Question Addlink SSD 2.0TB S95 M.2 2280 NVMe his data is strange

[roynany]

Addlink SSD 2.0TB S95 M.2 2280 NVMe his data is strange

all kinds of drives with high read and write speeds 7,000 / 6,800 MB / s

they all come with a new standard in the new NVMe 1.4 standard

the most stinking thing is that the writing rate is double the reading rate!
that means something stinks!
https://www.guru3d.com/news_story/addlink_s95_series_gets_nvme_1_4_compatibility.html

do not understand what the logic is in this ?
does the drive read the material and transfer to writing that can write at the maximum rate of reading ?
So why do I have a writing rate about 2 times the reading rate ?

there is no single road test for this drive, no one is checking its nature
?
although sold all over the world

 

[damian101]

It makes sense that the drive can't read the data as fast with a queue depth of only 1, compared to 8 in the first read speed result. Write speed isn't affected by that as much because of the write cache. Also, keep in mind that only 1GiB of data is written or read per run, which again has a positive effect on the write speed in the benchmark because of the write cache.
Looks all normal to me.

 

[roynany]

It makes sense that the drive can't read the data as fast with a queue depth of only 1, compared to 8 in the first read speed result. Write speed isn't affected by that as much because of the write cache. Also, keep in mind that only 1GiB of data is written or read per run, which again has a positive effect on the write speed in the benchmark because of the write cache.
Looks all normal to me.

this is not normal - there is no advantage here over Standard 1.3
What's the point of writing speed 10 times if reading speed is 1/10 !!
What determines actual writing ? speed is reading speed!

In laboratory tests there is no difference between standard 1.3 and 1.4
Note that there is no website that checks what the standard 1.4 is worth!
What are they afraid of?

 

[damian101]

this is not normal - there is no advantage here over Standard 1.3
What's the point of writing speed 10 times if reading speed is 1/10 !!
What determines actual writing ? speed is reading speed!

In laboratory tests there is no difference between standard 1.3 and 1.4
Note that there is no website that checks what the standard 1.4 is worth!
What are they afraid of?

What are you talking about? Who says 'Nvme 1.4 makes any significant performance difference for this kind of workload? I doubt the CrystalBench numbers would have looked much different with a similar Nvme 1.3 storage controller.
Nvme is just an open set of standards for communication with PCIe storage devices, that's it. With version 1.4 it got an update.

 

[roynany]

Nvme is just an open set of standards for communication with PCIe storage devices, that's it. With version 1.4 it got an update.

UPDATE to improve processes to bring about better productivity!
but there is nothing really new

 

[StrangerGuy]

NVME or not, SSD marketing has always been sleazy by putting technically true but practically worthless speeds at high queue depth numbers.

4K reads at QD1 is by far the most important metric and even good NVME drives struggle to break 70MB/s.

 

[UsandThem]

What are you talking about? Who says 'Nvme 1.4 makes any significant performance difference for this kind of workload? I doubt the CrystalBench numbers would have looked much different with a similar Nvme 1.3 storage controller.
Nvme is just an open set of standards for communication with PCIe storage devices, that's it. With version 1.4 it got an update.

It doesn't matter!