Bad blocks on new nvme

[ThrashZone]

Yeah it's brand new, not refurbished. I bought it today. Any replacement recomandation between Samsung 970 evo 1tb or Adata sx8200 pro 1tb ? or something else around this price range ? Also, isn't it weird the drive already had 1tb written? That made me think it wasn't new but the usage time was only 2 hours.

Hi,
I wouldn't buy another adata product.
Go with the sammy or even western digital

 

[RJARRRPCGP]

Its firmware, looks like poo! Looks like a false lifetime reading as well!

 

[maxfly]

Unfortunately i cant help with the issue at hand but i can say avoiding adatas 8200pro drives is a good idea. They have made several changes to the original controller and nand(very competitive) without providing a revision or even a press release.

Adata Switches to Slower Controller, NAND on SX8200 Pro

A change in components could mean slower performance.

www.google.com

 

[jsfitz54]

Hi @jsfitz54, here it is:
View attachment 184811

For good measure, here is the result from HDS:

View attachment 184815
An earlier HDS surface test found 13 bad blocks and 1 damaged one, but for some reason I can't find the export file I (think I) saved.

I already made peace with the fact that I'll have to get a new drive, but I'm curious to understand what is going on here. I'm tempted to nuke all the data and try a full format to see what happens. I'm still intrigued by @thesmokingman comment there, my system is indeed a Ryzen system with all those bells and whistles that I don't fully understand

I would not be concerned about the drive at this time. A FULL Format may reveal if there are any hard flaws.

 

[thesmokingman]

I already made peace with the fact that I'll have to get a new drive, but I'm curious to understand what is going on here. I'm tempted to nuke all the data and try a full format to see what happens. I'm still intrigued by @thesmokingman comment there, my system is indeed a Ryzen system with all those bells and whistles that I don't fully understand

WHEA errors due to memory/imc issues can manifest in drive errors. Raising soc voltage can actually exacerbate the issue instead of fixing it. It's a very Ryzen specific problem.

 

[AeliusRuivus]

WHEA errors due to memory/imc issues can manifest in drive errors. Raising soc voltage can actually exacerbate the issue instead of fixing it. It's a very Ryzen specific problem.

After reading your comment I tried to lower it and then run a test, but my knowledge on how to do it is limited, so what I did was to simply change the power plan hoping that the lower clock would result in lower voltage overall. Would you suggest another way to test this hypothesis?

 

[thesmokingman]

After reading your comment I tried to lower it and then run a test, but my knowledge on how to do it is limited, so what I did was to simply change the power plan hoping that the lower clock would result in lower voltage overall. Would you suggest another way to test this hypothesis?

First check your event viewer for any whea errors. Then you can setup hwinfo which has a whea error tracker, then set the whea stat to show on your taskbar. This allows you to know if you ever get a whea instantly.

After monitoring this stat, do you get whea errors? What's you system specs while we're on the the topic?

 

[AeliusRuivus]

I have to admit that I didn't knew what WHEA was until now, but to my surprise, I have none in Event Viewer:

My system is an Asus Zephyrus G14 with a Ryzen 4900HS and a GeForce 2060 Max Q. As we speak, those are the frequencies and voltages:

 

[RJARRRPCGP]

Looks more like just a faulty drive.

 

[jsfitz54]

Looks more like just a faulty drive.

How do you figure? MACRORIT says it's for SSD's but no mention of NVMe. https://macrorit.com/disk-surface-test/disk-surface-test.html

I would not trust the test, it appears to be more aimed at mechanical drives. Seems like an odd source for reading NVMe drives.

 

[RJARRRPCGP]

How do you figure? MACRORIT says it's for SSD's but no mention of NVMe. https://macrorit.com/disk-surface-test/disk-surface-test.html

I would not trust the test, it appears to be more aimed at mechanical drives. Seems like an odd source for reading NVMe drives.

The SMART values.

 

[AeliusRuivus]

So, small update on my end. Since there's nothing critical on this SSD (at least nothing that is not backed up on the cloud), I decided to buy HDS pro and try to run the surface test with the repair option - aparently it worked:

I understand that HDS it just writes it over and does the CRC, and if it passes, mark as solved. SMART still reports 100% health and no spare block swap. Another pass on the former bad blocks even got a good read speed (lighter green - and those tests are read only, so no harm to the SSD!):

In any case, I still don't know what caused the event viewer disk errors, or the bad blocks/crc fails, neither I'm confident that they won't come back. I haven't got any WHEA so far, so this is probably either file system or the SSD.

Someone on Reddit suggested that maybe the M.2 module was not properly seated in the board, since it is a double faced one. I opened the computer to check and, to my horror, Asus put the WiFi board directly under the M.2 drive! While there is still some room, they are almost touching each other, and I wonder if some sort of electric interference could be the culprit here...

I'll update this post in case I got any news, so anyone else in a similar situation can have it as a reference.

 

[R-T-B]

The SMART values.

They look perfect to me?

 

[Solaris17]

They look perfect to me?

Running tests meant for platter drives on flash memory, confused about results.

This world scares me.

 

[R-T-B]

Running tests meant for platter drives on flash memory, confused about results.

This world scares me.

Buddy, you ain't seen nothing yet.

 

[AeliusRuivus]

Running tests meant for platter drives on flash memory, confused about results.

This world scares me.

I appreciate your input on which tests I can run instead.

 

[AsRock]

So, small update on my end. Since there's nothing critical on this SSD (at least nothing that is not backed up on the cloud), I decided to buy HDS pro and try to run the surface test with the repair option - aparently it worked:

View attachment 185473

View attachment 185474


I understand that HDS it just writes it over and does the CRC, and if it passes, mark as solved. SMART still reports 100% health and no spare block swap. Another pass on the former bad blocks even got a good read speed (lighter green - and those tests are read only, so no harm to the SSD!):

View attachment 185477

In any case, I still don't know what caused the event viewer disk errors, or the bad blocks/crc fails, neither I'm confident that they won't come back. I haven't got any WHEA so far, so this is probably either file system or the SSD.

Someone on Reddit suggested that maybe the M.2 module was not properly seated in the board, since it is a double faced one. I opened the computer to check and, to my horror, Asus put the WiFi board directly under the M.2 drive! While there is still some room, they are almost touching each other, and I wonder if some sort of electric interference could be the culprit here...

I'll update this post in case I got any news, so anyone else in a similar situation can have it as a reference.

Sorry forgot to get back to you about HDS, yeah it's a good program.

Running tests meant for platter drives on flash memory, confused about results.

@Solaris17 it's only a read test.

 

[R-T-B]

@Solaris17 it's only a read test.

It still won't work if it does not understand ssd timings.

Hard Disk Sentinel does though, despite the misleading name.