ComputerSci

Limited lifetime?

Recommended Posts

I was just reading up getting an SSD and have noticed that they use flash memory, reading up a bit more I found that flash memory does not support as many read/write cycles as a traditional hard disk. I was just wondering how limited the lifetime of an SSD would be?

Share this post


Link to post
Share on other sites

http://blog.mpecsinc.ca/2010/05/intel-ssd-expected-lifetime-and-gross.html

has a slide from IDF 2009 saying 5 years at 20GB/day of random writes (which are the most intensive as far as wear goes).

# X25-M Series in a Client setting

* 5 Years of random writes

* 35 Terabytes of data volume

I think you'll be good as most people write far less than that per day... :P

Share this post


Link to post
Share on other sites

http://blog.mpecsinc.ca/2010/05/intel-ssd-expected-lifetime-and-gross.html

has a slide from IDF 2009 saying 5 years at 20GB/day of random writes (which are the most intensive as far as wear goes).

I think you'll be good as most people write far less than that per day... :P

As long as you limit that statement to the Intel drives.

When you include the Sandforce and Marvell based drives it's another story.

The erase cycle rating of the NAND used on an SSD drive, critical to longevity, is unadvertised on some.

Edited by dabl

Share this post


Link to post
Share on other sites

On that note - the phrasing of Intel makes it seem as if they pretty much guarantee 20GB per day for 5 years which makes it quite likely that the actual lifespan of the flash used in Intel SSDs is much higher.

Share this post


Link to post
Share on other sites
As long as you limit that statement to the Intel drives.

When you include the Sandforce and Marvell based drives it's another story.

True, but at least given the write combining that Sandforce does, those drives are likely to live longer, not shorter, in typical desktop use.

Share this post


Link to post
Share on other sites

True, but at least given the write combining that Sandforce does, those drives are likely to live longer, not shorter, in typical desktop use.

Again it depends on the quality of the NAND involved.

I've seen a claim from an end user that the NAND used in the OCZ Vertex drives for instance is rated for 5000 erase cycles vs the NAND in the Intel drives which is rated for 10000 erase cycles.

The real world user reports I've seen for the OCZ Vertex 2 drives failing after 6 months don't make me want to run out and buy one of those.

Edited by dabl

Share this post


Link to post
Share on other sites

Tony from OCZ says that Vertex 2 uses the same flash as Intel G2.

Well, it can't if the OCZ fails regularly, and Intels don't, don't you think so?

I suppose the controller can make a difference - but a few months, maybe weeks vs. drives lasting well over a year - I don't think the difference on wear would be that large.

Share this post


Link to post
Share on other sites

Tony from OCZ says that Vertex 2 uses the same flash as Intel G2.

Interesting, would be good to confirm that.

For whatever reason there do seem to be significantly more reported failures from users with OCZ Vertex 2 drives vs the Intel drives.

It's possible OCZ has sold significantly more than Intel and that's why I suppose.

Share this post


Link to post
Share on other sites

Interesting, would be good to confirm that.

For whatever reason there do seem to be significantly more reported failures from users with OCZ Vertex 2 drives vs the Intel drives.

It's possible OCZ has sold significantly more than Intel and that's why I suppose.

I don't think Intel has sold particularly few.

And still - how many reports of Intels bricking have you seen?

There was the firmware issue - but that's not "normal operation".

After that there was just one guy on Notebookreview who's drive failed - just died - I guess he was unlucky, that's it.

Share this post


Link to post
Share on other sites

OCZ saying it's the same flash probably means it's true, plus they were showing the SF drive with Intel NAND at the IDF last week.

As to units sold - while there's little data out there, I'd guess Intel leads OCZ by at least 4 to 1. I would expect that margin to be in decline though from mid this year through the release of the G3.

Share this post


Link to post
Share on other sites

OCZ saying it's the same flash probably means it's true, plus they were showing the SF drive with Intel NAND at the IDF last week.

As to units sold - while there's little data out there, I'd guess Intel leads OCZ by at least 4 to 1. I would expect that margin to be in decline though from mid this year through the release of the G3.

If by "they" you mean SandForce, that maybe correct. If by "they" you mean OCZ, that is unverified, speculation on your part (unless you have inside NDA info :P)

via this info Phil posted via Legitreviews the only thing we can see is an OCZ Revo X2 PCI-e SSD with claimed SF controllers, where it is *not* apparent what manufacturers flash is being used :)

http://www.legitreviews.com/article/1410/

Nothing in the image below with the supposed Intel NAND 25nm chips, posted on LR, says *that* 2.5" SSD (assuming standard 2.5" size, could be 3.5" for all we know) has been produced by OCZ...clearly reads SandForce on the green PCB...it's a prototype SSD, with the green wire clearly in evidence on the PCB. Actually the green wire, only looks like a green wire, the image is not that detailed/sharp :P

sandforce_25nm_intel.jpg

Edited by udaman

Share this post


Link to post
Share on other sites

If by "they" you mean SandForce, that maybe correct. If by "they" you mean OCZ, that is unverified, speculation on your part (unless you have inside NDA info :P)

via this info Phil posted via Legitreviews the only thing we can see is an OCZ Revo X2 PCI-e SSD with claimed SF controllers, where it is *not* apparent what manufacturers flash is being used :)

http://www.legitrevi...m/article/1410/

Nothing in the image below with the supposed Intel NAND 25nm chips, posted on LR, says *that* 2.5" SSD (assuming standard 2.5" size, could be 3.5" for all we know) has been produced by OCZ...clearly reads SandForce on the green PCB...it's a prototype SSD, with the green wire clearly in evidence on the PCB. Actually the green wire, only looks like a green wire, the image is not that detailed/sharp :P

Does this count?

Hardware Canucks

The Vertex 2 100GB uses sixteen, Intel branded, 29F64G08CAMDB chips. These chips are very similar to those found in the Corsair and are the exact same ones we found in the mid tier G. Skill Falcon 2. To be honest this is not all that surprising as SandForce makes a big deal out of the fact that you can use mid grade NAND on these drives and still get best in class performance.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now