Can anyone recommend a specific NVMe drive for FreeBSD?

I was having trouble installing FreeBSD 14.0 on a Teamgroup TM8FP6256G0C101 250GB NVMe drive and it appears that this drive has issues with working with FreeBSD.
Does anyone have any recommendations of drives they are using now?
Is there a NVMe drive list somewhere that lists what people have successfully used with FreeBSD? I was unable to find one.

Thanks
 
There are some with buggy firmware which apparently can cause issues when using ZFS but it's not that common afaik.
Crucial P5 Plus works great and so does SanDisk Extreme PRO M.2 NVMe 3D SSD

 
What do you want to use it for? Is performance a factor or more about pricing?

Consumer-grade or enterprise?

Samsung and Micron work well for me on FreeBSD but Samsung seemed to lose their way a bit last year with buggy firmware and other issues.
 
Does anyone have any recommendations of drives they are using now?
I have made absolutely flawless experiences with Intel's SSDPEDMW012T4 over the years.
Not sure what the current equivalent is.
Code:
jbo@hydrogen01:~ % doas smartctl -a /dev/nvme1
Password:
smartctl 7.4 2023-08-01 r5530 [FreeBSD 13.2-RELEASE-p4 amd64] (local build)
Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Number:                       INTEL SSDPEDMW012T4
Serial Number:                      CVCQ542600AZ1P2BGN
Firmware Version:                   8EV10171
PCI Vendor/Subsystem ID:            0x8086
IEEE OUI Identifier:                0x5cd2e4
Controller ID:                      0
NVMe Version:                       <1.2
Number of Namespaces:               1
Namespace 1 Size/Capacity:          1,200,243,695,616 [1.20 TB]
Namespace 1 Formatted LBA Size:     512
Local Time is:                      Tue Feb  6 19:48:46 2024 UTC
Firmware Updates (0x02):            1 Slot
Optional Admin Commands (0x0006):   Format Frmw_DL
Optional NVM Commands (0x0006):     Wr_Unc DS_Mngmt
Log Page Attributes (0x02):         Cmd_Eff_Lg
Maximum Data Transfer Size:         32 Pages

Supported Power States
St Op     Max   Active     Idle   RL RT WL WT  Ent_Lat  Ex_Lat
 0 +    25.00W       -        -    0  0  0  0        0       0

Supported LBA Sizes (NSID 0x1)
Id Fmt  Data  Metadt  Rel_Perf
 0 +     512       0         2
 1 -     512       8         2
 2 -     512      16         2
 3 -    4096       0         0
 4 -    4096       8         0
 5 -    4096      64         0
 6 -    4096     128         0

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

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        28 Celsius
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    20%
Data Units Read:                    341,074,304 [174 TB]
Data Units Written:                 475,020,821 [243 TB]
Host Read Commands:                 7,219,946,800
Host Write Commands:                9,392,694,182
Controller Busy Time:               549
Power Cycles:                       427
Power On Hours:                     61,096
Unsafe Shutdowns:                   52
Media and Data Integrity Errors:    0
Error Information Log Entries:      3

Error Information (NVMe Log 0x01, 16 of 64 entries)
Num   ErrCount  SQId   CmdId  Status  PELoc          LBA  NSID    VS  Message
  0          3     1       -  0x400c      -            0     -     -  Internal Error
  1          2     1  0x003e  0x4500  0x000   4294967295     1     -  Write Fault
  2          1     1       -  0x400c      -            0     -     -  Internal Error

Self-tests not supported
Code:
jbo@hydrogen01:~ % doas smartctl -a /dev/nvme0
Password:
smartctl 7.4 2023-08-01 r5530 [FreeBSD 13.2-RELEASE-p4 amd64] (local build)
Copyright (C) 2002-23, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Number:                       INTEL SSDPEDMW012T4
Serial Number:                      CVCQ648400761P2JGN
Firmware Version:                   8EV10171
PCI Vendor/Subsystem ID:            0x8086
IEEE OUI Identifier:                0x5cd2e4
Controller ID:                      0
NVMe Version:                       <1.2
Number of Namespaces:               1
Namespace 1 Size/Capacity:          1,200,243,695,616 [1.20 TB]
Namespace 1 Formatted LBA Size:     512
Local Time is:                      Tue Feb  6 19:50:33 2024 UTC
Firmware Updates (0x02):            1 Slot
Optional Admin Commands (0x0006):   Format Frmw_DL
Optional NVM Commands (0x0006):     Wr_Unc DS_Mngmt
Log Page Attributes (0x02):         Cmd_Eff_Lg
Maximum Data Transfer Size:         32 Pages

Supported Power States
St Op     Max   Active     Idle   RL RT WL WT  Ent_Lat  Ex_Lat
 0 +    25.00W       -        -    0  0  0  0        0       0

Supported LBA Sizes (NSID 0x1)
Id Fmt  Data  Metadt  Rel_Perf
 0 +     512       0         2
 1 -     512       8         2
 2 -     512      16         2
 3 -    4096       0         0
 4 -    4096       8         0
 5 -    4096      64         0
 6 -    4096     128         0

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

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        28 Celsius
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    22%
Data Units Read:                    319,819,356 [163 TB]
Data Units Written:                 628,766,467 [321 TB]
Host Read Commands:                 7,202,323,913
Host Write Commands:                14,354,521,590
Controller Busy Time:               200
Power Cycles:                       107
Power On Hours:                     58,562
Unsafe Shutdowns:                   11
Media and Data Integrity Errors:    0
Error Information Log Entries:      0

Error Information (NVMe Log 0x01, 16 of 64 entries)
No Errors Logged

Self-tests not supported

For desktop builds, I usually reside with Samsung's 8xx or 9xx Pro series m.2 drives.
You want to be sure to do a firmware update on those tho.
 
TeamGroup is just a crappy brand. I tried one of their SSDs because it was cheap - and now I know why it was cheap.

OP should go with SSDs from Samsung, ADATA, Western Digital, Crucial... the same brands that are known for high quality USB sticks. Yeah, good brands cost a bit more, but you get what you pay for.

Heck, I compiled my way into KDE on a rig where I installed a Samsung 970 PRO 250 GB M.2 - and no complaints from the system. ?
 
I was having trouble installing FreeBSD 14.0 on a Teamgroup TM8FP6256G0C101 250GB NVMe drive and it appears that this drive has issues with working with FreeBSD.
Does anyone have any recommendations of drives they are using now?
Is there a NVMe drive list somewhere that lists what people have successfully used with FreeBSD? I was unable to find one.

What are the error messages?

I currently buy the WD SN850x line (when I don't go for U.2). I also have a FreeBSD install on a HP SSD showing up as "SSSTC CL1-8D256-HP L183", but I don't recall any of my other SSDs giving me trouble with FreeBSD. Hence my question about error messages.
 
Performance and pricing, the Sandisk Extreme Pro is "legacy" by now and the P5 Plus's SSDs where heavily discounted for quite some time here in europe frequently.
P5 Plus's for various tasks incl boot and the Sandisk is mainly abused for Poudriere builds.
The P5 Plus series have now been superseded by T500 series
 
Sadly, "known brand" alone is not enough to ensure quality. I have a pair of Samsung SSD 970 EVO Plus 500GB NVMe M.2 drives, and they are incapable of running at 100% duty cycle without risk of damage. Despite the addition of heat sinks and direct fans, they over-heat (rapidly exceed their 70C maximum operating temperature) under sustained load. Google tells me that I am not alone with this problem.
 
Sadly, "known brand" alone is not enough to ensure quality. I have a pair of Samsung SSD 970 EVO Plus 500GB NVMe M.2 drives, and they are incapable of running at 100% duty cycle without risk of damage. Despite the addition of heat sinks and direct fans, they over-heat (rapidly exceed their 70C maximum operating temperature) under sustained load. Google tells me that I am not alone with this problem.
Wow, what's your computing load/task??? what in the world are you doing so that something THAT good is stressed out THAT bad??? ?

I'm ready to run Poudriere on mine, overnight - except that in my case, I have a Ryzen 5 7600 on that rig, which definitely shortens compile time to the point that it not making sense to plan for an overnight job ?

?
 
what in the world are you doing so that something THAT good is stressed out THAT bad??? ?
Does it matter? The point of my post was that the Samsung SSD 970 EVO Plus is not THAT good. Sustained I/O of any kind is enough to overheat it. I added a heatsink to one and fan directly above the other (no clearance for a heatsink each). Mine still easily get well past 70C under any sustained I/O load, where throttling is invoked by the firmware. Last time I instrumented (with smartctl(8)), I stopped the tests as the temperature was still rising towards 80C.
I have a Ryzen 5 7600 on that rig
You posted above that you have "Samsung 970 PRO 250 GB M.2". That's not the same as "Samsung SSD 970 EVO Plus".
 
You posted above that you have "Samsung 970 PRO 250 GB M.2". That's not the same as "Samsung SSD 970 EVO Plus".
Yeah, and then I came home and looked at the box the SSD came in.

Does it matter?
It does. Was it compiling, mining, rendering, streaming, snort, encryption/decryption, or something else? And, if you pair a good SSD with a crappy celeron, it will be a slow job and the entire rig will overheat.
 
I had very good results with WD Black SSDs, Blue appear to be OK as well. Not so great results with Kingston, but that was not FreeBSD-specific. BUT: all reports here will be highly andecdotal, unless somebody is running thousands of these SSDs. So not sure if any answer here will help the OP. My general advice would be to stay away from cheap/unknown brands. Samsung, WD, Sabrent, Crucial - they all appear to work OK. All this is for consumer-grade SSDs which also work well in servers that have reasonably limited write cycles.
 
Back
Top