Two M.2 NVMe Enclosure's Advertised as the same speed, yet one is significantly slower?

[HadesCore]

Some background info;

I own two M.2 NVMe SSD to USB 3.1 Gen 2 Type C Enclosures, they both advertise a 10Gpbs throughput, one is the Simplecom SE503 and the other is the Volans VL-UCM2.

(From now on, for time's sake, I'm going to refer to these as the Simplecom and Volans respectively)

Generally, in the Simplecom, I have a 256GB NVMe drive that came factory installed in my 2019 Razer blade 15, and a 1TB Crucial P1 NVMe drive in the Volans.

 

 

Here's the issue;

For the past few days, the Volans has been noticeably slower than the Simplecom, transfer speeds feel like first-gen SATA and even slower when it gets hot from sustained reads/writes. On top of that, it would regularly disconnect and reconnect from windows mid transfer causing failures to move game installations and having to start again.

So I tested to find the cause of these issues, which brought me to the conclusion that the Volans is the culprit.

 

 

Here's how I tested the enclosure;

I tried the same transfers on the same computer, with the same USB C cable plugged into the same port with the same NVMe drive (Crucial P1) in both the Simplecom and Volans. The Simplecom performed as expected, transferring at full speeds, getting just as hot as the Volans would and only slowing down marginally, if it all.

I didn't record exact transfer rate values but regardless, the tests made it obvious that the Volans was the weak link in the chain.

(I can't speak confidently as to whether its performed this way all along or only started to recently, as I've owned it for a few months now but only started using it full time in the past few days.)

 

 

The reason I'm posting this;

If someone has the technical skills to, by looking at the photos I've provided in the link below, be able to understand if there are differences between the two enclosure's PCB's that might indicate if the Volans is fundamentally slower and falsely advertised, or if the issue I'm having is software related, even if you'd owned the same model Volans I have and could tell me your experience with it.

Any help will be appreciated.

 

 

 

Link to each enclosure from where I bought it:

Volans https://www.scorptec.com.au/product/Hard-Drives-&-SSDs/Enclosures/76762-VL-UCM2?gclid=CjwKCAiAg9rxBRADEiwAxKDTukKZAczGIfiG5CxBC2jwdlGh8KeBHtG4jo7rBonPg1CmMNdCZThY3BoCI2IQAvD_BwE

Simplecom https://www.scorptec.com.au/product/Hard-Drives-&-SSDs/Enclosures/76042-SE503

 

 

 

[Kisai]

36 minutes ago, HadesCore said:

Some background info;

I own two M.2 NVMe SSD to USB 3.1 Gen 2 Type C Enclosures, they both advertise a 10Gpbs throughput, one is the Simplecom SE503 and the other is the Volans VL-UCM2.

(From now on, for time's sake, I'm going to refer to these as the Simplecom and Volans respectively)

Generally, in the Simplecom, I have a 256GB NVMe drive that came factory installed in my 2019 Razer blade 15, and a 1TB Crucial P1 NVMe drive in the Volans.

 

 

Here's the issue;

For the past few days, the Volans has been noticeably slower than the Simplecom, transfer speeds feel like first-gen SATA and even slower when it gets hot from sustained reads/writes. On top of that, it would regularly disconnect and reconnect from windows mid transfer causing failures to move game installations and having to start again.

So I tested to find the cause of these issues, which brought me to the conclusion that the Volans is the culprit.

 

 

Here's how I tested the enclosure;

I tried the same transfers on the same computer, with the same USB C cable plugged into the same port with the same NVMe drive (Crucial P1) in both the Simplecom and Volans. The Simplecom performed as expected, transferring at full speeds, getting just as hot as the Volans would and only slowing down marginally, if it all.

I didn't record exact transfer rate values but regardless, the tests made it obvious that the Volans was the weak link in the chain.

(I can't speak confidently as to whether its performed this way all along or only started to recently, as I've owned it for a few months now but only started using it full time in the past few days.)

 

 

The reason I'm posting this;

If someone has the technical skills to, by looking at the photos I've provided in the link below, be able to understand if there are differences between the two enclosure's PCB's that might indicate if the Volans is fundamentally slower and falsely advertised, or if the issue I'm having is software related, even if you'd owned the same model Volans I have and could tell me your experience with it.

Any help will be appreciated.

 

 

 

Link to each enclosure from where I bought it:

Volans https://www.scorptec.com.au/product/Hard-Drives-&-SSDs/Enclosures/76762-VL-UCM2?gclid=CjwKCAiAg9rxBRADEiwAxKDTukKZAczGIfiG5CxBC2jwdlGh8KeBHtG4jo7rBonPg1CmMNdCZThY3BoCI2IQAvD_BwE

Simplecom https://www.scorptec.com.au/product/Hard-Drives-&-SSDs/Enclosures/76042-SE503

 

 

 

Is there no TIM (Thermal Transfer material/pad) on the SSDs ? Both of those are PCIe x 4 devices, so I don't think it's the drives themselves. It looks like the chassis with the black PCB might have heat-sinking material on the board where as the blue PCB does not.

 

One thing I'll point out, and YMMV, the bigger the SSD, the faster it is. But that would only affect the peak speed, and would really only be noticed if the drive is >50% filled. The drive on the left is a Samsung PM981 which never has any firmware updates, and has hardware encryption, and typically found in Dell Precision laptops. The one on the right is a Crucial CT1000P1SSD08

 

Samsung:

2800 MB/s Max Read, 1100 MB/s Max Write

 

Crucial

2,000 MB/s Read, 1,700 MB/s Write

 

So those drives do not have comparable specs.

 

In fact if you look at the chassis PCB, they have the exact same reference layout and chips. They're only different in the last few mm closest to the USB port. So I would check for a firmware version.

 

[mariushm]

The adapters use same controller (jms583) so they should behave similarly. The only differences seem to be in the blue one having locations where leds can be installed (for activity, writing, standby etc etc)

 

I would look into the actual usb cable and I'd also look into issuing TRIM on the SSDs using some software or defragmenting program (ex O&O Defrag can force TRIM)

 

The Crucial P1 has 1 GB of RAM and uses QLC memory which is slow, but luckily it uses a portion of that memory in SLC mode, to cache writes.  The amount in SLC mode varies between 12 GB and 100 GB, depending how full the drive is. Think of it like 100 GB of SLC mode memory can be around 250-300 GB of QLC memory.

Let's say if the SSD is less than half full, it will have around 100 GB of SLC mode memory to cache writes. If you're 75% full, you probably have only around 30-50 GB of SLC memory available for caching writes.

 

The drive will take in data super fast until that cache SLC memory gets filled and then the write speeds go down to 2-300 MB/s ... you can see this in Anandtech review here (scroll down) : https://www.anandtech.com/show/13512/the-crucial-p1-1tb-ssd-review/7

With drive empty, they started writing and when they got to around 155 GB written, the write speeds went down to around 2-300 MB/s because the SLC buffer filled.

If you're not writing tens of GB of data often, the drive will be fine.  Issue that TRIM command from time to time to force the SSD to make memory "pages" available to be used in SLC mode or as write locations.

 

The Samsung SSD is MLC or TLC and should not have this behavior, the write speed should be more consistent throughout. But, if you have the Samsung software, that may use the ram in your computer to cache content or transfer in background to the SSD (you may see the copy reported as complete but in reality files are still transferred in background)

 

 

[HadesCore]

1 hour ago, Kisai said:

Is there no TIM (Thermal Transfer material/pad) on the SSDs ? Both of those are PCIe x 4 devices, so I don't think it's the drives themselves. It looks like the chassis with the black PCB might have heat-sinking material on the board where as the blue PCB does not.

 

One thing I'll point out, and YMMV, the bigger the SSD, the faster it is. But that would only affect the peak speed, and would really only be noticed if the drive is >50% filled. The drive on the left is a Samsung PM981 which never has any firmware updates, and has hardware encryption, and typically found in Dell Precision laptops. The one on the right is a Crucial CT1000P1SSD08

 

Samsung:

2800 MB/s Max Read, 1100 MB/s Max Write

 

Crucial

2,000 MB/s Read, 1,700 MB/s Write

 

So those drives do not have comparable specs.

 

In fact if you look at the chassis PCB, they have the exact same reference layout and chips. They're only different in the last few mm closest to the USB port. So I would check for a firmware version.

 

Thanks for the info,

There's no TIM on either enclosure, 

But as I've stated, I already know the Volans is the problem (That's the blue PCB) the drives themselves are working completely fine, I'm just hoping to find out why the no matter what drive I put into it, the blue PCB is significantly slower and unreliable

[HadesCore]

1 hour ago, mariushm said:

The adapters use same controller (jms583) so they should behave similarly. The only differences seem to be in the blue one having locations where leds can be installed (for activity, writing, standby etc etc)

That's good to know, with that knowledge I'd say it's safe to assume, the adapter is definitely not working correctly and not that I bought the wrong kind of adapter so I might try to see if I can dig up a receipt to return it.

1 hour ago, mariushm said:

 

I would look into the actual usb cable

I tested both adapters with the same drive and same USB cable as I stated, so it can't be either of them it's definitely the adapter itself.