• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."

Question Silicon Power 1TB NVMe (SP001TBP34A80M28) Write Speed Issue

Aditya Kumar

Junior Member
Oct 16, 2018
2
0
11
I ran some benchmarks and I am getting very low write Seq. Q32T1 speed from few weeks in crystal disk mark. Originally Seq. Q32T1 write speed was always above 3000mbps when I bought it last year which is the advertised speed by company.

I am not sure what could be the issue. Maybe the SSD is fried ? Does anyone know?

Original Benchmark from last year:


Recent Bechmark :


------------------------------------------------------------------------------------------------------------------------------------------------------

I could send the SSD for RMA but problem is I don't want to reinstall everything. I have 2 external HDD lying around. So I can clone full disk image of my SSD into a USB HDD then when I get it back I can restore image on it. Can anyone let me know good software to do that.

Also I think first I need to create image of SSD onto an external HDD and then when I get it back I will have to install Windows on another spare HDD so I can restore that image back to the replaced SSD. It's very long and tedious process. Let me know if there is any easier method. I am not sure if its possible to restore the cloned image without installing any OS.
 

razel

Platinum Member
May 14, 2002
2,330
87
101
Your test size changed from 16GB to 2GB so it should be the same or faster. Luckily it only appears that writes are affected. Ensure TRIM has ran on it. Also ensure that nothing else was running in the background at the same time.

I honestly wouldn't worry about it myself, but looking at your post, you don't want it anymore since your gut reaction is to RMA.
 

Aditya Kumar

Junior Member
Oct 16, 2018
2
0
11
Your test size changed from 16GB to 2GB so it should be the same or faster. Luckily it only appears that writes are affected. Ensure TRIM has ran on it. Also ensure that nothing else was running in the background at the same time.

I honestly wouldn't worry about it myself, but looking at your post, you don't want it anymore since your gut reaction is to RMA.
Same result in 16GB also, only write has gone down. I ran trim already before posting here but it didn't help. Ran benchmark in safe mode also.
 

mikeford

Diamond Member
Jan 27, 2001
5,577
130
106
If you try to RMA, vendor is bound to have some qualifying questions/tests, or worse the dreaded, not covered by warranty. I'd ask and find out.
 

VirtualLarry

No Lifer
Aug 25, 2001
50,897
6,313
126
Welcome to "Steady-State" SSD performance, which is ALWAYS LOWER than "out of the box" performance.

Nothing is "wrong", there is no reason to RMA.

Just educate yourself on SSDs.

Btw, is that a DRAM-less drive? Does it have SLC cache? Did you give it 10-30 minutes to empty the SLC cache, before trying the write tests?
 

UsandThem

Elite Member
Super Moderator
May 4, 2000
14,564
5,136
146
The drive size (and percentage of the filled percentage) is also different.

The first one is 32%, and the 2nd one is 73%. Along with the test size changing, I'm not sure you are able to compare them and be confident there is anything wrong with the drive.

Edit:////

I just noticed this was also posted over at Tom's and LTT:
https://forums.tomshardware.com/threads/silicon-power-1tb-nvme-sp001tbp34a80m28-write-speed-issue.3634157/#post-21899446
https://linustechtips.com/main/topic/1230742-silicon-power-1tb-nvme-sp001tbp34a80m28-write-speed-issue/

Pretty much all the same suggestions/comments we are saying here. Although they posted a different partition size (253GB) at LTT, so who knows what the actual size of the drive is.



I say if you're concerned about the speed, set up an RMA with Silicon Power, and let them test the drive.
 
Last edited:

VirtualLarry

No Lifer
Aug 25, 2001
50,897
6,313
126
I say if you're concerned about the speed, set up an RMA with Silicon Power, and let them test the drive.
I'll add:

Consider doing a system image backup (using something like Macrium Reflect Free), to a (portable) external HDD of sufficient size to hold the image, and then "Secure Erase" the SSD, and then do a fresh install of Windows onto it, and then do some benchmarks. It should show nearly "factory fresh" performance, in theory. If it does NOT, then that may be cause for RMA.

Then "Secure Erase" again, and then restore the image backup to the SSD. Benchmark again. See if the scores don't go down a bit. 73% used is a fairly big percentage of space used up. Oftentimes, the SLC cache size is scaled to the available free space, so you may not have any SLC cache space available any more, as the drive tries to make room for your stuff. That could well explain the difference, 3000MB/sec writing to SLC cache, 1000MB/sec writing to TLC or QLC NAND directly, without SLC caching.

Edit: My "gut" feeling is that you are overreacting, a bit of "computer OCD", that your benchmark scores went down, without compensating for issues such as actual space occupied by user data on the SSD, and commensurate with that, the reduced or eliminated SLC cache size/space, and the resultant drop in benchmark performance.

It's a well-known fact that SSDs slow down when filled. There were some models, that if you got over 50% space filled, performance dropped BY HALF.
 

ASK THE COMMUNITY