Shrinkflation, SanDisk Style

To me this reeks of Western Digital style dishonesty[1][2]. I was wondering how long it would take for the new parent company to rub off on SanDisk and I guess I have my answer. My suggestion is to stay the hell away from anything Western Digital (and now SanDisk) if accurate specs are important to your use case.<p>[1] <a href="https:&#x2F;&#x2F;arstechnica.com&#x2F;gadgets&#x2F;2020&#x2F;05&#x2F;western-digital-gets-sued-for-sneaking-smr-disks-into-its-nas-channel&#x2F;" rel="nofollow">https:&#x2F;&#x2F;arstechnica.com&#x2F;gadgets&#x2F;2020&#x2F;05&#x2F;western-digital-gets...</a><p>[2] <a href="https:&#x2F;&#x2F;arstechnica.com&#x2F;gadgets&#x2F;2020&#x2F;09&#x2F;western-digital-is-trying-to-redefine-the-word-rpm&#x2F;" rel="nofollow">https:&#x2F;&#x2F;arstechnica.com&#x2F;gadgets&#x2F;2020&#x2F;09&#x2F;western-digital-is-t...</a>

Here&#x27;s my problem: I&#x27;m glad to pay a premium for Sandisk if it guarantees reliable, honest, and accurate devices. I can buy generic flash drives for a fraction of the cost.<p>If Sandisk is starting to cheat, why would I pay the premium? And if I&#x27;m willing to pay a premium, whom should I buy from?<p>Sandisk, prior to WD, was expensive but honest. I was concerned this might happen, since WD has a bad track record here.

I&#x27;ve tried to find relevant EU law about this. IANAL, of course.<p>According to &#x27;misleading and comparative advertising&#x27; - <a href="https:&#x2F;&#x2F;eur-lex.europa.eu&#x2F;legal-content&#x2F;EN&#x2F;TXT&#x2F;HTML&#x2F;?uri=CELEX:32006L0114&amp;from=EN" rel="nofollow">https:&#x2F;&#x2F;eur-lex.europa.eu&#x2F;legal-content&#x2F;EN&#x2F;TXT&#x2F;HTML&#x2F;?uri=CEL...</a> - it seems this could fall under article 2b and 2c, defining misleading resp comparative advertising. The definition of misleading here is:<p><pre><code> [..] any advertising which in any way, including its presentation, deceives or is likely to deceive the persons to whom it is addressed or whom it reaches and which, by reason of its deceptive nature, is likely to affect their economic behaviour or which, for those reasons, injures or is likely to injure a competitor [..] </code></pre> This is a directive, unfortunately, so the actual law resides with (and is different in) each member state. OTOH, this means inhabitants might complain to their regulator, and don&#x27;t need to sue sandisk themselves.<p>I was hoping this one would help. It says which errors are allowed when measuring things. Unfortunately it only speaks of weights and (physical) volumes, not other measurements like number of bits. <a href="https:&#x2F;&#x2F;eur-lex.europa.eu&#x2F;legal-content&#x2F;EN&#x2F;TXT&#x2F;HTML&#x2F;?uri=CELEX:01976L0211-20190726#tocId14" rel="nofollow">https:&#x2F;&#x2F;eur-lex.europa.eu&#x2F;legal-content&#x2F;EN&#x2F;TXT&#x2F;HTML&#x2F;?uri=CEL...</a><p>In general, I don&#x27;t have much idea about how to navigate the EU consumer protection landscape. So while I do have a lot of protections, it is sometimes hard to know which ones. Does anyone know a good site or something?

So they need to overprovision, and compensate for effects - that is fine.<p>However, only advertising &quot;usable size&quot; is truthful, everything else is misleading.<p>It is sad that the legal system tolerates such cheating because it makes it harder to find the actual properties of the product. Or even makes it impossible to actually know how the product will actually work.

I have had the same happen to me with Toshiba MicroSD cards. I ended up returning them.<p>I also bought a Transcend 128GB USB stick, which i unfortunately did not return (yet). I have communicated with Transcend support and they said it was an issue with partitioning and filesystems - which is wrong, because the tools i use show the capacity at a lower level. They kept stubbornly repeating these statements.<p>I may end up returning it which shouldn&#x27;t be a problem even after almost 2 years because the defect has been there since the beginning.<p>Here&#x27;s what diskutil has to say about it:<p><pre><code> % diskutil info &#x2F;dev&#x2F;disk4 Device Identifier: disk4 Device Node: &#x2F;dev&#x2F;disk4 Whole: Yes Part of Whole: disk4 Device &#x2F; Media Name: Transcend 128GB Disk Size: 123.7 GB (123718336512 Bytes) (exactly 241637376 512-Byte-Units) Device Block Size: 512 Bytes </code></pre> For comparison, i get this for a Lexar 32GB USB stick:<p><pre><code> Disk Size: 32.0 GB (32008830976 Bytes) (exactly 62517248 512-Byte-Units) </code></pre> You can clearly see that more than 4GB are missing on the Transcend 128GB stick. It should be around 119.2GiB (128GB) but it is 115.2GiB (123.7GB)

Looks like it&#x27;s just a way to utilize chips with defects, instead of throwing them away.<p>Some of the missing gigabyte is probably used for a memory mapping algorithm.<p>Anyone who has dug deeper into this issue?

The author mentions they don’t “trust” VenToy yet (they don’t state which kind of trust) my question is why?<p>I’ve used VenToy just fine for a while now. Am I missing something?

I sometimes wish a power of two didn&#x27;t end up so close to a power of 1000.

This is simple to resolve. If I buy a 16 GB flash drive with fewer than 16 billion bytes of usable capacity then I will demand a refund and return the device at the retailer&#x27;s expense. I will also complain to trading standards.

Hmm. I also have a 16GB SanDisk Ultra made in China like the old one. I don&#x27;t remember exactly when I bought it but it was several years ago. Mine shows up with a usable size of 14.77 GB (30965760 blocks) from the same f3probe command the article runs. Testing my terrible no name 16GB stick shows 14.65GB usable.

There are tools out there that let you reconfigure the flash controller to have less overprovisioning. You&#x27;re losing reliability this way of course, but if you want to get exactly 16&#x27;000&#x27;000&#x27;000 bytes out of the stick there are ways to accomplish this.

What the hell? Why is this legal? I assume it&#x27;s very illegal to say &quot;12 eggs&quot; and then the box only contains 9, how is this allowed?

Burying some way more interesting unnoticed metric in the logs.<p>In practice, I would trade a ~75% faster drive for &lt;10% of storage in a thumb drive use case with zero hesitation:<p>Old: Probe time: 1&#x27;49&quot; Read: 463.8ms &#x2F; 4814 = 96us Write: 1&#x27;48&quot; &#x2F; 4192321 = 25us<p>New: Probe time: 1&#x27;02&quot; Read: 338.2ms &#x2F; 4814 = 70us Write: 1&#x27;02&quot; &#x2F; 4192321 = 14us

Tangent, but I think the practice of shrinkflation should be made illegal altogether (although not sure how you would enforce it). If you need to increase prices, fine, but it&#x27;s a complete waste of packaging materials to just give less of something for the same price in the same package, not to mention the environmental harm.

I wonder why there are so few usb flash drives that have a physical &quot;read only&quot; lock.<p>That was so useful to defend against viruses.<p>A cheap flash drive will cost say 15 dollars and one with a physical button&#x2F;lock to make it read only will be 150?<p>The cost of materials should be maybe 5 more dollars...<p>On a side note, if you go into one of those &quot;print shops&quot; that can print things for you and use an USB stick, then how can you sanitize your stick from the multiple viruses &#x2F; worms that you will get? Only thing that comes to my mind is to use a computer where the hard drives are physically disconnected and some live CD linux. But even then it is unclear if you can really format the usb flash memory to make it safe.

It would be nice to to have something like caniuse.com or dpi.lv, but for “usable storage space on device”. I doubt it would shame most manufacturers into actually providing what they advertise, but it might just create enough visibility for one or two manufacturers to commit the extra resources to delivering on the promise.

This has been going on my entire career, starting in at least 1996 or so. Maybe SanDisk was an exception? People have been complaining about this for a very long time so I don’t think it’s quite right to call it Shrinkflation.<p>But, maybe I’ve been conditioned into complacency. I do agree, manufacturers should have to post usable space. I don’t see this changing though.<p>Edit: Removed note about 1,000,000,000 bytes. The article isn’t specifically about that, but I’m not convinced that’s ever been the only factor.

I&#x27;m confused by the report from F3 probe even for the &quot;good&quot; one.<p>It says:<p><pre><code> Device geometry: *Usable* size: 14.91 GB (31266816 blocks) Announced size: 14.91 GB (31266816 blocks) Module: 16.00 GB (2^34 Bytes) </code></pre> So why it&#x27;s only 14.91? I assume it uses GiB (1024) -- but then, why &quot;Module&quot; is 2^34 B? Shouldn&#x27;t it also be 16000000000 B or 14.91 GB?

Any other reasonable explanations to explain a measurement error? Partitioned differently? I might be a bit too critical but it seems farfetched for me that they&#x27;ve withheld 579 mb only for no one to notice before. I&#x27;m predicting this will become major news on tech blogs. People love &quot;company X misleads customers&quot;. It just seems too good to be true.

The 32GB and 64GB SanDisk variants are actually cheaper on Amazon than the 16GB one. Yes, the advertising is misleading, but OP could&#x27;ve saved money and quadrupled his space if he wanted to.

I&#x27;d bet it&#x27;s because your new one uses an MBR partition as your tools show you clearly, that indeed both sticks are 16 GB sticks

I stopped buying WD many years ago because of their dishonest marketing tactics and, frankly, shit products.

what is wrong with using ventoy?

How can SanDisk redefine a SI unit to suit them? Don’t we have GiB for that?

The new one also has lower performance

one thing that isn’t clear from this article is where he bought it from and what it said in the listing. all he says is that he bought one of the same size and brand. if the listing says 15gb, then I don’t see a problem. it’s up to you to see that and perhaps choose a different option

one thing I&#x27;ve learned - IODD is cool

That should be illegal, they should use the same rules as for food packaging.<p>For a 1kg bag of flour bought in the EU, that would be<p>- On average, you should get at least 1kg of flour<p>- No bag should be less than 985g<p>- Packaging does not count<p>For a USB stick, that would be<p>- On average, the actual capacity must be no less than what&#x27;s advertised<p>- For a given product, the actual capacity must be no less than what&#x27;s advertised, within a small margin<p>- Only usable capacity counts

Sadly OP is behind the times here. The storage industry has pretty much co-opted GB to mean 1000^3 bytes, which is why you see folks in the know refering to GiB for the power-of-two-numbers. This is super-frustrating, but it&#x27;s been like that for decades (literally: this was finally &quot;officially&quot; resolved in the late 90&#x27;s by the IEC).<p>It&#x27;s frustrating that SanDisk used to give you extra bytes and they stopped -- and everyone including me HATES it when products get worse with no external indication that they changed.......but let&#x27;s be honest: it&#x27;s kind of SanDisk&#x27;s primary MO to buy the cheapest NAND they can find and sell it on the consumer market.