SSD Thread: (its good, ignore the anime)

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > SSD Thread: (its good, ignore the anime)

Guigui: Jan 19, 2010; Winner of January '10 Lux Aeterna "Best 2010 Poster" Award

Not sure if this should go into a hardware discussion forum, but it's (somewhat) SSD related, so...

I found an old HP HP 2000-224 laptop in my parts bin, which I've ressurected (somewhat) by cannibalizing parts from other machines. The HDD is no longer functional, nor is the Wireless antenna card (Realtek RTL8188CE). I'm tempted to repurpose this machine solely to run pihole (given the cost these days of Raspberry pi-4's) and maaayyybe a micrecraft server for the kids...

Normally I'd go up and buy a cheap SSD and be done with it; but a buddy is selling a cheap 10$ mSATA 32 gig - which would be more than big enough to fit whatever linux OS + programs I'd put in.

Thing is, I have no clue if the... port?... the Realtek WLAN adapter fits would accomodate this mSATA chip, or if they would be interchangeable. The pins look very similar (it's not an M2 port, that's for sure). Looking up old HP sites hasn't helped for this 10-year old laptop, so I figure I'd come to the goons for thoughts on anything mSATA based.

(A similar picture, found in ifixit... https://guide-images.cdn.ifixit.com/igi/Es5jy6KrHFtxVLTQ.medium )

# ? Oct 12, 2022 20:32

Adbot: ADBOT LOVES YOU

# ? Apr 20, 2024 00:43

Thanks Ants: May 21, 2004; #essereFerrari

You can get a converter board that will let you put an mSATA SSD into a 2.5" SATA bay. 32GB sounds like it's eMMC though, which is going to perform like poo poo.

# ? Oct 12, 2022 20:41

Klyith: Aug 3, 2007; GBS Pledge Week

Guigui posted:

Thing is, I have no clue if the... port?... the Realtek WLAN adapter fits would accomodate this mSATA chip, or if they would be interchangeable. The pins look very similar (it's not an M2 port, that's for sure). Looking up old HP sites hasn't helped for this 10-year old laptop, so I figure I'd come to the goons for thoughts on anything mSATA based.

Mini PCIe and mSata share the same slot type and electrical compatibility, the slot can accommodate both. But only if the sata part is hooked up. Absent any documentation that says yea or nay, the only way to know is to try it. (I'd generally assume it won't work on an old cheap laptop. Those traces cost a tenth of a cent!)

This confusion is why they made m.2, which is also confusing but at least tells you useful info via the slot type some of the time.

Thanks Ants posted:

You can get a converter board that will let you put an mSATA SSD into a 2.5" SATA bay. 32GB sounds like it's eMMC though, which is going to perform like poo poo.

Plus a cheapo 128gb 2.5" ssd is probably cheaper than the $10 msata card plus the adapter.

# ? Oct 12, 2022 20:55

Flipperwaldt: Nov 11, 2011; Won't somebody think of the starving hamsters in China?

Plus you can recycle a sata ssd in some other project later. Anything else that ever contains an msata slot is going to be of an extremely similar vintage as that current laptop.

# ? Oct 12, 2022 21:22

Farmer Crack-Ass: Jan 2, 2001; this is me posting irl

I've heard that SSDs can't be trusted to retain data when powered off for extended periods of time. How long would a system need to be powered off for before you would expect to start to see data degradation?

# ? Oct 17, 2022 14:54

BIG HEADLINE: Jun 13, 2006; "Stand back, Ottawan ruffian, or face my lumens!"

Farmer Crack-rear end posted:

I've heard that SSDs can't be trusted to retain data when powered off for extended periods of time. How long would a system need to be powered off for before you would expect to start to see data degradation?

I think it's overstated. I found and plugged in an old 128GB 830 Pro about two years ago after it had been disconnected from power for a good five years and it still had accessible, uncorrupted data on it.

# ? Oct 17, 2022 15:02

Klyith: Aug 3, 2007; GBS Pledge Week

Farmer Crack-rear end posted:

I've heard that SSDs can't be trusted to retain data when powered off for extended periods of time. How long would a system need to be powered off for before you would expect to start to see data degradation?

tl;dr it depends on many factors but you probably don't have to worry about it in less than 1-2 years.

Some of those factors:

How warm was the flash while you were writing, and how cool is the storage. NAND likes to be warm while operating and cool in storage. Do the opposite, to an unusual degree, and bitflips can happen in mere weeks. (Another reason why big nvme heatsinks are dumb in a normal pc.)

Is the drive one of the models with one of the last generations of 2d transistors? Those have weaker endurance and worse retention.

What level of MLC is used? More bits per cell = less drift for an error. AFAIK this is a pretty weak effect though.

How full is the drive? If your OS drive is let's say having 100 errors per year and you only have 1gb of data you care for on a 1tb drive, it's only a 10% chance that one of those errors is on your important data. And many formats have some redundancy: a single-bit error on most videos is likely unnoticeable.

# ? Oct 17, 2022 18:24

BobHoward: Feb 13, 2012; The only thing white people deserve is a bullet to their empty skull

The most important power off retention factor is simply how many write/erase cycles the flash media accumulated before you shut it down and put it on the shelf.

Flash memory wears out because the insulation trapping charge in a cell gets leakier each time the cell is erased and programmed. If you take two identical SSDs out of the box and write torture test only one till it's almost dead, the fresh drive will retain data much longer than the worn drive.

Klyith posted:

What level of MLC is used? More bits per cell = less drift for an error. AFAIK this is a pretty weak effect though.

No, not weak at all. It's fundamentally the same reason why it's possible to use a region of TLC/QLC flash in SLC mode as write cache. That cache region accumulates P/E cycles at an impressive rate compared to the rest of the media, but you can tolerate far worse leakage when reads only need to sense two levels. (Also helps that the SLC write cache isn't expected to need much retention.) There's a reason why 2-bit MLC P/E life ratings are generally at least a factor of 10 worse than SLC.

So... don't use old SSDs which spent years being thrashed by swapping as offline backup, I guess? If you really want to shove a SSD in a box and read data five or ten years later, it's definitely best if that SSD is as few bits per cell as possible and has as few write cycles as possible.

# ? Oct 18, 2022 04:24

Harik: Sep 9, 2001; From the hard streets of Moscow
First dog to touch the stars; Plaster Town Cop

For 2.5" SATA to upgrade an older machine, is there any drawback to dramless TLC? The big bottleneck is going to be the SATA interface but does lack of DRAM matter at those slower speeds?

# ? Oct 18, 2022 06:15

Klyith: Aug 3, 2007; GBS Pledge Week

Harik posted:

For 2.5" SATA to upgrade an older machine, is there any drawback to dramless TLC? The big bottleneck is going to be the SATA interface but does lack of DRAM matter at those slower speeds?

For SATA drives the dramless penalty is much worse than NVMe. Dramless NVMe drives use a bit of system ram for caching (portions of) their map table. SATA drives can't do that. Dramless SATA often puts a bit of memory on the controller itself to help compensate, but not a lot.

How bad / noticeable your dramless penalty is depends on the application. On an old PC where you just want any SSD to replace an old HDD, it's still a massive upgrade. Gaming is probably also fine -- games tend to be fairly sequential, which is where dramless has the least effect.

BobHoward posted:

No, not weak at all. It's fundamentally the same reason why it's possible to use a region of TLC/QLC flash in SLC mode as write cache. That cache region accumulates P/E cycles at an impressive rate compared to the rest of the media, but you can tolerate far worse leakage when reads only need to sense two levels. (Also helps that the SLC write cache isn't expected to need much retention.) There's a reason why 2-bit MLC P/E life ratings are generally at least a factor of 10 worse than SLC.

I don't think this is exactly right. The pSLC writes produce far less wear on the cell than higher levels. I think because they only use the lowest charge level, less charge means less strain on the trap. And many drives, especially QLC ones, move their pSLC cache around to help distribute wear on the drive. Not because the pSLC has more wear, but because it's less. When QLC first came out the thing that got attention was the 14 writes per cell of endurance. And finally, the QLC drives that use expanding pSLC space will keep data in the SLC area for a long time.

USB flash sticks are generally using the highest number of bits per cell they can, and they spend a lot of time unpowered. Though they don't get as much wear as an internal drive.

BobHoward posted:

So... don't use old SSDs which spent years being thrashed by swapping as offline backup, I guess?

Yeah. And maybe just don't use SSDs as cold-storage offline backup, in general.

And for people that are putting their PC into storage for 2 years because they've joined the peace corps or something, I would think about things like moving all your data onto HDDs / the cloud.

# ? Oct 18, 2022 13:51

Farmer Crack-Ass: Jan 2, 2001; this is me posting irl

In my case, I've got a project system that I installed an OS on a few months ago, and then haven't been arsed to do anything further with; I didn't think it was likely that just a few months would lead to problems, and I could "easily" just reinstall at this stage anyway, but i did have the thought of "hmm, at what point have I let this thing sit too long that I really do need to wipe and reinstall"

One other question out of curiosity - suppose you've got a worn SSD that spent long enough unpowered that it lost data. Is the flash media itself damaged at that point, or would the drive still behave properly?

# ? Oct 19, 2022 15:29

Klyith: Aug 3, 2007; GBS Pledge Week

Farmer Crack-rear end posted:

One other question out of curiosity - suppose you've got a worn SSD that spent long enough unpowered that it lost data. Is the flash media itself damaged at that point, or would the drive still behave properly?

No, the drive itself is still 100% functional. If you wipe the drive* you can go on using it for however much life the drive has left.

It will probably even appear to be fine, like nothing has happened. The only problem is that a small amount of the data on the drive is corrupt. This may be detected by the drive: reported uncorrectable errors or bad sectors when it tries to read stuff. Or it may just be invisible bitflips.

(As an aside, it's unlikely that a normal user will ever have a "worn" drive -- SSDs have vastly more endurance than most desktop systems can consume.)

*the best way to wipe SSDs is with an ATA Secure Erase, which you can get either with the drive manufacturer utility or a linux boot stick. Even if you don't care about the "secure" aspect, for SSDs it has benefits for drive performance and possibly health.

# ? Oct 19, 2022 16:09

mobby_6kl: Aug 9, 2009; by Fluffdaddy

I'll probably be building a new PC finally, but my current one has a relatively recent 512GB WD Blue SATA SSD in it which is working fine. I'm sure the newer stuff is faster, but it's not going to be super obvious, is it? As long as I'm not running database benchmarks or something.

# ? Oct 21, 2022 15:26

Klyith: Aug 3, 2007; GBS Pledge Week

mobby_6kl posted:

I'll probably be building a new PC finally, but my current one has a relatively recent 512GB WD Blue SATA SSD in it which is working fine. I'm sure the newer stuff is faster, but it's not going to be super obvious, is it? As long as I'm not running database benchmarks or something.

Yep that's still generally the case. A few games will load a few seconds faster with a nvme drive, but nothing really needs a nvme drive today. And you'd be hard pressed to tell the difference in a blind test, for anything besides file copies and such.

# ? Oct 21, 2022 15:58

Thanks Ants: May 21, 2004; #essereFerrari

I buy NVMe drives because SATA cables seem to like going bad, but I wouldn't be replacing a working SATA SSD with one.

# ? Oct 21, 2022 16:10

priznat: Jul 7, 2009; Let's get drunk and kiss each other all night.

Thanks Ants posted:

I buy NVMe drives because SATA cables seem to like going bad, but I wouldn't be replacing a working SATA SSD with one.

This to me is their biggest advantage, m.2 drives just slapped on the motherboard make the cases just slightly tidier!

# ? Oct 21, 2022 16:57

WhyteRyce: Dec 30, 2001

you guys aren't buying U.2 NVMe drives?!

# ? Oct 21, 2022 17:00

priznat: Jul 7, 2009; Let's get drunk and kiss each other all night.

WhyteRyce posted:

you guys aren't buying U.2 NVMe drives?!

Lol AIC NVMe is less of a hassle for anyone with consumer motherboards for sure. (I know you�re kidding!)

I would want a consumer case (like fractal etc) that has a nvme u.2/edsff drive bays (4 would be good!) with hotplug that would rule though.

# ? Oct 21, 2022 17:54

WhyteRyce: Dec 30, 2001

I'm waiting for some coked out engineer at asrock or whatever find some way to shove 8 E1.S slots onto a mini-itx board. Or at least replace all the M.2 slots with something that will take 9.5mm E1.S drives

For U.2, mobo makers could actually start putting mini-sas (8644) connectors on the board which we can easily use a U.2 drive with. Given the weird ways mobo makers try to stand out I'm surprised that's not more of a thing and would be more of a product differentiator for enthusiasts than some red colored 6 port sata tower

WhyteRyce fucked around with this message at 18:14 on Oct 21, 2022

# ? Oct 21, 2022 18:07

priznat: Jul 7, 2009; Let's get drunk and kiss each other all night.

WhyteRyce posted:

I'm waiting for some coked out engineer at asrock or whatever find some way to shove 8 E1.S slots onto a mini-itx board. Or at least replace all the M.2 slots with something that will take 9.5mm E1.S drives

For U.2, mobo makers could actually start putting mini-sas (edsff 8644) connectors on the board which we can easily use a U.2 drive with.

Some mobos already have those from supermicro and asrock rack etc! Just not many consumers get em. And then you still got cables, boo cables.

A bunch of E3S bays would be the shiiiit.

The thing I hate the most about m.2 is that loving tiny mounting screw. Solve that like how e1.s did (iirc) and it is the best solution!

# ? Oct 21, 2022 18:14

WhyteRyce: Dec 30, 2001

I believe the m.2 spec specifically calls out the mounting screw must be grounded, so those nifty looking nylon push/twist setups are actually bad (in theory).

I'm not too familiar with what E1.S specifies other than maybe where the screw holes for the latches can be mounted on to. The latches themselves should be specific to the server you are using. I've used some really lovely latches and stopped bothering to use them on said systems

WhyteRyce fucked around with this message at 18:24 on Oct 21, 2022

# ? Oct 21, 2022 18:22

priznat: Jul 7, 2009; Let's get drunk and kiss each other all night.

WhyteRyce posted:

I believe the m.2 spec specifically calls out the mounting screw must be grounded, so those nifty looking nylon push/twist setups are actually bad (in theory).

I'm not too familiar with what E1.S specifies other than maybe where the screw holes for the latches can be mounted on to. The latches themselves should be specific to the server you are using. I've used some really lovely latches and stopped bothering to use them on said systems

Ah that explains why more manufacturers didn't use them, the nylon ones on Supermicros were great. They must just YOLO it.

# ? Oct 21, 2022 19:01

Potato Salad: Oct 23, 2014; nobody cares

priznat posted:

Ah that explains why more manufacturers didn't use them, the nylon ones on Supermicros were great. They must just YOLO it.

strong case for not touching the drat thing powered on

# ? Oct 21, 2022 19:19

Saukkis: May 16, 2003; Unless I'm on the inside curve pointing straight at oncoming traffic the high beams stay on and I laugh at your puny protest flashes.
I am Most Important Man. Most Important Man in the World.

They could have used grounded SO-DIMM latches.

# ? Oct 22, 2022 17:53

Harik: Sep 9, 2001; From the hard streets of Moscow
First dog to touch the stars; Plaster Town Cop

Thanks, I grabbed a cheap teamgroup vulcan Z SATA since it's for an older machine.

WhyteRyce posted:

you guys aren't buying U.2 NVMe drives?!

I am. 12-bay hotswap NVMe, all direct-to-cpu thanks to epyc. it's v. fast even with older 4tb u.2 drives. glorious iops for everyone.

Anecdatapoint: DPG SX8200 Pro 2TB died at around 18 months. 90 TBW before it started taking errors. That seems ... a bit early? I was using it more than most do with 45 erase cycles but the smart data didn't say it was bad.

code:

=== START OF INFORMATION SECTION ===
Model Number:                       ADATA SX8200PNP
Serial Number:                      2L0929AA6JYR
Firmware Version:                   42B9T6TB
PCI Vendor/Subsystem ID:            0x1cc1
IEEE OUI Identifier:                0x707c18
Controller ID:                      1
NVMe Version:                       1.3
Number of Namespaces:               1
Namespace 1 Size/Capacity:          2,048,408,248,320 [2.04 TB]
Namespace 1 Utilization:            1,694,576,869,376 [1.69 TB]
Namespace 1 Formatted LBA Size:     512
Namespace 1 IEEE EUI-64:            707c18 00002092aa
Local Time is:                      Sat Oct 29 22:00:43 2022 EDT
Firmware Updates (0x14):            2 Slots, no Reset required
Optional Admin Commands (0x0017):   Security Format Frmw_DL Self_Test
Optional NVM Commands (0x005f):     Comp Wr_Unc DS_Mngmt Wr_Zero Sav/Sel_Feat Timestmp
Log Page Attributes (0x0f):         S/H_per_NS Cmd_Eff_Lg Ext_Get_Lg Telmtry_Lg
Maximum Data Transfer Size:         64 Pages
Warning  Comp. Temp. Threshold:     75 Celsius
Critical Comp. Temp. Threshold:     80 Celsius

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

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        38 Celsius
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    8%
Data Units Read:                    55,144,738 [28.2 TB]
Data Units Written:                 182,457,896 [93.4 TB]
Host Read Commands:                 1,098,800,082
Host Write Commands:                1,581,754,121
Controller Busy Time:               22,262
Power Cycles:                       49
Power On Hours:                     13,262
Unsafe Shutdowns:                   26
Media and Data Integrity Errors:    115
Error Information Log Entries:      27
Warning  Comp. Temperature Time:    0
Critical Comp. Temperature Time:    0
Thermal Temp. 1 Transition Count:   57
Thermal Temp. 1 Total Time:         321

Error Information (NVMe Log 0x01, 16 of 256 entries)
Num   ErrCount  SQId   CmdId  Status  PELoc          LBA  NSID    VS
  0         27     1  0x007d  0x4502      -    792809553     1     -
  1         26     1  0x007c  0x4502      -    792809552     1     -
  2         25     1  0x007b  0x4502      -    792809553     1     -
  3         24     1  0x007a  0x4502      -    792809552     1     -
  4         23     1  0x0079  0x4502      -    792809552     1     -
  5         22     1  0x001e  0x4502      -    792755289     1     -
  6         21     1  0x001d  0x4502      -    792755288     1     -
  7         20     1  0x001c  0x4502      -    792755289     1     -
  8         19     1  0x001b  0x4502      -    792755288     1     -
  9         18     1  0x001a  0x4502      -    792755288     1     -
 10         17     1  0x001e  0x4502      -    792664033     1     -
 11         16     1  0x001d  0x4502      -    792664032     1     -
 12         15     1  0x001c  0x4502      -    792664033     1     -
 13         14     1  0x001b  0x4502      -    792664032     1     -
 14         13     1  0x001a  0x4502      -    792664032     1     -
 15         12     1  0x002d  0x4502      -    792504201     1     -
... (240 entries not read)

Lolling that it considers itself healthy while reads keep stalling and it had unrecoverable errors on a couple sectors. This is why I've got nightly backups.

E: Rated for 1280 TBW, which I'm not remotely close to.

E2: Replaced with samsung 980 pro. Waste of my weekend cloning the disk. I don't backup cccache (and other caches) for obvious reasons but i'd prefer to not do 24 hours of builds to get it populated again.

Harik fucked around with this message at 03:34 on Oct 30, 2022

# ? Oct 30, 2022 03:23

WhyteRyce: Dec 30, 2001

The is healthy check is probably just checking if the critical warning field >0

A lot can die on a flash drive before a critical warning gets flagged

# ? Oct 30, 2022 05:47

Harik: Sep 9, 2001; From the hard streets of Moscow
First dog to touch the stars; Plaster Town Cop

Agreed, which is why I'm restoring the lost files from nightly backup instead of raging that they're gone. Now I just need to extend that backup some more since i see some things i could have potentially lost that weren't critical but would have annoyed me to rebuild.

Sadly the clone wasn't viable, too much metadata corruption and it was slowly eating itself. Wiped it again, re-copied the less important stuff over and letting the backup restore overnight.

Going to RMA it and see if they give a reason why it died so soon. It's got a 5-year on it and I've only had it since last april.

# ? Oct 31, 2022 04:56

redeyes: Sep 14, 2002; by Fluffdaddy

Harik posted:

Thanks, I grabbed a cheap teamgroup vulcan Z SATA since it's for an older machine.

I am. 12-bay hotswap NVMe, all direct-to-cpu thanks to epyc. it's v. fast even with older 4tb u.2 drives. glorious iops for everyone.

Anecdatapoint: DPG SX8200 Pro 2TB died at around 18 months. 90 TBW before it started taking errors. That seems ... a bit early? I was using it more than most do with 45 erase cycles but the smart data didn't say it was bad.

code:

=== START OF INFORMATION SECTION ===
Model Number:                       ADATA SX8200PNP
Serial Number:                      2L0929AA6JYR
Firmware Version:                   42B9T6TB
PCI Vendor/Subsystem ID:            0x1cc1
IEEE OUI Identifier:                0x707c18
Controller ID:                      1
NVMe Version:                       1.3
Number of Namespaces:               1
Namespace 1 Size/Capacity:          2,048,408,248,320 [2.04 TB]
Namespace 1 Utilization:            1,694,576,869,376 [1.69 TB]
Namespace 1 Formatted LBA Size:     512
Namespace 1 IEEE EUI-64:            707c18 00002092aa
Local Time is:                      Sat Oct 29 22:00:43 2022 EDT
Firmware Updates (0x14):            2 Slots, no Reset required
Optional Admin Commands (0x0017):   Security Format Frmw_DL Self_Test
Optional NVM Commands (0x005f):     Comp Wr_Unc DS_Mngmt Wr_Zero Sav/Sel_Feat Timestmp
Log Page Attributes (0x0f):         S/H_per_NS Cmd_Eff_Lg Ext_Get_Lg Telmtry_Lg
Maximum Data Transfer Size:         64 Pages
Warning  Comp. Temp. Threshold:     75 Celsius
Critical Comp. Temp. Threshold:     80 Celsius

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

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        38 Celsius
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    8%
Data Units Read:                    55,144,738 [28.2 TB]
Data Units Written:                 182,457,896 [93.4 TB]
Host Read Commands:                 1,098,800,082
Host Write Commands:                1,581,754,121
Controller Busy Time:               22,262
Power Cycles:                       49
Power On Hours:                     13,262
Unsafe Shutdowns:                   26
Media and Data Integrity Errors:    115
Error Information Log Entries:      27
Warning  Comp. Temperature Time:    0
Critical Comp. Temperature Time:    0
Thermal Temp. 1 Transition Count:   57
Thermal Temp. 1 Total Time:         321

Error Information (NVMe Log 0x01, 16 of 256 entries)
Num   ErrCount  SQId   CmdId  Status  PELoc          LBA  NSID    VS
  0         27     1  0x007d  0x4502      -    792809553     1     -
  1         26     1  0x007c  0x4502      -    792809552     1     -
  2         25     1  0x007b  0x4502      -    792809553     1     -
  3         24     1  0x007a  0x4502      -    792809552     1     -
  4         23     1  0x0079  0x4502      -    792809552     1     -
  5         22     1  0x001e  0x4502      -    792755289     1     -
  6         21     1  0x001d  0x4502      -    792755288     1     -
  7         20     1  0x001c  0x4502      -    792755289     1     -
  8         19     1  0x001b  0x4502      -    792755288     1     -
  9         18     1  0x001a  0x4502      -    792755288     1     -
 10         17     1  0x001e  0x4502      -    792664033     1     -
 11         16     1  0x001d  0x4502      -    792664032     1     -
 12         15     1  0x001c  0x4502      -    792664033     1     -
 13         14     1  0x001b  0x4502      -    792664032     1     -
 14         13     1  0x001a  0x4502      -    792664032     1     -
 15         12     1  0x002d  0x4502      -    792504201     1     -
... (240 entries not read)

Adata uses poo poo-tier nand.

# ? Oct 31, 2022 21:01

Harik: Sep 9, 2001; From the hard streets of Moscow
First dog to touch the stars; Plaster Town Cop

what about sk hynix P31 gold?

that drive would be much more annoying to recover so I don't want to be doing it every 18 months.

# ? Oct 31, 2022 22:38

Klyith: Aug 3, 2007; GBS Pledge Week

Harik posted:

what about sk hynix P31 gold?

good drive, uses nand from SK hynix themselves

If it's priced well in your area it's a good pick, but it looks like it's being discontinued -- in the US it seems mostly gone from online stores and the prices are way too high. Though possibly it's just gonna be rebranded as this new 'Solidigm' thing (god what a terrible name).

# ? Oct 31, 2022 22:59

mobby_6kl: Aug 9, 2009; by Fluffdaddy

So I went on aliexpress as I usually do on Sundays, and I found this:

https://www.aliexpress.com/item/1005003819383727.html

Perfect for all your SSD cooling needs.

# ? Nov 7, 2022 01:55

BIG HEADLINE: Jun 13, 2006; "Stand back, Ottawan ruffian, or face my lumens!"

mobby_6kl posted:

Perfect for all your SSD cooling needs.

https://www.corsair.com/us/en/Categ...p/CX-9029002-WW

And that's not the only one out there.

# ? Nov 7, 2022 03:15

SlayVus: Jul 10, 2009; Grimey Drawer

BIG HEADLINE posted:

https://www.corsair.com/us/en/Categ...p/CX-9029002-WW

And that's not the only one out there.

Really though, cooling is only needed for the controller chip. The NAND chips work better when ran warm.

# ? Nov 7, 2022 03:44

BIG HEADLINE: Jun 13, 2006; "Stand back, Ottawan ruffian, or face my lumens!"

There have been arguments that heatsinking the NAND can transfer *more* heat to the controller chip, or at least form an insulating barrier of sorts.

# ? Nov 7, 2022 03:54

priznat: Jul 7, 2009; Let's get drunk and kiss each other all night.

BIG HEADLINE posted:

There have been arguments that heatsinking the NAND can transfer *more* heat to the controller chip, or at least form an insulating barrier of sorts.

I dunno about that, U.2 enclosures seem to be just fine with thermal pads on the nand and controller all going to the metal case. However these should be in enclosures with constant airflow so should be kept relatively cool.

# ? Nov 7, 2022 03:57

mobby_6kl: Aug 9, 2009; by Fluffdaddy

BIG HEADLINE posted:

https://www.corsair.com/us/en/Categ...p/CX-9029002-WW

And that's not the only one out there.

Lol I should've known!

I thought for sure it was just a wacky aliexpress thing, like a chicken helmet, and nobody would actually put a cooler with heat pipes and a fan on an SSD.

# ? Nov 7, 2022 09:12

Saukkis: May 16, 2003; Unless I'm on the inside curve pointing straight at oncoming traffic the high beams stay on and I laugh at your puny protest flashes.
I am Most Important Man. Most Important Man in the World.

They are the perfect product. Impressive looking upgrades for high performance device, that are useless and harmless, or at least don't do harm in a noticeable way.

# ? Nov 7, 2022 10:46

mobby_6kl: Aug 9, 2009; by Fluffdaddy

Saukkis posted:

They are the perfect product. Impressive looking upgrades for high performance device, that are useless and harmless, or at least don't do harm in a noticeable way.

You are talking about the chicken helmets, right?

# ? Nov 7, 2022 11:45

Klyith: Aug 3, 2007; GBS Pledge Week

BIG HEADLINE posted:

There have been arguments that heatsinking the NAND can transfer *more* heat to the controller chip, or at least form an insulating barrier of sorts.

The flash chips are producing much less heat than the controller, particularly when the workload is 100% read, which is the only time a controller will throttle. And insulating barrier is just what. So that's a dumb argument.

Heatsinking, as in anything with a fin stack, is absolutely superfluous in any DIY PC that has reasonable ambient case temps and some airflow. A heatspreader with some thermal mass is all that is needed, if at all. Some recent drives, like the SK Solidigm P44 Pro, don't throttle even with no heatspreader and no directed airflow. That's the fastest drive on the market right now.

Making the controller throttle in normal operation is unusual. A "hard" workload -- random IO and mixed read/write -- does not do it. The max temperatures are from sustained sequential reads or writes. The main way to observe throttling is to run benchmarks.

# ? Nov 7, 2022 15:42

Adbot: ADBOT LOVES YOU

# ? Apr 20, 2024 00:43

sirbeefalot: Aug 24, 2004; Fast Learner.; Fun Shoe

Is there a particular recommended pick for a 2TB drive for a general use/gaming/CAD laptop? I'm looking at the WD SN770 on Amazon for $180 and wondering if I should save a bit and look at Gen 3 drives instead. This would go into a new ASUS ROG Flow X16. I have used a couple different laptops with Gen 3 drives for the past few years and don't really think they have felt slow to load at any point, but eh.

# ? Nov 9, 2022 22:52

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > SSD Thread: (its good, ignore the anime)

«‹›214 »