ars Praefectus
3,287
Subscriptor++
Here unlike SMR debacle I would say they don't mean bad. From performance point of view 7.2 k rom is better than 5400. However, that changes the power drawn, hear dissipation and probably the disk failure distribution. So not an intentional deception but rather they wanted to have the freedom to swap 5400 and 7200 rpm drivers depending on their supply chain. Still, I would be pissed if the performance of the better same drive model changed from batch to batch
I think in this case these 7200 drives have the same performance as a 5400 because they are slowing down the writes on failed 12TB drives by 33% to get marginal 8TB capacity drives they can sell instead of throwing them away.

So you are getting drives that are as fast as 5400 rpm (hence the '5400 rpm class') while still running noisier, hotter, and with more power consumption. And then they're having the drive lie about how fast it's spinning. This is terrible if you think you're actually buying a 5400 rpm drive. I think it is rather malicious.
 
Upvote
26 (26 / 0)

Jim Salter

Ars Legatus Legionis
17,287
Subscriptor++
Why the fuck is it so hard for HDD companies recently to just tell us what we're buying?

There is no world were a "5400 RPM" class means anything other than a ~5400 RPM drive. Plus or minus a few hundred RPM is one thing, but what's next, "10 Watt" class drives that pull 15?

"What are you gonna do? Go buy some 8 TB SSDs? Eat shit and thank us for it, fuckface."

Not sure if it's the contempt that comes with companies being one of an exceedingly few number of competitors in a given industry, the fact that their days are numbered as SSDs continue to get cheaper per TB or both.

I hate to rain on your parade, here, but Western Digital is absolutely an SSD player.

For example: WD Blue SSDs. Local shops in my area have stocked these for a few years now, and are using them pretty frequently in their default builds.
 
Upvote
18 (18 / 0)

AmanoJyaku

Ars Tribunus Angusticlavius
16,197
If you cannot tell the truth on a spec sheet I see no need to even consider your products. Accurate labeling is not optional.
I think it's also not legal. Anyone want to file a class action lawsuit and get $5?
AFAIK no law requires that a HD manufacturer provide any particular set of specifications. In this case WD appears to be truthful with what they say, in the sense that it's *at least as good* as the "class" specification. It's not uncommon to see things overprovisioned to make sure they meet a spec. As the article notes, they haven't said what they mean by "5400rpm class." So 1) are they lying? and 2) does it matter?

It would still be good if they were transparent about what's in the box and how it performs. Not being so enhances suspicion, considering other recent faux pas. It's interesting. But I wonder if there's a better hill to die on.
I would argue that it falls under deceptive advertising. If I wanted a 5,400RPM drive, I wouldn't buy this one after comparing the higher power consumption to real 5,400RPM drives. But, if I did I think I'd be pissed if I wanted a quieter, cooler drive and discovered this isn't it. So:

1) Yes, they are lying
2) Yes, it does if you are impacted by the differences in characteristics

I was being tongue-in-cheek about the class action lawsuit. Poe's law and all, but I thought the $5 gave it away.

That being said, it's not my hill to die on because I don't purchase hard disks anymore. SSD or bust.
 
Upvote
15 (16 / -1)

Jim Salter

Ars Legatus Legionis
17,287
Subscriptor++
What benefit is there from labelling a higher rpm drive as lower rpm from a marketing perspective? Unless you think you'll trick power conscious buyers into going a 7200 drive because they don't look at power specs.

My guess is that power-conscious consumers wanting 5400rpm drives hadn't occurred to them, and this was more a case of trying to preserve the market segmentation associated with higher spindle speeds, without actually having to maintain two separate product lines.

This is completely speculation here, but imagine: "We want people to really, really know these are the cheap drives, not the fast ones. But we're putting 7200rpm motors in them now because it's not economically feasible to maintain two separate product lines."
 
Upvote
17 (17 / 0)

Jim Salter

Ars Legatus Legionis
17,287
Subscriptor++
What *are* the best NAS drives if WD Red/Red Pro aren't an option? I've got some 7 year old WD Reds in a ZFS NAS. At some point here soon I might need to replace them.
Seagate IronWolf is the usual recommendation.


I've had several opportunities but passed on WD since their bullshit with SMR drives. Apparently they don't want to earn my business back.

Seagate Exos is far and away better, unless you're desperate for the "File Recovery" bundle that you get with Ironwolf drives. Anandtech (link above) did an excellent and rather damning overview of consumer & NAS drives.

A word of warning: don't buy Exos (or Ultrastar) unless you're willing to read the datasheet and do so thoroughly. Exos—or WD's Ultrastar—can mean CMR, or it can mean host-managed SMR, or it can mean basically anything else, because those lines are intended for sale to enterprise/datacenter customers who absolutely do read the datasheet. The whole datasheet.

If you buy an Exos or Ultrastar and aren't careful, you could end up with a host-managed SMR disk that doesn't even work as a normal drive (instead needing to be managed with, eg, Linux's dm-zoned).
 
Upvote
36 (36 / 0)

AmanoJyaku

Ars Tribunus Angusticlavius
16,197
SMR and now this. Is there anything on WD's spec sheet which isn't a pack of lies?
Soon people will complain that their 8tb drive isn't actually 8tb...
Didn’t this happen about 15 years ago? It’s actually 8 tibbibytes or something.
HDDs are 8 terabytes, which is 8,000,000,000,000 bytes. In other words, powers of 10.

An SSD would be 8 tebibytes, which is 8,796,093,022,208 bytes. In other words, powers of 2.
 
Upvote
-17 (2 / -19)

raxx7

Ars Legatus Legionis
17,132
Subscriptor++
SMR and now this. Is there anything on WD's spec sheet which isn't a pack of lies?
Soon people will complain that their 8tb drive isn't actually 8tb...
Didn’t this happen about 15 years ago? It’s actually 8 tibbibytes or something.
HDDs are 8 terabytes, which is 8,000,000,000,000 bytes. In other words, powers of 10.

An SSD would be 8 tebibytes, which is 8,796,093,022,208 bytes. In other words, powers of 2.

A 512 GB SSD is 512,000,000,000,000 bytes.
 
Upvote
13 (14 / -1)

Andrewcw

Ars Legatus Legionis
19,046
Subscriptor
SMR and now this. Is there anything on WD's spec sheet which isn't a pack of lies?
Soon people will complain that their 8tb drive isn't actually 8tb...

They will bring this concept to the SSD market. Soon with Hardware compression! The 6TB drive! will hold billions of pictures or 2TB of video!. /s
 
Upvote
5 (5 / 0)
Upvote
1 (1 / 0)

Eldorito

Ars Tribunus Angusticlavius
8,012
This is a storm in a tea cup. WD isn't listing the rotational speed as 5400 RPM, they're listing the "performance class" of the drive as "5400 RPM-class". Frankly, the word "class" appended to "5400 RPM" should be enough to tip folks off that what is being described is not the actual spindle speed, but even if not, they explicitly claimed it to be something else entirely!

I look forward to Ars doing a deep dive into motorcycle manufacturers and their "deceptive practice" of using model numbers that do not actually align with the engine capacity in cubic centimeters, even though those manufacturers never claimed that they did.

Ignoring that WD completely made up that class and, like usual, don't explain anywhere what it means, what attributes of a 5400rpm drive does this have?

Also, what motorbike is named 750CC-class when it's really a 500CC? Model numbers are entirely different to spec sheets to, I don't know of any motorcycle manufacturer lying in their spec sheets.
 
Upvote
11 (12 / -1)

Jim Salter

Ars Legatus Legionis
17,287
Subscriptor++
SMR and now this. Is there anything on WD's spec sheet which isn't a pack of lies?
Soon people will complain that their 8tb drive isn't actually 8tb...
Didn’t this happen about 15 years ago? It’s actually 8 tibbibytes or something.
HDDs are 8 terabytes, which is 8,000,000,000,000 bytes. In other words, powers of 10.

An SSD would be 8 tebibytes, which is 8,796,093,022,208 bytes. In other words, powers of 2.

A 512 GB SSD is 512,000,000,000,000 bytes.

Roughly 512*10^9 bytes. But not exactly 512*10^9 bytes, because the page size is going to be a power of two, and the total device size is an even multiple of the page size, and 10^9 isn't a power of two.

I'm never going to stop being mad about either 1. us nerds way back in the day being sloppy with our language and just using powers-of-ten prefixes for powers-of-two units, or 2. the damn marketers for exploiting that loophole to allow them to sell devices with less capacity than any reasonable person would think they're being advertised with.

If we'd been less sloppy, the damn marketers wouldn't have had a loophole to exploit. If the marketers weren't, you know, marketers, they wouldn't have exploited the loophole because seriously how do you fucking sleep at night when you're that scummy? But both conditions were true, and now we have drives sold in different units than they actually function upon, so yay for that.

And no, Apple, just making Finder start reporting in SI prefixes (despite the fact that your filesystem is still using powers-of-two sector sizes) doesn't "fix it."
 
Upvote
27 (30 / -3)

sswilson

Ars Centurion
362
Subscriptor
This is a storm in a tea cup. WD isn't listing the rotational speed as 5400 RPM, they're listing the "performance class" of the drive as "5400 RPM-class". Frankly, the word "class" appended to "5400 RPM" should be enough to tip folks off that what is being described is not the actual spindle speed, but even if not, they explicitly claimed it to be something else entirely!

I look forward to Ars doing a deep dive into motorcycle manufacturers and their "deceptive practice" of using model numbers that do not actually align with the engine capacity in cubic centimeters, even though those manufacturers never claimed that they did.

You're ignoring the fact that these drives report as 5400 RPM drives in the firmware. If it weren't for that they might be able to weasel out with the "class" statement... the fact that they went so far as to code the false rotational speed into the smart reporting demonstrates that their intention was to deceive.
 
Upvote
37 (37 / 0)