Show us your SSD performance

[Dave76]

The only problem with that is they are using the manufacturers rating. Intel rates their lifetime at 5GB of writes per day for 5 years. Those tests they are running has proven that that estimate is grossly understated. If my calculations are correct, Intel is rating theirs at around 9TB. They are over 130TB. and still going.

I believe it is 20GB per day for 5 years = 36.5TB

Link to Datasheet

Minimum useful life 5 years
The drive will have a minimum of five years of useful life under typical client workloads with up to 20 GB of host writes per day.

My X25-V 40GB has 1TB after one year, by Intel's standard I'm good for 36.5 years

On the XS endurance thread the Kingston(Intel) 40GB drive is at 110TB, MWI 40%, and still going strong.
Guess mine will be good for 150+ years:shock:

Think I'm safe leaving my page file on the SSD, have since I got it. At that time many 'experts' were saying just leave it on there. Now I see why.
I have read so many posts (most of them not here) with ways of moving just about every possible write off of the SSD.
Thought it was a bit extreme at the time, but you never know until someone shows you

I think they will get tired of running the test before the drives wear out, my bet is 1+PB of writes can be reached.
This is what the other two tests that were linked made it to, after a year of heavy writes.

Interesting discussion on the OCZ v2 SF controller, they moved it to another thread as it was getting a bit long and... let's say, a debate.

 

[Hopalong X]

Interesting discussion on the OCZ v2 SF controller, they moved it to another thread as it was getting a bit long and... let's say, a debate.

Yeah I read one fellows posts the last few days. He insinuates he knows how to fix the problem with the Vertex 2 since the problem started weeks ago but has not posted until a few days ago.

He still hasn't posted the solution. Just making rambling useless posts.

Appears to be a U.S. Congressman in the making.
Might be running for Wieners empty seat.

 

[whs]

The one thing that is safe to assume from all this data is that the SSDs are not as fragile as some people suggest. I think most of us will long have moved to a later generation of storage devices before our cirrent SSDs wear out. Therefore it is not necessary to stand on your heads to reduce writes to the SSDs. Especially the page file, which really does not have a lot of writes when you have 3 or 4GBs of RAM, needs not to be moved.

 

[Dave76]

Interesting discussion on the OCZ v2 SF controller, they moved it to another thread as it was getting a bit long and... let's say, a debate.

Yeah I read one fellows posts the last few days. He insinuates he knows how to fix the problem with the Vertex 2 since the problem started weeks ago but has not posted until a few days ago.

He still hasn't posted the solution. Just making rambling useless posts.

Appears to be a U.S. Congressman in the making.
Might be running for Wieners empty seat.

:roflmao:

Just call him a Wiener :roflmao:

I was thinking the same thing, well almost the same thing.
Actually have seen his posts on the other forum, quite knowledgeable, don't know why he was going on about his qualifications, when he could have just gave a suggestion on how to do a different SE on the drive.
The usual cause is not reading much of the threads posts, just posting...

 

[Hopalong X]

SE won't help. Throttling is in the firmware and there is no fix except from OCZ.
A way to turn throttling off or ability to change the paramaters.

----------------------------------------------------------------------------------------------------------------

A Congressman knows everything about everything but.....needs picture instructions on the back of the stall door to use the paper distribution appliance.
Instructions from how to remove and where to use.

 

[essenbe]

SE won't help. Throttling is in the firmware and there is no fix except from OCZ.
A way to turn throttling off or ability to change the paramaters.

----------------------------------------------------------------------------------------------------------------

A Congressman knows everything about everything but.....needs picture instructions on the back of the stall door to use the paper distribution appliance.
Instructions from how to remove and where to use.

Even then, he will mess it up.

 

[Hopalong X]

Even then, he will mess it up.

:roflmao: :roflmao:

I'm glad I wasn't drinking anything when I read that.

The monitor would have received a shower!

The image that came to mind!

 

[essenbe]

Even then, he will mess it up.

:roflmao: :roflmao:

I'm glad I wasn't drinking anything when I read that.

The monitor would have received a shower!

That is funny. But the sad part is it's the truth.

 

[Hopalong X]

We keep electing them. "We" as in all the public. New ones are worse than the old batch.

Oh well enough on politics or I'll end up in trouble.

 

[essenbe]

We keep electing them. "We" as in all the public. New ones are worse than the old batch.

Oh well enough on politics or I'll end up in trouble.

You won't get in trouble with me. I think they are all criminals. Some just worse than others.

 

[essenbe]

Sounds like they are slowly getting there. SSDs are much more durable than most people give them credit for.

 

[Hopalong X]

SSD Write Endurance 25nm Vs 34nm - Page 24

Post #584 and #585

Kingston SSDNow 40GB- Intel 25 rebrand
116.86 TB writes
MWI- 36%
RaS-6

Intel 320 40GB
155 TB writes.
MWI- 20%
RaS-13

 

[essenbe]

I say they still won't reach 1PB. But, I'm surprised they have written as much as they have.

 

[Hopalong X]

Big jump from 155GB to 1000GB (1PB)

Whoever thought 1PB of writes were possible was let's say enthusiastic at least.

 

[essenbe]

Big jump from 155GB to 1000GB (1PB)

Whoever thought 1PB of writes were possible was let's say enthusiastic at least.

A couple of other sites tested the drives and wrote 1PB. However they gave no details on what kind of data was written and how the test was done.

 

[Hopalong X]

Hammered state as they are running in this test? Doubtful.

It would be like testing my Honda van engine at 7000RPM red line until it dies vs. 1900RPM 55MPH cruising speed or it might even be like testing at 1200-1500RPM.

Of course they are doing sheduled "oil changes" etc. with TRIM.

That was the purpose of the xtreme forums tests to slam them hard and see how long they would last.
So far so good!

 

[Wishmaster]

Hammered state as they are running in this test? Doubtful.

It would be like testing my Honda van engine at 7000RPM red line until it dies vs. 1900RPM 55MPH cruising speed or it might even be like testing at 1200-1500RPM.

Of course they are doing sheduled "oil changes" etc. with TRIM.

That was the purpose of the xtreme forums tests to slam them hard and see how long they would last.
So far so good!

This all just re-enforces my belief the "reducing writes" to the SSD is blown way out of Proportion.
Which is why I always say, just leave everything where it is, except User DATA (for space) and dont worry. Even Pre-Fetch is fine if you want it.

If they can write this much DATA to them, I honestly do not see anyone wearing out thier SSD in 5 years time or more (Defective Hardware excluded)


Its quite intersting to say the least.

 

[essenbe]

I think a more intresting question is how much could they write to a 120GB SSD. These are 40GB. Many more nand cells on a 120.

 

[essenbe]

Hammered state as they are running in this test? Doubtful.

It would be like testing my Honda van engine at 7000RPM red line until it dies vs. 1900RPM 55MPH cruising speed or it might even be like testing at 1200-1500RPM.

Of course they are doing sheduled "oil changes" etc. with TRIM.

That was the purpose of the xtreme forums tests to slam them hard and see how long they would last.
So far so good!

This all just re-enforces my belief the "reducing writes" to the SSD is blown way out of Proportion.
Which is why I always say, just leave everything where it is, except User DATA (for space) and dont worry. Even Pre-Fetch is fine if you want it.

If they can write this much DATA to them, I honestly do not see anyone wearing out thier SSD in 5 years time or more (Defective Hardware excluded)


Its quite intersting to say the least.

I agree with that. I think these tests show it is hard to kill one, even if you try.

 

[Hopalong X]

I think a more intresting question is how much could they write to a 120GB SSD. These are 40GB. Many more nand cells on a 120.

My thoughts also.

In theory you could loose an entire NAND chip or two and still have 90GB? left to keep using on a 120GB.
As long as the controller stays working- only one of them.