Crucial P1 minicard flash drive? Not if you grabbed Intel's 660p Micron's consumer SSD brand Crucial has pushed out an NVMe minicard flash drive that looks almost identical to Intel's 600p. The M.2 2280 format Crucial P1 arrived just months after Intel's 660p and uses the same 64-layer QLC (4bits/cell) 3D NAND. Intel and Micron share the output from their IMTF flash foundry joint-venture. …
Out of, dunno - maybe 20+ SSDs that have been here so far, exactly two have failed. And in all time (let's just say decades going back to mfm) - are the only times data has actually been truly and instantly lost.
Both were Crucial M.2 sata, and they seemed to have failed due to parasitic body scr latchup. Got very hot, drug the power supply down to zip (intel NUCs in both cases) and were not recoverable even from an external adapter card - even in the one time in ten they didn't latch up again on power up.
Every other SSD I've bought - Samsung, Sandisk, Intel - quite a number of them, most in much more demanding applications - is still working fine, with many times the power cycles and runtime hours.
!00% failure rate of one brand says it all for me. FWIW, the same machines that had those drives fail in them had Crucial brand LPDDR3 too - and it's fine.
Just one, well, two, data points. Enough for me, though.
I bought a bucket of the cheapest Crucial junk SSDs I could find, lobbed them into any machine in work that couldn't take our >4Gb RAM upgrades (which tells you the age of those machines! They run 64-bit WIndows but the motherboard can't take more than 4Gb RAM) - so half the machines are 4Gb with an SSD, the other half are 8Gb with a normal hard drive.
Bear in mind that I *never changed a single option* - none of this caching rubbish, no "tool" running to optimise the SSD, no overprovisioning, no disabling of swap, etc. - literally a byte-for-byte image of whatever was on the same computer before the upgrade...
1) I've not had to replace one in over 4 years.
2) If I did, they are the cheapest things to replace, and literally replaceable because nothing is stored on the HD, just the OS and roaming profiles.
3) They would be much swapping harder than the 8Gb machines.
4) They OUTPERFORM the 8Gb machines, by a large margin. People use them in preference.
5) When I *do* runs the tools, there are zero failures and the estimated life is still 5 years +
6) These machines are hit hard every day, in use all through the working day, way into the evening, and sometime 24 hours a day in some locations. They get dozens of users a day sucking down their entire profile and then pushing back to the server, and doing all kinds in between and "Switch User"ing between half-a-dozen users all the time rather than logging off.
I honestly can't fault them... I have a Samsung in my personal stuff but they were a test to see if they were viable and whether I'd have to replace them every year, and they are still flying. If I had to replace them every year, I really wouldn't care at this point.
P.S. You should never lose data. Literally never. If you can afford one drive, you can afford two half the size and something to RAID between them, even if it's only a pathetic mirroring. And you shouldn't be storing anything critical on any machine that can't do that (we call those clients, they shouldn't be storing files on them and you should be able to code up a bare-metal machine to a working client with all your software and domain in minutes).
Now, if you'd said Seagate and hard drives - I'd be right with you. I burned through EVERY SEAGATE DRIVE in the workplace in that same time. Literally everyone failed, and every RAID resync with more Seagates inside them was a cross-your-fingers-and-check-your-backups moment. Every single drive that failed was Seagate (whether SAS or SATA, client or server or storage). Every Seagate drive has failed.
But the cheapest, junkiest, more useless, sacrificial Crucial SSDs... they are so impressive, I've worked out what I'm upgrading next rather than RAM.
Oh christ yeah, how bad are Seagate drives, every single one of them died on me, in similar set-ups and use cases as Western Digital drives, which not one has failed (few have wobbled and been pulled before given the chance, but not a fail, unlike the Seagates that would just go down hard).
Hmm, I'll steer clear of the Crucial M2 ssds for a while then, thanks.
We've been mainly using the WD Green nvme drives which work. I couldn't tell you any more than that because they've only been in service for about six months, and as for speed, once they're faster than a harddrive that's 'good enough' for our purposes.
The main problem I have is finding suitably small ones. A basic Linux Mint install only needs about 16GB, but the cheapest SSDs are around 120GB now, which isn't exactly a problem I suppose, it just offends my sense of efficiency.
Same here.
That is to say I have forsworn getting another Crucial NVMe thingy.
Not because the one I got was crap (although eventually that was proven to be the case) but because I was too late (thinking the problem may be with my machinery - which is quite picky when it comes to NVMe - and then trying different formulations) to get help from Amazon, and when (per Amazon suggestion) I emailed Crucial to recover the ridiculous amount I paid they did not even reply.*
Once bitten .....
FWIW one of the things I tried was a PCIe adapter which takes NVMe SSD's. The Crucial one did not register. Threw it in the bin. Bought another - non-Crucial - and it is fine.
*P.S. I do not live in the USA so getting redress is sometimes too much of a PITA.
It's good to see someone else agree with me about Seagate drives. With one exception, every other brand I had, had still been SMART = 100% when I threw it away for being too small (I still have a working 80GB HDD).
So that is ONE HDD failure since Win95 ruled (Fujitsu 3TB @ 8 years old)
As for Crucial, not had any experience with them, but I HAVE had two other SSD brands dies on me; both during normal low load instances, checking email, reading web etc; both 100% total loss.
One was 2 years old and one was 2 weeks old.
I dont trust any valuable data to them now; OS and Program files go on SSD, all data and documents go on spinning rust.
I guess it must be quality control or "luck of the draw".
FWIW, a little reading comprehension here - I was using m.2 sata drives, not something you could even put in an old machine. Assuming the guts are identical is well, an assumption, Lee, and I've noticed you're usually a lot smarter than that.
I've had great luck with seagate drives - and GM cars, but seem to have avoided the models everyone knows (now) were utter crap. I have some pretty old stuff that's taken quite the beating and still works fine - I only took the 2 gb seagates offline because, why would you keep something that obsolete...
I've had IBM deathstars go bad, but our software dev outfit was cross backing them up on each other's machines. So no loss, other than $$ and time.
Hint -you can't raid in a single slot NUC. You can back up, of course, and I did - but not every minute.
And FWIW, neither is proof against a fat finger delete - even a backup if you don't notice in time and avoid backing up the mistake.
Lucky it wasn't real important data I lost to Crucial..else I'd have used better stuff (not some rebranded OCZ)..else I'd have been more careful.
And you know, there's a limit - hindsight by a third party always says you should have spent more than you made protecting what you made???? Things like raid and backups cost time and money too, and sometimes aren't worth it. It's almost always better to avoid the problem in the first place as a one time cost. A rate of expense - well, a drip can empty the ocean.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
