Question Seeking aid with the notorious Kingston A400S37 SSD

[MezMez]

Greeting everyone. Newcomer here.

So, to get to the issue at hand. I've noticed long loads and performance drops while playing data-heavy games, like Path of Exile. It usually happened whenever a new wave of data floods my rig, and I'd feel stutters. So I decided to dig a bit deeper, and the first suspect was SSD. I didn't need a long look to see something is fishy.

I'm aware that A400S37/240GB is a low-end SSD, and that my rig is out-dated. ASUS B75M-a, i5-3350p, RX470 nitro 4gb, 2x4GB Kingston dd3 1600 Mhz. That being said, I'd still like to get some help with this particular issue that has been bugging me for days. The help and ideas other than "your spec sucks, go by new SSD, not A400". Yeah, I know. Best I can do at the moment, unfortunately.

To start, I'm submitting the CrystalDiscMark benchmark scores. I've done them multiple times over the past few days - they've been all over the place. I'll submit the last few, as those scores are consistent, and I feel, on point in seq - but something odd follows. Here are the results.

SSD is a couple of months old. Clean Windows 10. Health is ok. Up to date drivers. ASCI mode in bios. Connected to SATA 3 port. IDE ATA/ATAPI controllers - Standard SATA AHCI controller. GPT partition style. I even flashed BIOS, with the latest version available for my old MB.

I didn't, however, manage to get the SSD's firmware updated. I've got the Kingston utility program, KSM - Kingston SSD Manager. It detects(?) the SSD, but it doesn't show it.
Here.

Quite a lot of info there, and I can't make heads or tails of it.

Possible Counterfeit Drive Detected (CSMI Path), Device 0 [phy=0] not available with reason code 2, what the heck are those lines?!

So, yeah. If anyone has any tip or idea on how to update the firmware or how to improve the performances of the aforementioned drive in general, I'd be utmost grateful!

 

[Obakasama]

Not sure, but feel compelled to comment....

What else have you monitored? Because the storage medium isn't what I would be looking at with respect to gaming performance. A new wave of data would also suggest it's something your CPU, GPU and RAM will have to contend with as well. Do you use something like MSI Afterburner to monitor those other aspects of gaming to see where performance might be affected? If not I would start there.

 

[MezMez]

Hello Obakasama,

It's all good, all kind of input is welcome. I suspect that other parts of my rig might be bottlenecking the smooth flow. RAM is the next main suspect, as the aforementioned Path of Exile isn't that CPU hungry to my knowledge. As far as GPU goes, I've minimized the graphical settings both in the game and in the AMD setting to optimize performances, so the game usually plays nicely, stutters do not seem to be related to the clutter on the screen. It happens when there's some new encounter or influx of data - PoE is the kind of a game that completely chokes up on the big in-game events when there is a whole bunch of new stuff - the old and fragmented HDD's and/or slow SSDs that can't keep up with the data flow - it's the most common issue in PoE. I'm not adamant that's its the only problem, it's just the reason I tried to start examining the SSD. As far as monitoring, I've been using Sapphire Trix for the fan curves for my Rx470, and GPU-Z to track GPU load. I am pretty dim when it comes to hardware and spec, so to be completely honest, other than tracking performances in the task manager, I have no idea how to monitor other aspects of gaming, as you've put it. If you have any advice, I'm open to suggestions!

That being said, I'm still not over this SSD thingy. At this point, it's not even about the overall PC performances anymore, I'm genuinely way too curious about what is going on with the naughty SSD. Don't get me wrong, I'd like to address all potential issues with my setup, but this particular issue really grinds my gears.

 

[Obakasama]

That's fine. Heard some things about that SSD model way back, though I do put it under 'bait and switch'. Even if it isn't up to the original spec it's still faster than a mechanical HDD. I would have to assume the way the SSD is the issue in such a scenario is the game requires such an amount of read/write speed that it would completely overwhelm it; which itself is hard to comprehend under normal scenarios. I just have a bit of doubt about this factor, but hopefully others can address it.

As for other monitoring I mentioned MSI Afterburner. Personally I would monitor CPU (all core/threads), GPU and RAM frequencies and usages; CPU and GPU temperatures; fps and frame times; and page file. The general idea is to log ingame performance to see how the hardware copes. Depending on the onscreen experience, the stuttering you mention, cross reference the fps and frame time drops and peaks to see whether any other component is also dropping or abnormally high. For all we know it could be there isn't enough RAM to cope and the SSD has to be used frequently for the page file.

 

[MezMez]

I'll get the MSI AB, I was under the assumption it's just used as an overclocking tool for GPUs. Shows how little I know. I'm gonna get it tonight and play with it, see if anything catches my attention.

 

[fzabkar]

Scammed and curious about the origin of this SSD?
https://forum.hddguru.com/viewtopic.php?p=273999

SSD tools (for repair and info, not data recovery):
http://www.hddoracle.com/viewtopic.php?f=106&t=2819

Be warned that some of the SSD tools are data destructive.

 

[mdd1963]

You are getting 500+MB/sec sequential reads and writes...?

Do we suspect someone snuck a Crucial MX500 into your Kingston box, or...

 

[MezMez]

Seems like I've got SM2258 controller, not the Phison PS3111-S11 which should be in A400S37.

v0.554a
Drive: 0(ATA)
OS: 10.0 build 18362
Model: KINGSTON SA400S37 240GB
Fw : R0430A0
Size : 228936 MB
From smart : [SMI2258XT] [R0430A0 02]
Controller : SM2258
FlashID: 0x45,0x3e,0x98,0xb3,0x76,0x72,0x0,0x0 - Sandisk 64L BiCS TLC 16k 512Gb/CE 512Gb/die 2Plane/die
Channel: 4
CE : 1
Plane : 2
Die/Ce : 1
Ch map : 0x0F
CE map : 0x01
Inter. : 1
First Fblock : 3
Total Fblock : 2958
Total Hblock : 3578
Fblock Per Ce : 2958
Fblock Per Die: 2958
Original Spare Block Count : 151
Vendor Marked Bad Block : 0
Bad Block From Pretest : 132

I guess that's definitive proof that it's a counterfeit?

@mdd1963 The results of my benchmarks are linked in the first post - write in seq is somewhere between 200-300MB, read is 500+.

 

[fzabkar]

I would contact Kingston and ask whether they switched flash controllers.

 

[MezMez]

I did, we shall await their reply.

 

[fzabkar]

SSD shadiness: Kingston and PNY caught bait-and-switching cheaper components after good reviews:

https://www.extremetech.com/extreme...itching-cheaper-components-after-good-reviews

Bait and Switch - The Sad State of SSDs:

https://www.tweaktown.com/blogs/Chris_Ramseyer/93/bait-and-switch-the-sad-state-of-ssds/index.html

 

[MezMez]

I've got the answer from their tech support.


"Our A400 SSD product line is build with 3D TLC Nand with variable BOMs.

Please find information on the specifications for this product on the data sheet on the following link:
https://www.kingston.com/datasheets/sa400_en.pdf"

I don't think this answers anything?

Edit: I've mailed them again, asking specific questions. Waiting for the reply.

 

[MezMez]

I've gotten a new reply. No straight answer.

I've mentioned the KSM issues and the different controllers, but I haven't mentioned performance issues yet. That's my bad.

" Could you please confirm if you are encountering any issues with your SA400S37/240G?


As previously stated our A400 SSD product line is build with 3D Nand with variable BOM's.

Please provide us with clear pictures of both sides of the SSD (front side all label information and back side the entire golden finger connector), as well as front and back of the packaging if available.

Please may we ask you to save any images in JPG format and to send them to us as e-mail attachments only - this means please do not send them from a smartphone or tablet PC. Thank you for your co-operation.

We will then investigate your request further and get back to you. "

I guess I have some image taking to do.

 

[fzabkar]

The variable Bill Of Materials statement seems to be suggesting that Kingston are entitled to change the internals of the product as long as it continues to meet the stated specs. IMO, a product which uses a different flash controller is essentially a different product. The other annoying thing about Kingston is that they sometimes rebrand flash controllers with their own part number, further obfuscating the specs.

 

[MezMez]

One more update, as a new reply came in.

"Please note that there are counterfeit products in circulation sold in authentic looking packaging.

In order to verify if this is an original or counterfeit product we will need to do a physical inspection on the SSD.


If you would like to arrange for us to verify this further, please reply to this mail providing the following details:

• A valid and up to date mailing/shipping address
• A daytime contact telephone number
• A proof of purchase of the product

Please may we ask you to save any images in JPG format and to send them to us as e-mail attachments only - this means please do not send them from a smartphone or tablet PC. Thank you for your co-operation.

As soon as we receive your reply we will investigate your request further. "

I guess it's simpler and cheaper to buy new SSD at this point, rather than shipping this one and waiting for it to come back.

 

[fzabkar]

I would have thought that Kingston could immediately verify whether your model was ever manufactured with a Silicon Motion controller. Furthermore, why won't they accept your photos of the PCB? Of course they would need to authorise you to open the case, but where's the problem with that?

 

[MezMez]

Yeah, the twice mentioned issue of the SMI controller got "our A400 SSD product line is build with 3D Nand with variable BOM's" reply.

About the latter, I think I'll get aggressive with my reply if their 'best and final solution' to the issue at hand is to ship the drive across the continent.

 

[ssdtechsupport]

Definitely fake. There is also a clue in the model number. There should not be a space here -> "SA400S37 240GB". Serial number is probably bogus too. There are A400 SSDs with SMI controllers though but I think all 2259.

 

[Saga Lout]

Welcome ssdtechsupport albeit with a necro post but the information looks as though it comes from the inside of the manufacturers premises.

 

[jeblad]

The message “Possible Counterfeit Drive Detected (CSMI Path)” might be reported due to the operating mode of the SATA controller. Check whether the BIOS has some kind of toggle to switch from IDE mode and to AHCI mode. It might be named differently.

Still note that changing the mode might create trouble while accessing the IDE drives, and that Win10 might have trouble booting. This can sometimes be fixed by starting Win10 in recovery mode. It does not seem to be necessary to do anything except booting up in recovery mode.