Question WD BLACK 1TB SN850X significantly slower than Lexar 1TB NM790

[exm]

I ran both Passmark and CrystalDiskMark. Specs: W11, ASRock Riptide Z790 with Intel i9-12900k CPU. 32 Gb ram, no workload. No GPU.

The read performance is similar, but the write performance is significantly slower on the WD drive.

WD 2619 Seq vs Lexar 5461

WD 156 Random vs Lexar 868

Thermal throttling doesn't seem to be an issue. Latest firmware for everything.

Here's the full data:

Round 1:
HW Info:

https://imgur.com/9ACMxYx

View: https://imgur.com/9ACMxYx


WD:

https://imgur.com/5s6e450

View: https://imgur.com/5s6e450


Lexar:

https://imgur.com/AYFHa4t

View: https://imgur.com/AYFHa4t


Round 2:
I made the recommendations suggested on Reddit, and performed some additional tests:

• The WD is definitely running X4 speed. WD Dashboard confirms this.
• WD Game mode enabled
• I have run tests where the WD is the boot drive, as well as the Lexar
• I also swapped the WD and Lexar physically on my motherboard as far as nvme slots are concerned

Lexar as boot (C: drive:

https://imgur.com/9hvuDnz

View: https://imgur.com/9hvuDnz
WD as boot (C: drive:

https://imgur.com/ZZ5rcV5

View: https://imgur.com/ZZ5rcV5


As you can see, not much difference. Swapped nvme slots on motherboard, Lexar as boot (C: drive:

https://imgur.com/9hvuDnz

View: https://imgur.com/9hvuDnz


Finally, some real life testing: Files copied from a secondary SSD (non-WD or Lexar)
18.8Gb, 1 file
WD 9.37 secs
Lexar 9.65 secs

164Gb,3.080 files, 129 folders
WD 134 secs
Lexar 121 secs

Files copied from a local drive (e.g. WD to WD)
18.8Gb, 1 file
WD 22.59 secs
Lexar 10.69 secs

164Gb,3.080 files, 129 folders
WD 387 secs
Lexar 167 secs

As you can see, the Lexar is much, much faster. I believe I've tried every possible scenario here.

Another side note, I found these images from the Amazon reviews:

SK hynix Platinum P41:

https://imgur.com/8J7SwEu

View: https://imgur.com/8J7SwEu


SAMSUNG 990 PRO:

https://imgur.com/xacKsTz

View: https://imgur.com/xacKsTz


The results I'm getting from the Lexar is better than ANYTHING else, but the WD Black definitely under-performs.

At this point I'm sticking with the Lexar, unless anyone can think of anything else to try?

 

[Maxxify]

Optimize (TRIM) before benchmarking. Also, boot to a live Linux environment and install/run KDiskMark, and see if matches with CDM. You can also secure erase/format (nvm format) the SN850X (be careful) in that environment if testing is still off. That rules out software and optimization or write caching issues. If there's still an issue, I'd point to hardware compatibility; the SN850 and SN850X had some compatibility issues in the past (albeit mostly on AMD systems) with firmware updates.

 

[Bob.B]

I ran both Passmark and CrystalDiskMark. Specs: W11, ASRock Riptide Z790 with Intel i9-12900k CPU. 32 Gb ram, no workload. No GPU.

The read performance is similar, but the write performance is significantly slower on the WD drive.

WD 2619 Seq vs Lexar 5461

WD 156 Random vs Lexar 868

Thermal throttling doesn't seem to be an issue. Latest firmware for everything.

Here's the full data:

Round 1:
HW Info:

https://imgur.com/9ACMxYx

View: https://imgur.com/9ACMxYx


WD:

https://imgur.com/5s6e450

View: https://imgur.com/5s6e450


Lexar:

https://imgur.com/AYFHa4t

View: https://imgur.com/AYFHa4t


Round 2:
I made the recommendations suggested on Reddit, and performed some additional tests:

• The WD is definitely running X4 speed. WD Dashboard confirms this.
• WD Game mode enabled
• I have run tests where the WD is the boot drive, as well as the Lexar
• I also swapped the WD and Lexar physically on my motherboard as far as nvme slots are concerned

Lexar as boot (C: drive:

https://imgur.com/9hvuDnz

View: https://imgur.com/9hvuDnz
WD as boot (C: drive:

https://imgur.com/ZZ5rcV5

View: https://imgur.com/ZZ5rcV5


As you can see, not much difference. Swapped nvme slots on motherboard, Lexar as boot (C: drive:

https://imgur.com/9hvuDnz

View: https://imgur.com/9hvuDnz


Finally, some real life testing: Files copied from a secondary SSD (non-WD or Lexar)
18.8Gb, 1 file
WD 9.37 secs
Lexar 9.65 secs

164Gb,3.080 files, 129 folders
WD 134 secs
Lexar 121 secs

Files copied from a local drive (e.g. WD to WD)
18.8Gb, 1 file
WD 22.59 secs
Lexar 10.69 secs

164Gb,3.080 files, 129 folders
WD 387 secs
Lexar 167 secs

As you can see, the Lexar is much, much faster. I believe I've tried every possible scenario here.

Another side note, I found these images from the Amazon reviews:

SK hynix Platinum P41:

https://imgur.com/8J7SwEu

View: https://imgur.com/8J7SwEu


SAMSUNG 990 PRO:

https://imgur.com/xacKsTz

View: https://imgur.com/xacKsTz


The results I'm getting from the Lexar is better than ANYTHING else, but the WD Black definitely under-performs.

At this point I'm sticking with the Lexar, unless anyone can think of anything else to try?

In device manager is write caching enabled for the WD?