Has my SSD died?

Just tried to boot up my laptop and after the initial boot screens it tells me to insert proper boot media and restart. In the bios it should read as Supertalent SSD but it no longer does. Also when I put vista install disk in to try and do a fresh install it could not pick up a drive to install on.

So does this mean my SSD is dead?

 

sounds like it. maaybe open up the laptop and check if it's plugged in correctly?

maybe mount it in some pc to test if it's visible there? plug another hdd/ssd into the laptop to see if it works there.

but it could have died.

 

probably. welcome to the world of over-hyped zero- and first- generation 'technology'.

 

if all would fail, i'd support you, newsposter. but the chance for failure is in all components since the first computer in existance, so no need for some random-sarcastic comment.

to the op, hope you a) have a backup or b) get your data off the drive

the only other thing that comes to mind would be switching from sata to ide-emulation mode in the bios if possible? maybe then the drive would respond again? (unlikely, but still worth some little try?)

 

Thanks dave I will try to change to IDE in the bios. I didn't have a backup for anything but there was nothing too important on there. Just will lose my pictures. I can't afford to do a data removal and buy a replacement lol. It lasted 18 months or so and it was quite an early SSD so I can't completely complain. I think vista killed it for me in the end, it is such a rubbish OS. I did take it out of the laptop then put it back in with no result also.

Is the 5400rpm Hitachi 5k500.B 500GB drive any good to replace it? Don't think I want to get another SSD for a few years yet to ensure the technology becomes completely mainstream.

edit: the bios doesn't pick it up at all I think it refers to the ram installed now as it says 4.2gb capacity SSD..

 

Your SSD should still be covered by the SuperTalent warranty:

http://supertalent.com/support/Warranty Policy.pdf

It says 2 years for MLC and 3 years for SLC. You shouldn't have trouble getting a replacement from them.

 

Hmm I wonder if I could contact them direct as I bought it on newegg when I was in the USA and now I am back in the UK!

 

So what are you waiting for?

 

Lol email sent, thank god the newegg email with the invoice number from a year back is still there!!

 

theres one for longevity.

How much did it cost back then?

also can you tell us how you use it? do you treat it with special care? (I'm guessing you treat it 'normal' with no special considerations right?). how much memory were you writing per day roughly? do you hibernate your comp often etc? what do you think caused the failure?

 

Back in June last year this MLC 120GB SSD cost $620!! I never hibernate my machine and always disable unnessary services and startup programs including on vista these horrible superfetch, readyboost and indexing services, also keep it pretty empty at all times and ran regular ccleaner to ensure everything was "tidy", never ran defrag and disabled it after installing vista recently. So i guess I did look after it pretty well and had no trouble with it on XP. I gave vista a 2 year break and still I hate it lol. Lets hope supertalent let me RMA it and give me a replacement heck I deserve it after putting such an investment into their possibly half baked product.

 

Have you tried flashing the firmware? I think there is a updated firmware for the 120 GB model. http://www.supertalent.com/support/driver_download.php#

 

I did that when I installed vista on it recently. Flashing any sort of firmware on it now is impossible though as it is no longer recognised due it being dead probably lol. Thanks for the idea though

 

you definately should. at the very least get their newest edition for free and send them their failed device so they can find out what went wrong with their product.

what is your guess on how much memory writing you did per day, roughly?

btw did you try booting with a ubuntu usb drive or add a new HDD in there to load windows to make everything but the SSD is workin?

 

I ain't got another HDD to hand but my bootable usb stick which I use to do vbios flashes works perfectly. Anyways just got email back from supertalent with RMA number. Btw what does it stand for lol! Off it goes from UK to California guess no laptop for another month...
I can't says how much memory I wrote. I just used it to surf, use word 2000, excel etc on and occasionally game.

 

I don't understand why people seem to think that it is the writes to the drive that caused this? Excessive writes don't cause an SSD to do this (unless the drive controller died or overheated or the like), the write limit on MLC just means that the cells can't be written to, they can still be read and the drive would still work. This just sounds like a hardware failure, not an issue of excesssive writes.

 

Hello there. Floating gate transistor based electrically erasable read only memory is currently used in most 'solid state memory' for non-volatile bit storage, and is responsible for 'memory' like features in these electrical devices.

Reading cycles on the floating gate have fairly low effect on device reliability as voltage measurements on the floating gate does not tend to degrade the material. Programming/erasing cycles on floating gate transistors have limited reliability because of a variety of physical phenomena occuring at the nanoscopic material level induced as a result of the charging and discharging of the floating gate. Theses phenomenon are responsible for the theoretical figures of finite endurance of solid state memory from cycled programming/erasing being quoted by manufacturers of solid state memory (and as it also relates to the nature of memory cell configurations).

Research has shown that high cycles of writing/erasing (or program/erase, PE) increases the rate of charge leakage from the floating gate. What this means is data will eventually be lost over time. Some research also show certain floating gates have a saturation of charge leakage with increasing PE cycles, but that nontheless charge leakage increases with increasing PE cycles.

So in other words, over time the failure mechanism on floating gate transistors is most likely this: You cannot read useful data over time because even data written before will be lost if the floating gate (the actual single device responsible for bit assignment) fails to store charge (the bit value). Effectively, you will read all 111111111, or 00000000 depending on convention, taken ad infinitum in time. From a software programming standpoint, this may be disasterous very quickly because bits representing coding of a program will be gone over time (and there will be statistical randomness as to how important these lost bits are to the device operation). Therefore, one should not expect your windows or linux operating system to still turn on but that you simply cant save your office document. What is likely is your computer may respond as if you just put a plank piece of single crystal silicon in there, that is presumably nothing except for the BIOS will load.

Popular news articles like to quote manufacturers stating that the 'controller' will fail before floating gates fail. This may or may not be true depending on how good the semiconductor microchip is fabricated and designed, although it is entirely possible that a well made controller can outlast even the best current floating gate transistors engineers can fabricate. Think of it this way: semiconductor processors can perform for many years running extremely high frequency electrical signal cycles. Even high power semiconductor (non-floating gate) devices can perform for many years. Floating gate transistors fabricated with current technology (and for the forseeable future) however have much much less life than conventional semiconductor devices. Just take a look at your microprocessor. They are at gigahertz clock speeds and can survive many years. The controller in solid state memory devices are made using these conventional technologies and engineering design protocols. Floating gate transistors are also a special class of semiconductor devices, but the nature of 'floating gate' for charge storage specifically for non-volatile memory applications, and together with the neccessity to rely on quantum tunnelling as it requires lower voltages suitable for commercial applications, is a huge problem for engineers! Degradation of material properties at the sub-nanoscopic level is simply unavoidable. And because its unavoidable, reliability problems of floating gates are also unavoidable.

 

what does what stand for? rma? = return merchandise authorization

are they giving you the same old one or a newer generation one, presumably better. i doubt they will keep you in the loop about what went wrong with the SSD right?

 

1.5years is not bad for "such an old ssd". in those 1.5 years, ssd improved a lot, so getting a new one should result in a longer lasting disk. hope you get a good replacement trough warranty.

and vista is a great os and does NOT kill a hdd or an ssd. they die by themselves

and for the next time, get a backup solution.. when you got the ssd, you had a spare hdd left, not? that could have been your backup disk

technology is quite mainstream right now. they matured much. yours was, by a big chance, a jmicron one, which are known to be quite bad and possibly had bad wear leveling. and the chip most likely doesn't know what to do if he overwrote the data too much and the flash gets readonly. i know the intel ssd f.e. should not fail once it can't write anymore. but we first have to wait to get one filled 100000 times

anyways, check out now if you get a fresh one trough warranty. that's the first job. hope you get an indilinx version from supertalent back. they should work quite well. haven't had one myself yet.

 

what if the controller is too dump to understand what to do when it gets a "can't write anymore error" on a cell? maybe at that point, the controller just died.

but anything over a year old can die just out of fun, seen enough devices just die for no reason after 1, 2 years. mostly cheap stuff. and jmicron was chosen because of one reason: it's a cheap controller.

 

I think it is Jmicron because that is what the bios page flashes up before it boots into windows, still does even now but I guess the controller has died now. My question is why would they use a cheap controller in a product they charged $620 for? That certainly is not cheap

 

to make money?

at first, there where mostly just the mtrons, the samsungs, the memorights. all of them where much more expensive than 620$ for 120gb. the mtrons where around 800$ for 64gb back then (or more? .. ).

the jmicrons where the only cheap alternative. cheap by back-thens days. it was relatively cheap. nowadays you could get a very good ssd for that price, but first, get the replacement. then, maybe, sell it and get something new? for the 620$ you can get a nice intel 160gb

 

Indeed. Hopefully they give me a new SSD worth $620 that would be leaps and bounds ahead of my old one

 

that would rock wish you luck, you deserve it.