Western Digital Caught Bait-and-Switching Customers With Slow SSDs (extremetech.com) 92
An anonymous reader quotes a report from ExtremeTech, written by Joel Hruska: According to a report from Chinese tech site Expreview, the WD SN550 Blue -- which is currently one of the best-reviewed budget SSDs on the market -- has undergone a NAND lobotomy. While the new SSD variant performs on-par with the old drive that WD actually sampled for review, once you exhaust the SLC NAND cache, performance craters from 610MB/s to 390MB/s. The new drive offers just 64 percent of the performance of the old drive.
This is unacceptable. It is unethical for any company to sample and launch a product to strong reviews only to turn around and sell an inferior version of that hardware at a later date without changing the product SKU or telling customers that they're buying garbage. I do not use the term "garbage" lightly, but let me be clear: If you silently change the hardware components you use in a way that makes your product lose performance, and you do not disclose that information prominently to the customer (ideally through a separate SKU), you are selling garbage. There's nothing wrong with selling a slower SSD at a good price, and there's nothing right about abusing the goodwill of reviewers and enthusiasts to kick bad hardware out the door.
As a reviewer of some twenty years, I do not care at all about the fact that SLC cache performance is identical. While I didn't realize it at the time I wrote up the Crucial bait-and-switch on August 16, I've actually been affected by this problem personally. The 2TB Crucial SSD I purchased for my own video editing work is one of the bait-and-switched units, and it's always had a massive performance problem -- as soon as it empties the SLC cache, it falls to what I'd charitably call hard drive-level performance. Performance can drop as low as 60MB/s via USB3.2 (and ~150MB/s when directly connected via NVMe) and it stays there until the copy task is done. The video upscaling projects I work on regularly generate between 300-500GB of image data per episode, per encode. Achieving ideal results can require weaving the output of 3-5 models together. That means I generate up to 1.5TB of data to create a single episode. God help you if you need to copy that much information to or from one of these broken SSDs. It's not literally as bad as a spinning disk from circa 2003, but it's nowhere near acceptable performance.
This is unacceptable. It is unethical for any company to sample and launch a product to strong reviews only to turn around and sell an inferior version of that hardware at a later date without changing the product SKU or telling customers that they're buying garbage. I do not use the term "garbage" lightly, but let me be clear: If you silently change the hardware components you use in a way that makes your product lose performance, and you do not disclose that information prominently to the customer (ideally through a separate SKU), you are selling garbage. There's nothing wrong with selling a slower SSD at a good price, and there's nothing right about abusing the goodwill of reviewers and enthusiasts to kick bad hardware out the door.
As a reviewer of some twenty years, I do not care at all about the fact that SLC cache performance is identical. While I didn't realize it at the time I wrote up the Crucial bait-and-switch on August 16, I've actually been affected by this problem personally. The 2TB Crucial SSD I purchased for my own video editing work is one of the bait-and-switched units, and it's always had a massive performance problem -- as soon as it empties the SLC cache, it falls to what I'd charitably call hard drive-level performance. Performance can drop as low as 60MB/s via USB3.2 (and ~150MB/s when directly connected via NVMe) and it stays there until the copy task is done. The video upscaling projects I work on regularly generate between 300-500GB of image data per episode, per encode. Achieving ideal results can require weaving the output of 3-5 models together. That means I generate up to 1.5TB of data to create a single episode. God help you if you need to copy that much information to or from one of these broken SSDs. It's not literally as bad as a spinning disk from circa 2003, but it's nowhere near acceptable performance.
Intel & Corsair (Score:4, Informative)
haven't failed me yet.
I've been leery about WD products for a while although I have a 5TB MyBook that still hums along fine.
This is totally unacceptable and sue worthy.
Re: (Score:3, Insightful)
Re: (Score:1, Informative)
Sadly, they do:
https://www.tomshardware.com/n... [tomshardware.com]
It's right there in the URL.
Re:Intel & Corsair (Score:5, Informative)
In any case, the manufacturer has changed the packaging, part numbers and updated the product sheet for the 970 Evo Plus, giving consumers a friendly heads-up.
Re: (Score:2)
Better, but they're still dishonestly cashing in on the good name the drive had earned. If someone making a purchasing decision based on reviews and benchmarks of the 970 Evo Plus versus other options (aka being a responsible, informed, consumer), they're almost certain to end up basing their decision on reviews and analysis of the original design, since there are going to be a *lot* more reviews of a drive when it comes out than when it undergoes a revision a year or two later.
If you want to sell a new pr
Re: (Score:2)
Oh, look at that.
https://hardware.slashdot.org/... [slashdot.org]
Re:Intel & Corsair (Score:5, Informative)
And when you read the article, you notice that Samsung did nothing like the crappy thing the others did. They probably even kept performance intact and just had to do it because they could not get enough of the same controllers. They also adjusted SKU, Package design and datasheet. That is quite different.
Re: (Score:2)
It's not clear from the article that WD didn't change the part number, only that they didn't change the model number, just as Samsung doesn't.
Re: Intel & Corsair (Score:2)
And replaced a part with a slower cheaper one. Not one that is similar or faster like Samsung
Re: (Score:2)
The article at the very least strongly indicated that WD kept the SKU.
Re: (Score:3)
Sadly, they do:
https://www.tomshardware.com/n... [tomshardware.com]
It's right there in the URL.
And the tag line after the title is "Component swap done right".
So clearly the reviewer does not share your criticism.
Re:Samsung Drives (Score:2)
I had SIX 1Tb Samsung drives fail me over a 6 month period. That was it for me. Samsung is on my do not buy list for good.
Re: (Score:2)
Strange because I found the Samsung 1TBs to be very reliable, although you should know that they sold that business to Seagate.
If anything the handling of the samsung 840 series SSDs has been their biggest failing, apart from intentionally making bad cellphones for the past several years.
Re: (Score:3)
Strange because I found the Samsung 1TBs to be very reliable, although you should know that they sold that business to Seagate.
If anything the handling of the samsung 840 series SSDs has been their biggest failing, apart from intentionally making bad cellphones for the past several years.
I've killed two Samsung SSDs in my Tesla. It has constant video writing to it while driving and when sentry mode is on.
I've never had a problem with Samsung or Intel SSDs in any normal computer.
Re: (Score:3)
I'm a fan of Samsung SSDs too. They are solid and have a decent implementation of OPALv2/eDrive for self encrypting.
The only real downside is that their Samsung Magician app has got crappier over the years, but then again compared to other manufacturer's apps it's still decent.
Re: (Score:2)
I hat to disappoint you, but:
https://hardware.slashdot.org/... [slashdot.org]
Re:Intel & Corsair (Score:4, Interesting)
The Intel flash storage line is being bought by SKHynix, so, keep an eye for them in the future.
Even though corsair do not produce their own Flash chips (like Western Digital or crucial do) they are smart about their brand, and do not engage in these practices...
Re: Intel & Corsair (Score:2)
Why do you need 5 Terabytes of space? You sick fuck!
Thinking that the answer is evident in your question.
Re:Intel & Corsair (Score:4, Informative)
haven't failed me yet.
I was a solid fan of Intel also, till I found out they were fudging the benchmarks on some (rather expensive) 750 Series 1.2TB PCie SSDs I purchased. They were rated up to 460,000 IOPS (read) and 290,000 IOPS (write). These drives internally were actually like two separate SSDs, and I found out that to achieve the published benchmarks, they were restricting their testing to not cross the boundary between one internal SSD and the other, which it turned out made a big difference.
Re:Intel & Corsair (Score:4, Insightful)
Given the whole debacle of rigging the Intel compiler suite to sandbag AMD for years, that doesn't surprise me that Intel cheated.
WD spinning disks (Score:3, Insightful)
I still exclusively buy WD spinners but I wouldn't buy their SSDs, only Cosair. I had a problem with a "enterprise" Intel SSD getting slow then failing right outside the warranty period of 1 year so I don't buy them, they're overpriced anyway.
Re: (Score:2)
I still exclusively buy WD spinners.
Yep. Their black USB "game drives" are awesome value.
Re: WD spinning disks (Score:2)
Re: (Score:3)
Agreed. I bought some 10k rpm WD spinners in the year or two before SSD really took off, and the only reason they're not still being used is because the storage density isn't worth the SATA connection in the face of SSD costs these days.
I won't touch Seagate anything. For spinning drives, I've found that Toshiba is really hard to beat in NAS applications - better warranties, usually double the cache of the competition at the same or lower price point, and if you do have to warranty it, they just give you
Re: (Score:3)
> I've found that Toshiba is really hard to beat in NAS applications
--Agreed. SG Ironwolf is a good NAS drive for price, but performance-wise they slow WAAAAYY down past the ~5TB mark when doing burn-in testing.
Toshiba N300 maintains a sequential speed of over 100MB/sec for the whole test. They're my new go-to for NAS drives
Re: (Score:3)
Intel gave me the shits a few years ago when I started to notice that reviewers of CPU's would get review units, that they'd be able to overclock to some stupid numbers, which anyone buying off the shelf would have no chance of achieving. Clearly they were getting specially binned CPU's.
Unfortunately, the whole industry does this, because it's so easy for them to do some sort of misrepresentation, or bait and switch. Especially when so many people look at reviews, and use them to inform their purchasing dec
Re: Intel & Corsair (Score:2)
Add Kingston to the list. They haven't deviated much from mediocre in any direction - including not downard.
If you're looking for a reasonably-priced drice that holds pretty much what says on the box, there you have it.
Re: (Score:3)
I'd never buy another Intel SSD after their ingenious self-bricking 320-series drives, where at some random point on power-up they'd go from being an xxxGB SSD to an 8MB flash drive. Intel never fixed the problem but just quietly discontinued the line, leaving owners with an SSD that could commit hara-kiri at any point, destroying all data on it with no hope of recovery.
Sure, other companies have had problems, but AFAIK only Intel has gone out of its way to trumpet how safe and reliable their drives are a
Re: (Score:2)
Some of them deliberately self-destructed. When some of their enterprise models reach end of life they flag up a SMART warning. If they get power cycled after that they can't even be read for data recovery, they are fully bricked.
Re: (Score:2)
Intel didn't abandon 320-series users. When the problem was discovered they quickly released a critical firmware update that prevented the problem from happening.
When the update was released I distinctly remember rushing to update the firmware on around ~200 drives we were using at the time and I have never seen a drive with the firmware update fail in this way.
The only problem was you had to install the update before the drive failed. If the drive had already failed you had to send it to Intel to get it
Re: (Score:2)
Intel didn't abandon 320-series users. When the problem was discovered they quickly released a critical firmware update that
claimed to
prevent the problem from happening.
FTFY. There were still plenty of reports of the 8MB bug striking with the updated firmware, that's why Intel discontinued the 320 series, they never managed to actually fix the problem.
Re: (Score:3)
haven't failed me yet.
Careful with your selection bias. You could have used exclusively these products all your life and it would make no difference to the topic at hand unless you're ordering every component in almost commercial quantities.
Case in point: I have in service a functioning OCZ Vertex 3, and until I threw it in the bin in fully working condition, an IBM Deathstar 60GB drive. Both of them infamously known for being such a shit product they effectively sank the companies (or in IBM's cases their HDD division).
Yet neit
Re: Intel & Corsair (Score:2)
OCZ's sin was using Sandforce controllers in a mode that required a supercapacitor for backup power (to guarantee the drive would never lose power mid-write)... then omitting it to save about $2 in manufacturing costs.
Sandforce controllers had two operating modes. One wrote every update multiple times, using bit flags to mark old/intermediate copies as 'committed' when finished. The other wrote everything exactly once. In 'safe' mode, the drive could recover from mid-write power loss... but ran slowly (comp
Re: (Score:2)
I haven't used Intel, but I have seen too many failures with Corsair drives. I'm dealing with one now. It seems to have corrupted its SMART data.
"Why" missing from summary (Score:4, Informative)
Crucial has swapped the TLC NAND it originally shipped with QLC NAND â" and not terribly good QLC NAND, at that. The new version of the P2 has two fewer NAND chip packages than the original, and significantly fewer total dies. This reduces the total potential bandwidth the SSD controller can achieve and further harms the performance of the 500GB drive. Average power consumption on the QLC drive is lower, at 1.49W, but total power efficiency is actually worse because the savings do not make up for the dramatically slower performance. If full drive performance on the P2 was already bad, itâ(TM)s downright abysmal on the P2 with QLC NAND
Re: (Score:2)
This links to an article about WD, not Crucial.
Bait and Switch (Score:4, Funny)
The article is actually about WD HDDs.
Re: (Score:3)
Re: (Score:2)
No it's about the SN550 Blue which is a SSD.
Whoosh!!! (I say that both to you, and to whoever modded GP down).
Re: Bait and Switch (Score:2)
Re: Bait and Switch (Score:2)
contextualize Have you never heard "bait and switch" in another context? I mean FFSWTF?
Re: (Score:2)
Re: (Score:3)
The bait is rotten fish, it was switched to rotten fish, so no foul.
Re: (Score:2)
Re: (Score:2)
Bait: SSD
Switched for: HDD
HTH, because I don't think anyone's going to be able to break it down any more than that. :)
Re: (Score:3)
The article is actually about WD HDDs.
I particularly liked the part where it says: "The 2TB Crucial SSD I purchased for my own video editing work is one of the bait-and-switched units,"
Isn't "Crucial" a whole different brand than Western Digital?
Demand an full refund use an CC change back if nee (Score:4, Insightful)
Demand an full refund use an CC change back if needed.
Re: (Score:3)
Re:Demand an full refund use an CC change back if (Score:4, Informative)
That puts the hammer to retailers, but not necessarily to the manufacturer.
PSA: WD == Sandisk (Score:5, Informative)
Public Services Anouncement:
Western Digital owns Sandisk since 2016. So, if you will stop buying Western Digital because of this, Sandisk is not an option.
Also, Western Digital is in talks to buy Kioxia, so, if that merger goes to fruition, Kioxia would not be an option either.
On a final note, Kioxia (nee Toshiba Memory) and Western Digital (nee Sandisk) are Fab partners, meaning they own the fabs 50-50, they invest in engineering and development 50-50, and share the outut 50-50, so, if the emrger happens, is still the same flash chips, no more or less.
Re: (Score:2)
Western Digital owns Sandisk since 2016. So, if you will stop buying Western Digital because of this, Sandisk is not an option.
Well, shit.
Great heads up as that's where I would have headed next.
PSA Number 2 (Score:4, Informative)
ADATA has also done this. Seagate has also done this. And Kingspec, Dataram, and Avant.
It's been reported that Kingston has done this. It's rumored that Crucial has done this.
Samsung has NOT done this. Hynix has done this but has updated SKUs and descriptions to reflect that it's made with different parts.
https://www.pcworld.com/articl... [pcworld.com].
Re: (Score:2)
Samsung also doesn't lie in the ATA commands like most other brands. This is important for error correction down in the stack.
Intel is also on good behavior but Samsung has a major pricing advantage. Either are trustworthy for critical data, at least in the 'pro' editions.
If you just have petabytes of pr0n to back up then pick whatever is cheap and mirror it.
Anyone remember the SMR/CMR issue a year back? (Score:2, Insightful)
This reminds me of the SMR/CMR issue for 6 TB drives. SMR drives may be OK as a single drive, but if used in any type of RAID, can take days to weeks to finish syncing or rebuilding. One had to go carefully by model and serial numbers in order to not to buy a dud drive, especially for a home NAS.
Re: Anyone remember the SMR/CMR issue a year back? (Score:2)
Re: (Score:2)
Not even as a single drive for many scenarios. Picture rotating off-site backup disks. Good performance copying data to it at the beginning, then dazzle at the 30MB/s that soon come up.
Why is the subject about Western Digital (Score:2)
And the body about Crucial?
Re: (Score:2)
Because the author wrote a piece about two weeks ago detailing where Crucial did basically the same thing. He was thinking the Crucial bit was a one-off screw up, but it now looks like other mfgs are doing the same thing. WD, at least.
Fool me twice? (Score:5, Informative)
After silently swapping CMR WD reds, for SMR drives which simply are not fit for purpose in a NAS It became really hard to suggest that WD even understood their own product lines. This just seems to be more of the same behavior, although at least in this case its just _slower_ not that its possible for the system to catastrophically fail.
It does suggest that internally WD is looking for ways to command a premium price and switch the product for something that doesn't perform as well. This isn't entirely uncommon in the computing industry as the laptop manufactures and large OEM's have been playing this game for decades. For example when the I7 was announced the assumption was that it was a high end desktop processor, but then they started putting them in laptops where the actual sustained/etc perf was much closer to the i3 product lines. Similarly for GPUs where the mobile variants were named similarly to the desktop parts but provide a fraction of the perf. So this is probably more of that thinking, sell a product labeled one way hoping that no one notices that it sucks.
Re: (Score:2)
I used to be a big WD fan, very loyal customer.
I recently selected Seagate for my new cluster.
I've got WD HDDs that I've been spinning for over 10 years that are still working great, but I just don't trust them anymore, or their new drives. For SDD I switched to Samsung.
In the old days, Seagate was who I didn't trust. It made me sad to change who I trust, but such is the world. Don't hate Brand X too much, they may be better some day!
Re: (Score:2)
Same. I used to run all my RAIDs with WDs. Had a great experience, both with their performance as well as WD's customer service when having to deal with failed drives.
Ok. 'twas good while it lasted. But I'm not going to honor getting bullshitted by a company. NEXT!
Re: (Score:2)
I wish more people would get this. The number of times I've had to bite my tongue because someone has told me they'll never deal with a company again because of something that happened 10+ years ago, as though they are doing anything more than cutting off their nose to spite their face; often that same company is now seen as one of the better options; meanwhile the company they swapped to is often producing poorer products by that point. Learn from yo
Re: (Score:2)
I recently selected Seagate for my new cluster.
Good luck with that. Seagate still has the worst unpredictable reliability out there.
That's just dandy (Score:2)
Along with the recent hack on their NAS and SMR HD contraversy, WD is now on my shit list. My WD MyCloud NAS is reporting 6TB free after the new update to My Cloud OS 5. It's a 2TB drive. Yippie free upgrade /s.
Re: (Score:2)
WD has been on my shit list for well over a decade. I still have stacks of dead 8GB WD drives from back when I did break / fix PC repair that I haven't used for target practice yet. At least a Seagate would warn you with the click of death that it was going to die... eventually.
WD's just up and died with zero warning.
They seem to be the first to admit errors (Score:2)
https://hardware.slashdot.org/... [slashdot.org]
Honestly since the days of MFM/RLL drives, I've always used them as my go to for storage solutions. Maybe this is too dated to mention but I never had a WD with striction issues, nor have I EVER had an issue (in the rare case) with warranty.
So, everyone screws up, but I've always found WD to stand up.
My experience..wtf do I know(A E. Newman)?
PSA #2: WD Too Little, To Late comunique (Score:3)
Western Digital Acknowledged the screw-up and pinky swore to never do it again:
https://www.theregister.com/20... [theregister.com]
IMHO: Too Little, Too Late
Toshiba for HDDs and Samsung and Corsair for SSDs (Score:2)
Let's see:
WD screewed the pooch with their shingled shenanigans.
Since Seagate are not the nicest peas in the pod, for my CONSUMER (including my DS1515+) HDD needs, Toshiba it is.
What? You forgot Toshiba made HDDs too?
(on a side note, since 2015, HitachiGST drives are in reality WD disk drives).
And since WD/Sandisk (they are one and the same) screw the pooch again with their SSD shenanigans, like Adata, and Crucial, and Kingston, and... Then Corsair and and Samsung it is for my CONSUMER Flash storage needs..
Re: (Score:2)
Seagate sucked 20 years ago, it's true, but is there a contemporary reason to avoid them?
Re: (Score:2)
No, some models are good. But you still have to research very damned model. Perhaps it's best to say there's a reason to be suspicious
Re: (Score:2)
That's true any time you make a purchase, though.
Re: (Score:2)
The problem with Seagate is that some of their drive models are entirely fine and some are catastrophic. They never managed to reach consistent good quality and they still have not. Hence with Seagate you can only buy older models that you have researched carefully for reliability.
Re: (Score:2)
My research has not uncovered reports of these problems.
The question was, is there a contemporary reason to avoid them?
I asked about a reason . Not just, is there a guy who says they're bad?
Re: (Score:2)
Reason? MTBF & $ per GB. There are branding effects in consumer segment, but if you look at anyone doing storage at a scale, there's no loyalty sg/wd, There's only loyalty to specific production runs.
Re: (Score:2)
I've been super happy with Toshiba NAS drives. Tons of cache for speeding things up, great price point, great warranty, no hassle with service on the rare occasion I've needed it.
Honestly I care more about reliability. (Score:2, Insightful)
But for me, as long as it is not slow under normal pleb use I care more that is doesn't corrupt or lose my data over it's claimed lifetime (per smartctl Percent_Lifetime_Remaining).
Re: (Score:2)
But for me, as long as it is not slow under normal pleb use I care more that is doesn't corrupt or lose my data over it's claimed lifetime (per smartctl Percent_Lifetime_Remaining).
In some cases this is because they've secretly swapped out the TLC flash for QLC flash, in which case reliability is going to go down the pan as well as the data rate. This is definitely the case for the Crucial drive mentioned, and other articles point to ADATA doing this. If WD are playing that game as well, then yeah, I'd be worried.
Who cares? (Score:1)
If you fill up any SSD's buffer, then performance suffers. If you want a real SSD, get a samsung. If you want cheaper options for a friend get a WD (they won't notice the difference anyway if they aren't smart enough to install the SSD themselves.)
How can I tell? (Score:1)
I bought a WD SN550 Blue recently. How can I tell if mine's "bait and switch", or whether the info given on my shop's website is correct? https://www.komplett.no/produc... [komplett.no]
I tried userbenchmark,com utility and it gave the following info: https://i.imgur.com/utb23AG.jp... [imgur.com]
Which is a bit hard to compare.
IDK if the store info Intern datahastighet (internal speed) 2400 MBps (read) / 1950 MBps (write)
is comparable to the benchmark: Read 1,596 Write 1,779
Both should be MegaByte per second. Store info is "internal
Re: (Score:1)
Aaaand reading the article further I got my answer. With pictures, even! :-/
Sabrent (Score:2)
Many of my favorite "budget" NVMe drives as of late have come from Sabrent. Are they pulling the same stupid stunts as Crucial/Western Digital/Adata?
while we're at it (Score:2)
" It is unethical for any company to sample and launch a product to strong reviews only to turn around and sell an inferior version of that hardware at a later date without changing the product SKU or telling customers that they're buying garbage. "
Looks at Pyrex...
Re: (Score:2)
Buy import PYREX intended for sale in other regions instead of domestic Pyrex (note capitalization difference) because the PYREX is still borosilicate glass.
Re: (Score:2)
Consumer protection would rip them a new one and their sales would vanish if they tried that in Europe.
Re: (Score:2)
I never claimed the average American wasn't dumb, or willing to work against their own interests.
However, for the rest of us, we can do a little research and find out if we can work around some of the flaws in the system.
I've never seen WD behaviour in tech before. (Score:1)
Saying that, Seagate and the rest as bad, with SMR discs.
The assumption that technology is always improving just does not hold for hard drive manufacturers. From about 2002 to 2007, huge amounts of them failed due to flux defects and I've got few left from archiving back then.
Now with them silently bringing in SMR, I'm seeing hard drives I've bought (with little information to confirm are SMR) write at 1MBytes a SECOND! This is shockingly 1/100th of the normal speed. This is worse than early 90s speed with
It's just "v2" of the same device (Score:2)
If it's that different then it should have a different model number. I hate it when manufacturers use the same model number and just add "v2" on the label when they change the hardware. If they make significant changes then it should clearly be a different model number and UPC. The "v2" should be PART of the model number and not just a footnote. Look at HP lab equipment. They used the same model number but added A, or B, or C, etc. to the number. It made it clear that while the basic design is still the sam
For the right customer? (Score:2)
It looks as if this would be a decent compromise for the right people. IF the hard drive was advertised like that, and IF it was sold cheaper than other drives. And a larger cache wouldn't really solve the problem for people affected by
Western Dataloss? (Score:2)
Seriously, who uses Western Dataloss SSDs?
Why?
Masochism?
If their SSDs are even 10% as bad as their garbage hard drives they're going to be awful.
Bring back Bigfoot! (Score:3)
I wish somebody would resurrect the "Quantum Bigfoot" idea for big laptops & make a spinning drive with DVD-diameter platters (possibly augmented by a few gigs of ssd cache) with the form factor of an optical drive... so people with big laptops could swap the optical drive for a big, slow 4-8TB second drive for bulk tonnage.
Or hell, how about making laptops with ~12mm 3.5" drives instead of 2.5", so we can finally break the damn 2TB laptop-drive barrier? Modern laptops have plenty of "horizontal" space thanks to big, wide displays. Why not take full advantage of it by going larger than 2.5", so we can have a ssd card as the primary drive, augmented by one or more 4TB+ additional drives?
For that matter, why are ExpressCard SSDs not a real thing? A 2TB ssd is about 1/4 the price of a 4TB ssd, ExpressCard is bigger than mSATA, and is basically a 1x PCIe port. It's one more place bay-constrained laptop users could stick another semi-affordable secondary drive.