Phison E16 / Sabrent Rocket 4.0 Failures?

[Zarathustra[H]]

I'm curious,

Has anyone else experienced this?

I went to boot up this morning, and my 2TB Sabrent Rocket 4.0 drive was not detected in BIOS.

I keep all of my files stored on external (redundant and backed up NAS) so I haven't really lost anything crucial or irreplaceable (except game saves, which is annoying)

It looks like I am not alone. LOTS of failures reported for the Sabrent Rocket 4.0 drives in this thread on Reddit.

I'm about to reseat and/or pull the drive and see if I can rescue anything off of it, but judging from others experiences with these (in reddit link above) that is unlikely to help me. It seems like when they are gone, they are GONE.

I guess, take this thread as a little bit of a warning if you have one of these drives. Would also appreciate any suggestions when it comes to data rescue off of m.2 drives. I'm guessing the good old hard drive "freezer trick" is useless.

This is reminding me of the old OCZ days. Guess I should have stuck with my instincts and only used Intel and/or Samsung drives after all.

Wish me luck!

 

[Zarathustra[H]]

Surprisingly, with registration Sabrent offers a 5 year warranty on these things, so I can get another one, but considering the reported high failure rate, I'm wondering if I actually want to...

 

[learners permit]

Running 3 these in Raid 0 on my system with the oldest being over 2 years old and have experienced no issues other than much higher chipset and phison controller temperatures. Large file transfers started to get a bit sluggish on several ocassions so I started looking for the cause and found the controller temps were in the low 70's. I added some extra ghetto style cooling fans to help alleviate the issue and every thing has been fine since. The phison controllers on these drives generate quite a lot of heat that requires some sort of active cooling implementation and may be the cause of the failures being reported.

 

[Zarathustra[H]]

Running 3 these in Raid 0 on my system with the oldest being over 2 years old and have experienced no issues other than much higher chipset and phison controller temperatures. Large file transfers started to get a bit sluggish on several ocassions so I started looking for the cause and found the controller temps were in the low 70's. I added some extra ghetto style cooling fans to help alleviate the issue and every thing has been fine since. The phison controllers on these drives generate quite a lot of heat that requires some sort of active cooling implementation and may be the cause of the failures being reported.

Hmm.

I don't remember having elevated temps, but honestly, I haven't paid attention since I first installed it.

I would certainly have checked temps when it was new right after I installed it, and 70C would definitely have raised my eyebrows.

It is installed with the motherboards heatsink, without any active fans, except the nearby chipset fan, and whatever air the case fans pull over it.

I've never noticed any slowdowns. Today's failure to boot was the first indication of any trouble.

 

[Zarathustra[H]]

So, no suggestions on what I might do even temporarily to revive it and pull some data off of it?

You know like the good old temperature tricks? Heat it up to see if whatever connection that is marginal resäestavöishes contact?

Cool it down to see if whatever connection makes contact?

Reflow in an oven like we used to do with GPU's?

Thoughts?

 

[Zarathustra[H]]

Running 3 these in Raid 0 on my system with the oldest being over 2 years old and have experienced no issues other than much higher chipset and phison controller temperatures. Large file transfers started to get a bit sluggish on several ocassions so I started looking for the cause and found the controller temps were in the low 70's. I added some extra ghetto style cooling fans to help alleviate the issue and every thing has been fine since. The phison controllers on these drives generate quite a lot of heat that requires some sort of active cooling implementation and may be the cause of the failures being reported.

Hmm.

I don't remember having elevated temps, but honestly, I haven't paid attention since I first installed it.

I would certainly have checked temps when it was new right after I installed it, and 70C would definitely have raised my eyebrows.

It is installed with the motherboards heatsink, without any active fans, except the nearby chipset fan, and whatever air the case fans pull over it.

I've never noticed any slowdowns. Today's failure to boot was the first indication of any trouble.

Actually, yeah, I found some smartctl output I ran back when it was installed and operational in my system.

=== START OF INFORMATION SECTION ===
Model Number:                       Sabrent Rocket 4.0 2TB
Serial Number:                     
Firmware Version:                   RKT401.1
PCI Vendor/Subsystem ID:            0x1987
IEEE OUI Identifier:                0x6479a7
Total NVM Capacity:                 2,000,398,934,016 [2.00 TB]
Unallocated NVM Capacity:           0
Controller ID:                      1
Number of Namespaces:               1
Namespace 1 Size/Capacity:          2,000,398,934,016 [2.00 TB]
Namespace 1 Formatted LBA Size:     512
Namespace 1 IEEE EUI-64:            6479a7 2aa0933336
Local Time is:                      Thu Jan 13 19:27:09 2022 EST
Firmware Updates (0x12):            1 Slot, no Reset required
Optional Admin Commands (0x0017):   Security Format Frmw_DL Self_Test
Optional NVM Commands (0x005d):     Comp DS_Mngmt Wr_Zero Sav/Sel_Feat Timestmp
Maximum Data Transfer Size:         512 Pages
Warning  Comp. Temp. Threshold:     90 Celsius
Critical Comp. Temp. Threshold:     95 Celsius

Supported Power States
St Op     Max   Active     Idle   RL RT WL WT  Ent_Lat  Ex_Lat
 0 +     9.78W       -        -    0  0  0  0        0       0
 1 +     6.75W       -        -    1  1  1  1        0       0
 2 +     5.23W       -        -    2  2  2  2        0       0
 3 -   0.0490W       -        -    3  3  3  3     2000    2000
 4 -   0.0018W       -        -    4  4  4  4    25000   25000

Supported LBA Sizes (NSID 0x1)
Id Fmt  Data  Metadt  Rel_Perf
 0 +     512       0         2
 1 -    4096       0         1

=== START OF SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        46 Celsius
Available Spare:                    100%
Available Spare Threshold:          5%
Percentage Used:                    0%
Data Units Read:                    22,320,758 [11.4 TB]
Data Units Written:                 38,426,068 [19.6 TB]
Host Read Commands:                 127,189,541
Host Write Commands:                122,497,992
Controller Busy Time:               614
Power Cycles:                       370
Power On Hours:                     4,085
Unsafe Shutdowns:                   55
Media and Data Integrity Errors:    0
Error Information Log Entries:      2,185
Warning  Comp. Temperature Time:    0
Critical Comp. Temperature Time:    0

Error Information (NVMe Log 0x01, max 63 entries)
Num   ErrCount  SQId   CmdId  Status  PELoc          LBA  NSID    VS
  0       2185     0  0x100d  0x4004  0x028            0     0     -

Looks like it was only at 46C, but I don't think this was during a load test or anything.

Then again, I rarely have high loads on my SSD's

Looks like they have configured the warning temp for this device at 90C, and the critical temp at 95C. At that point, with just over 4000 power on hours it had spent 0 time over either the warning or critical temps.

With I could access it to pull that data now...

Either way, if it didn't overheat in 4085 hours of use, (which is about 6 months of use, power on time only, I shut this machine down when not using it) I doubt overheating was the issue, unless Sabrent effed up the solder like Nvidia did back on their GPU's 10-15 years ago.

 

[Zarathustra[H]]

Well,

I've tried the drive in a USB adapter, I've tried it in another motherboard, I even tried blocking off the data pins, and only allowing the power pins to make contact, leaving it on for 30 mins, disconnecting for 30 secs, leaving it on for 30 mins again, and then disconnecting for 30 secs, called the "power cycle method" that has been known to sometimes work for SSD's online, but no success.

Short of baking it to see if I can reflow the solder, this drive is almost certainly a goner.

It's a real shame. No files lost, but game progress and stored overclock settings will be a pain to get back. That and reinstalling and setting up everything the way I like it

I should have imaged the drives more often I guess, but I kind of felt like the bad old OCZ days of SSD's were over, and as long as I kept an eye on wear leveling counts in SMART data I'd be OK.

That may be the case for Samsung and Intel, but I guess not for Sabrent.

 

[Snowknight26]

That may be the case for Samsung and Intel, but I guess not for Sabrent.

Pretty sure most Phison-based SSDs are just reference design PCBs with each company's branding and firmware slapped on top. Sabrent doesn't actually manufacture them last I checked.

 

[trick0502]

I had a 1tb of these crap out after 9 months of use. Same as op, turned the pc on and it wasn’t detected. No warning sign of a failure. They were quick to replace it with a new one.

 

[D-EJ915]

Pretty sure most Phison-based SSDs are just reference design PCBs with each company's branding and firmware slapped on top. Sabrent doesn't actually manufacture them last I checked.

They are the reference design, intel consumer ssds are the only ones i've ever had fail on me lol. I think WD and Samsung are the only brands that actually design their own now.

 

[Zarathustra[H]]

I had a 1tb of these crap out after 9 months of use. Same as op, turned the pc on and it wasn’t detected. No warning sign of a failure. They were quick to replace it with a new one.

How has the replacement held up?

My 2TB lasted since 2019 before this happened, but a good chunk of that time my system was done due to stupid Gigabyte motherboard issues, so it wasn't in use, the rest of the time it was only in use a few hours a day.

I'm trying to decide if I actually want to RMA it or if I should just trash it. I don't want another one that is going to crap out on me

 

[philb2]

Hmm.

I don't remember having elevated temps, but honestly, I haven't paid attention since I first installed it.

I would certainly have checked temps when it was new right after I installed it, and 70C would definitely have raised my eyebrows.

It is installed with the motherboards heatsink, without any active fans, except the nearby chipset fan, and whatever air the case fans pull over it.

I've never noticed any slowdowns. Today's failure to boot was the first indication of any trouble.

I have used Hard Disk Sentinel Pro for a long time now, but don't usually look at it in detail. But today I checked the temps on my Sabrent drive. Same drive as the OP, but only 1 TB. Fortunately max temp never exceeds mid-40s, even though I have an ASUS board with a "heat sink" cover for the M.2 slots. I've had this Sabrent in use now for about 2 years now. Big improvement over the SSD it replaced for my C: and D: drives.

 

[trick0502]

How has the replacement held up?

My 2TB lasted since 2019 before this happened, but a good chunk of that time my system was done due to stupid Gigabyte motherboard issues, so it wasn't in use, the rest of the time it was only in use a few hours a day.

I'm trying to decide if I actually want to RMA it or if I should just trash it. I don't want another one that is going to crap out on me

The replacement has been fine. I use it as a steam drive now. If you can rma it I would.

 

[Zarathustra[H]]

The replacement has been fine. I use it as a steam drive now. If you can rma it I would.

That is good to know. Thank you. Maybe they only had a bad few runs or something like that, and the later ones are better...

It surprised me, but when I looked at the product page it says you get a 5 year warranty (but only if you register it)

Otherwise the warranty is 1 year.

I never registered mine, but I'm hopeful I can do it after the fact. It would be trivial to find the amazon receipt

 

[Zepher]

I just bought one of the gen 3 1tb drives for my dad recently, went ahead and registered for the longer warranty. Had no idea you had to do that.

 

[Zarathustra[H]]

Well, luckily Amazon had 2TB Samsung 980 Pro's with 1 day delivery, so I'll be reinstalling Ng everything today.

Going to RMA the Sabrent drive and use the replacement as a Steam games drive. That way if it is indeed a shitty drive and the replacement dies again, all I've lost is a night of downloading.

 

[Zarathustra[H]]

They are the reference design, intel consumer ssds are the only ones i've ever had fail on me lol. I think WD and Samsung are the only brands that actually design their own now.

Intel's SSD's have been up there as the best in the industry for me.

I still have some 320 series and some S3700 series as well as a PCIe 750 series that have seen TONS of hours and writes over the years and just continue to keep on chugging as reliably as ever.

I've also used some 30 Samsung 840 Pro, 850 EVO, 850 Pro, 860 EVO and 870 EVO's, as well as a 970 EVO I used as my boot drive for a while, but has now been moved to secondary duty in the desktop. I have beat these SSD's to absolute shit, using most of them in ways they were never designed for (high write server cache's, etc) over several years of 24/7 power on and I have still never had a Samsung orIntel SSD fail on me. They all keep on just working as some sort of miracle.

On the flip side I had several OCZ drives in the early SSD days, and 100% of those died within 2 years of use, including the RMA's sent out to replace them.

I've only used on Sabrent SSD, the Phison E16 "Rocket 4.0" one that this thread was about, and it lasted 2.5 years before dying.

So I guess over a combined 45+ Samsung and Intel SSD's most of which have been beat to shit I have never had a failure. Every other brand I've bought has died within 2-3 years (mostly OCZ but now also Sabrent)

This experience has only served to reinforce my experience with SSD's based on a rather large sample size for an individual customer, and that is, anything Samsung or Intel is absolutely bullet proof. Absolutely any other brand is viewed with suspicion.

 

[Zarathustra[H]]

Running 3 these in Raid 0 on my system with the oldest being over 2 years old and have experienced no issues other than much higher chipset and phison controller temperatures. Large file transfers started to get a bit sluggish on several ocassions so I started looking for the cause and found the controller temps were in the low 70's. I added some extra ghetto style cooling fans to help alleviate the issue and every thing has been fine since. The phison controllers on these drives generate quite a lot of heat that requires some sort of active cooling implementation and may be the cause of the failures being reported.

I have used Hard Disk Sentinel Pro for a long time now, but don't usually look at it in detail. But today I checked the temps on my Sabrent drive. Same drive as the OP, but only 1 TB. Fortunately max temp never exceeds mid-40s, even though I have an ASUS board with a "heat sink" cover for the M.2 slots. I've had this Sabrent in use now for about 2 years now. Big improvement over the SSD it replaced for my C: and D: drives.

This has me wondering if maybe I should leave the motherboard heatsink / heat spreader / whatever it is off. It's mostly just a flat piece of metal with some ridges, but not to the extent I would call it a "heat sink".

Maybe It would be better to just install them with those little m.2 heatsinks. that stick up off the slot.

For many this would be a problem due to the typical location of the primary m.2 slot underneath where a dual slot GPU would go, but I guess that is one of the benefits of water cooling, since my GPU is just a single slot wide.

 

[D-EJ915]

Intel's SSD's have been up there as the best in the industry for me.

I still have some 320 series and some S3700 series as well as a PCIe 750 series that have seen TONS of hours and writes over the years and just continue to keep on chugging as reliably as ever.

I've also used some 30 Samsung 840 Pro, 850 EVO, 850 Pro, 860 EVO and 870 EVO's, as well as a 970 EVO I used as my boot drive for a while, but has now been moved to secondary duty in the desktop. I have beat these SSD's to absolute shit, using most of them in ways they were never designed for (high write server cache's, etc) over several years of 24/7 power on and I have still never had a Samsung orIntel SSD fail on me. They all keep on just working as some sort of miracle.

On the flip side I had several OCZ drives in the early SSD days, and 100% of those died within 2 years of use, including the RMA's sent out to replace them.

I've only used on Sabrent SSD, the Phison E16 "Rocket 4.0" one that this thread was about, and it lasted 2.5 years before dying.

So I guess over a combined 45+ Samsung and Intel SSD's most of which have been beat to shit I have never had a failure. Every other brand I've bought has died within 2-3 years (mostly OCZ but now also Sabrent)

This experience has only served to reinforce my experience with SSD's based on a rather large sample size for an individual customer, and that is, anything Samsung or Intel is absolutely bullet proof. Absolutely any other brand is viewed with suspicion.

We switched over to using Samsung ones for new disk replacements or swaps for larger drives a while ago. Clients kind of get pissed when their drives die lol, maybe Intel consumer ones just hate full disk encryption I have no idea. (I think 480gb was the biggest available at the time, we had a lot of 120, 180 and 240 drives in the field) Datacenter ones are fine though, still have ones from like 8-10 years ago with a ton of sql server writes going strong today. I've got maybe 24? SN750 WD drives but most have light usage but no failures yet there.

Quick google brought this up lol https://www.reddit.com/r/sysadmin/comments/bawktf/anyone_else_been_having_a_lot_of_intel_ssd/

Had one on my personal vmware box that just went to shit by slowly corrupting everything lol, that was nice. Meanwhile I have a samsung pro drive from 2008 that someone spilled coke on and it still works fine now.

Bottom line is backup your data! I'm doing a restore now for my file server because playing around with oracle vm server (the old one) and apparently removing a repository deletes everything and I had some stuff in it, good times.

 

[philb2]

I Every other brand I've bought has died within 2-3 years (mostly OCZ but now also Sabrent)

This experience has only served to reinforce my experience with SSD's based on a rather large sample size for an individual customer, and that is, anything Samsung or Intel is absolutely bullet proof. Absolutely any other brand is viewed with suspicion.

Note to self: After 3 years, replace my current Sabrent with newer (larger) model as my boot drive. [That would be next May.] Repurpose current Sabrent as secondary drive, probably for swap and temp files.

I just put in a reminder in my calendar for May 1, 2023.