SSD's - a risk?

[farinha]

I have a OCZ Vertex 2 for 3 months and can't complain either. Still working flawless.

 

[C0RR0SIVE]

Ok, then lets wipe the slate clean, you can't say they are more reliable for another 10 years, considering there are plenty of HDD out there that have last that long, and beyond. I still have a 20mb HDD laying in the closet that works from like 25 years ago.

 

[Peter.Cort]

There is a write limit to flash based SSD's. So there's the risk that eventually it will die, but that's no different than any other piece of hardware, at some point it will die, it's just a matter of when.

 

[roweder]

How do you tell if a SSD is flash based or not?

 

[Peter.Cort]

Newegg.com - OCZ Vertex 2 OCZSSD2-2VTXE120G 2.5" 120GB SATA II MLC Internal Solid State Drive (SSD)
^ flash based,

Newegg.com - OCZ Z-Drive R2 M84 OCZSSDPX-ZD2M84512G PCI-E 512GB PCI Express MLC Internal Solid State Drive (SSD)
^ DRAM based

it's like a flash drive versus ram pretty much. also DRAM ssd's are wicked expensive.

 

[roweder]

I don't see on those pages anything that says whether the drives are DRAM based or flash based, is it safe to assume that all super expensive PCI SSD's are DRAM based and all SSD's that connect using SATA are flash based?

Do you think that a typical user could reach the write limit in under 10 years?

Is there a read limit to flash memory?

 

[Slaymate]

flash memory has a write limit, it won't have a read limit. The life of the SSD will be based off of the write limit. Each data block can only be written to a limited number of times (this is a large number not yet determined). To prolong the life of the drive, the controller in newer SandForce drives, won't allow the system to re-write a data block until all the other empty data blocks have been written to 1st. So if you only use 50% of the storage space you won't be writing to the same data blocks over and over again like you do with a std. HDD.

 

[Lexluethar]

I agree, there hasn't been enough data out there to determine which lasts longer. As Slaymate says any solid state device only has a limited amount of writes to it before it won't hold a charge. That is the what really determines the life cycle of a SSD. This write limit happens on platter drives too because they too lose charge, but the more likely failure is mechanical since it has moving parts (read head crash, drive not spinning, etc).

With any new device come speculation - and that is a good thing IMO, just take it with a grain of salt.

 

[roweder]

I agree, there hasn't been enough data out there to determine which lasts longer. As Slaymate says any solid state device only has a limited amount of writes to it before it won't hold a charge. That is the what really determines the life cycle of a SSD. This write limit happens on platter drives too because they too lose charge, but the more likely failure is mechanical since it has moving parts (read head crash, drive not spinning, etc).

With any new device come speculation - and that is a good thing IMO, just take it with a grain of salt.

This makes me wonder if both have the same limit, but we see it in SSD's because, with no moving parts, they last long enough to reach this limit.

 

[Peter.Cort]

The write limit on a platter based drive is different than a flash based drive, they won't lose a charge because the data isn't being written by an electrical charge. They will reach a write limit based on the mechanical failure over time. Obviously the more you use a drive the shorter it's life span will be, but there's no definitive limit. I would venture a guess that since each drive is a bit different they will have a different limit, and also depending on the intensity that you're using the drive at, that will affect the write limit. For instance I would say that if you're constantly writing to the drive over and over with complex data it would have a lower write limit than if you wrote a million 0's to the drive once a week until it failed.