Investigation uncovers SSD maker bait and switch practices

by Mark Tyson on 3 December 2020, 11:11

Tags: Adata (3260.TWO)

Quick Link: HEXUS.net/qaept5

Add to My Vault: x

An investigation has uncovered some alarming behaviour from SSD makers. According to analysis and tests by Tom's Hardware, certain SSD makers have grown bold in swapping product components willy nilly, likely done to trim costs and earn fatter margins on goods sold rather than to improve products over time.

Its most damning findings concern Adata, a company whose SSDs you can readily purchase in the UK. As per our sub-headline, one Adata SSD model which was tested was over 40 per cent slower in file copy operations, as well as exhibiting a 500MBps decline in light workloads after the unadvertised component changes.

After noticing the above, the investigation focussed comparisons and testing on the Adata XPG SX8200 Pro which won an award from Tom's Hardware when it was reviewed back in 2018 and was highly a recommended product to those looking to buy in subsequent months. The investigation noticed that Adata has since changed not just the controller but the firmware and the NAND – without any information shared about the change.

Adata has made a statement about the findings and asserts that controllers were updated to newer models from controller maker SMI, bringing along improvements such as greater storage durability. It went on to say it guaranteed that every XPG SX8200 Pro would perform to spec. However, Tom's Hardware did find some versions below performance spec and noted that the spec was rather broad to begin with. Adata SSD specs mention an SMI controller but not the model, providing them component switching latitude. Last but not least the SMI component switch didn't seem to be to a newer choice but a contemporary alternative choice.

To back up its report findings and conclusions Tom's Hardware tested three Adata XPG SX8200 Pro variants in various use cases and benchmarks. In summary it was rather disappointed that the original version reviewed was a noticeably better performer in a wide range of tests. Readers who followed the site's recommendation also had no way of knowing what component swaps had been made – as no information was provided about this by Adata.

Not all component choice changes, whether mentioned by the product maker or not, will be to the detriment of the product. The best companies will upgrade parts as they become available AND notify customers of the changes via spec sheets / packaging/ model number increments.



HEXUS Forums :: 39 Comments

Login with Forum Account

Don't have an account? Register today!
That's very disappointing, I have 3 of the drives as originally when they came out and the non pro version before it they represented excellent value for the performance.

Only just got the last of them 2 weeks ago and thought it strange that it reported a different drive size to the others when formatted, so something had changed between the models.

Guess I will need to do some performance tests to see if it is not as expected and send it back.
My initial reaction to that is that it's tantamount to fraud, if not legally quite there.

My second thought is that, if allowed to get away with it, it renders product reviews utterly pointless.

Most major review sites (and I'm talking pre-internet) used to require products to be sold as tested, or at least as well as tested, for a minimum period as a condition of submitting for testing. I wonder what happened to that?

My third thought is that it's a good reason to only buy from companies that have sufficient to lose from underhand tactics that they wouldn't risk the reputational damage in case they got caught. But … who?
Kanoe
Guess I will need to do some performance tests to see if it is not as expected and send it back.

According to the original Tom's Hardware article, not only was the newer Pro version provided by ADATA different from the original Pro version, but the one they bought themselves was different from both of those too.
Adata seems to be a brand to avoid then. Bait and switch alive and well.

Is there a utility available that can identify which controller and NAND is being used on any particular drive?
Excellent work - Adata,Kingston and Sabrent have all done this before.