Question *STABLE* NVMe - USB Adapter?

[cyberjedi]

I'm looking to buy a *RELIABLE* NVME-USB external adapter, that doesn't disconnect randomly, can keep up the close to 1GB/s speed.
So far all adapter's reviews contain frighteningly high % of reports of malfunction, random disconnects, sub-par speeds, falling to even under USB 2.0 speeds, etc.
There are 3 main chipsets in the market: JMicron JMS583, Asmedia ASM2362, and Realtek RTL9210 (no product has surfaced with the latter yet, so no REAL reviews, experiences yet UPDATE: it is on the market already).
Also ASUS has a new product, ROG Strix Arion, but there isn't enough feedback about it yet, and I mean real life feedback not just being able to run a few benchmarks, but real stress test to see if the device (adapter) can keep it up, without disconnecting or slowing down hard.
I would really appreciate any feedback from those who have GOOD experiences during stress-tests, like tons of random 4k writes with a stable speed, cloning 100's of GB-s at a stable high speed close to 1GB/s.
I know SSD's very well, so please stick to the point and don't start to educate me on how their speed falls after some writes, etc. I'm aware and this is NOT the point.
You can help if you can point me to a STABLE and RELIABLE NVMe-USB-C adapter that WORKS as expected.
Thank you in advance!

 

[Alpaca786]

@Jimminy That is quite a lot and indicates, that ASPM is not going into effect, as 1 W is typical idle power consumption without it enabled. Maybe try and and set ASPMDIS to 0x01 and measure it again. I would not trust those variable names 100 %.

 

[Jimminy]

Thanks @alpaca. I'm giving up on this POS RTL9210 for a while, but will keep monitoring this thread and hope for some fix.

In the meantime ... Has anyone tried the new firmware released by Jeyi? Released a few weeks ago.
Version 1.32.49?

For me, it was a real turd. Crashed the computer after being connected a while.

I'm beginning to think 10 gbps usb is just too flakey to trust.

I wonder how many times we can flash these chips before wearing them out ... LOL.

 

[PRSXFENG]

Thanks @alpaca. I'm giving up on this POS RTL9210 for a while, but will keep monitoring this thread and hope for some fix.

In the meantime ... Has anyone tried the new firmware released by Jeyi? Released a few weeks ago.
Version 1.32.49?

For me, it was a real turd. Crashed the computer after being connected a while.

I'm beginning to think 10 gbps usb is just too flakey to trust.

I wonder how many times we can flash these chips before wearing them out ... LOL.

Yeah 10Gbps is flakey

As for the flashchip
I poped open my enclosure and it has a FM25Q64
Datasheet says its good for 100,000 program/erase cycles so probably fine lol

 

[Jimminy]

Yeah 10Gbps is flakey

As for the flashchip
I poped open my enclosure and it has a FM25Q64
Datasheet says its good for 100,000 program/erase cycles so probably fine lol

Wow 100,000, That's a lot of writes. I wish there was a way to get these things to flash without having to remove the drive. That would speed up trying different things if I didn't have to remove and replace the drive each time.

 

[Shonky]

are there any firmware updates for ASM2364

I picked up an Integral UltimaProX2 USB 3.2 Gen 2×2

Its a turnkey Phison PU31-Pro that id's as Integral PU31-Pro

It's an ASM2364 bridge chip with a 256GB TLC Phison E13T

​

I had to fit the drive into my old Phison PU31 that has a JMS583 bridge chip
to dump the drive info as nandflash_id2 doesnt support the ASM2364 atm

Model   : PCIe SSD                               
Fw      : EDFMQ0.7
HMB     : 1640 - 279344 KB
Size    : 244198 MB [256.1 GB]
LBA Size: 512
AdminCmd: 0x00 0x01 0x02 0x04 0x05 0x06 0x08 0x09 0x0A 0x0C 0x10 0x11 0x14 0x18 0x80 0x81 0x82 0x84 0xD0 0xD1 0xD2 0xF4
I/O Cmd : 0x00 0x01 0x02 0x04 0x08 0x09
Firmware lock supported [02 01] [P001] [0100]
Drive unlocked [02 03]
F/W     : EDFMQ0.7
P/N     : 511-201208091  
Bank00: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Bank01: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Bank02: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Bank03: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Bank08: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Bank09: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Bank10: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Bank11: 0x98,0x3c,0x98,0xb3,0x76,0xe3,0x0,0x0 - Toshiba 96L BiCS4 TLC 16k 256Gb/CE 256Gb/die 2Plane/die
Controller    : PS5013-E13
CPU Clk       : 667
Flash CE      : 8
Flash Channel : 4
Interleave    : 2
Flash CE Mask : [++++---- ++++---- -------- --------]
Flash Clk,MT  : 800
Die per CE    : 1
Block per CE  : 1980
Page per Block: 1152
Bit Per Cell  : 3(TLC)
PMIC Type     : PS6103
PE Cycle Limit: 30000/3000
SLC cache     : 0x600000/0x340000

 

[Jimminy]

@Jimminy That is quite a lot and indicates, that ASPM is not going into effect, as 1 W is typical idle power consumption without it enabled. Maybe try and and set ASPMDIS to 0x01 and measure it again. I would not trust those variable names 100 %.

I think I figured it out. I never got good results with V 1.32 (it hangs or crashes the computer), so I went back to V1.31, which worked well, but wouldn't suspend during inactivity.

What finally worked was setting EN_U1U2 = 0x1. Apparently, this enables the USB sleep states, and is required as well as ASPMDIS = 0 .

I kept ASPMDIS = 0, and set a 5 minute timeout, DISK_IPS_THRES = 0x5.

Now, the measured current drops to zero when it sleeps. (I'm sure its not actually zero; my usb meter is a very cheap and crude instrument.)

 

[White7561]

I think I figured it out. I never got good results with V 1.32 (it crashes the computer), so I went back to V1.31, which worked well, but wouldn't suspend during inactivity. What finally worked was setting EN_U1U2 = 0x1. Apparently, this enables the USB sleep states.

I kept ASPMDIS = 0, and set a 5 minute timeout, DISK_IPS_THRES = 0x5.

Now, the measured current drops to zero when it sleeps. (I'm sure its not actually zero; my usb meter is a very cheap and crude instrument.)

Does it interfere with background garbage collection etc though? Since i do know that usually. They need it to be idle without any power saving mode to do their background thingy. Idk how true those are nowadays though.

 

[Jimminy]

Does it interfere with background garbage collection etc though? Since i do know that usually. They need it to be idle without any power saving mode to do their background thingy. Idk how true those are nowadays though.

@White7561 , I have no idea. I don't even know how to check on garbage collection. I know that Trim works, but checking it causes the drive to be active, and trim is not exactly the same thing anyway, from what little I know.

I know very little of how these nvme drives work internally. If you know some way to check on it, please let me know. I'm always happy to add to my meager knowledge.

 

[White7561]

@White7561 , I have no idea. I don't even know how to check on garbage collection. I know that Trim works, but checking it causes the drive to be active, and trim is not exactly the same thing anyway, from what little I know.

I know very little of how these nvme drives work internally. If you know some way to check on it, please let me know. I'm always happy to add to my meager knowledge.

Yeah idk for sure since those are processes that run internally. But yeah i think if trim works it should be fine but yeah idk if the process where it moves the data from the pSLC cache to the TLC drive (if it's a TLC drive) still runs when it's immediately on power saving mode.

 

[AlexMars]

In the meantime ... Has anyone tried the new firmware released by Jeyi? Released a few weeks ago.
Version 1.32.49?

I adapted this firmware for my device with a fan and backlight. Everything works well. No problem. But I always disable sleep mode in all firmware. Sleep mode is a compatibility issue across devices. I don't need him.

 

[AlexMars]

Yeah 10Gbps is flakey

For 1 year of use there were no problems. Everything works well.