Show us your SSD performance

[essenbe]

Hopalong, I don't think you need to worry about your SSD. The xtreme systems test is up to 80TB written to it and it's still going. Kind of like a timex watch.

 

[Hopalong X]

I think you are right. Might not be the fastest but should out last my 10 year old greatnephew.

Year is 2060 and still running. :shock:

 

[essenbe]

I think you are right. Might not be the fastest but should out last my 10 year old greatnephew.

Year is 2060 and still running. :shock:

Kind of like that old (ancient to some of you) Timex watch commercial. Takes a likin' but keeps on tickin'

 

[ElisD]

Hi, First post, I am using a Crucial M4 64 GB



Are theese results good? and is there a better driver than msahci?

 

[Wishmaster]

Welcome to the forums

Nice scores too.

 

[essenbe]

Welcome to the forums. Great scores. As far as the driver, With my intel and vertex2 I found msahci performed better than the AMD driver. I'm sure the Crucial forums will have a lot of info on that.

 

[A Guy]

Welcome to Seven Forums ElisD. Great numbers

A Guy

 

[Wishmaster]

Welcome to the forums. Great scores. As far as the driver, With my intel and vertex2 I found msahci performed better than the AMD driver. I'm sure the Crucial forums will have a lot of info on that.

You're not the only one.
I have a IntelP45 with a ICH10 (not the ICH10R)

Mine seems to do just as well, maybe slightly better just using the generic MSAHCI driver over the Intel iaStore, for whatever reason.

Although admittedly, the differnce is very little.


Some do however see nice gains from the Intel driver. I suppose it really depends on the setup.


The standard MSAHCI driver actually works really well.

 

[essenbe]

Can this be true?

This site experimented to see how much data could be written to an X25-V before the NAND expired. They wrote to the device 24/7. (Seems they paused once every 12 hour to run TRIM from the toolbox).

Translation makes it hard to understand exactly how they ran the experiment, but they state:
...
In total there were able to achieve 0.73PB in 6,185 hours! That is a phenomenal amount of writes, which appears to be way over spec of 20GB of host writes per day for a min of 5 years.

Here is another one 0.86PB in 7,224 hours!

Does that come down to the work load or are Intel specs extremely conservative?

From the Intel Forums.

 

[Dave76]

They went to a whopping 789.9TB host writes before they stopped.
Can't tell for sure from the translation but, seems like it was still going :shock:

Post #1532

As mentioned already, some SSD endurance tests have been seen and one is currently going.
SSD Write Endurance 25nm Vs 34nm

They reached the 20GB of host writes per day for 5 years, 36.5TB mark, and are still going strong.

Some interesting links in this thread.

SSD NAND Endurance​

Can this be true?​

This site experimented to see how much data could be written to an X25-V before the NAND expired. They wrote to the device 24/7. (Seems they paused once every 12 hour to run TRIM from the toolbox).​

Translation makes it hard to understand exactly how they ran the experiment, but they state:​

A class (Sequential?)

• 10~100KB 80% 10 ~ 100KB 80% (System thumbnails and the like)
• 100~500KB 10%100 ~ 500KB accounted for 10% (JPG images and the like)
• 1~5MB 5% 5% 1 ~ 5MB (big picture, MP3 and the like)
• 5~45MB 5% 5% 5 ~ 45MB (video clips and the like)

Where B is random write categories: (Random?)

• 1~100 100% 1 to 100 bytes of 100% (System log and the like)

In total there were able to achieve 0.73PB in 6,185 hours! That is a phenomenal amount of writes, which appears to be way over spec of 20GB of host writes per day for a min of 5 years.

Here is another one 0.86PB in 7,224 hours!​

Does that come down to the work load or are Intel specs extremely conservative?​

Looks like SSD endurance is much better than previously indicated.
The 20GB per day for 5 years = 36.5TB
The XS thread have reached that with a 80% Media Wearout Indicator, 80% life remaining.

One of the other threads linked above have reached 187.2TB with a 1% Media Wearout Indicator, 1% life remaining.

A whopping 789.9TB host writes before they stopped.

That's 432.8GB per day for 5 years, or 20GB per day for 108 years.

Guess we don't have to be so concerned about the amount of writes on our SSDs.

 

[Hopalong X]

Expanding on Dave76's previous post.

SSD Write Endurance 25nm Vs 34nm - Page 15 - XtremeSystems Forums

In progress. Running approximately 3 weeks. 5/16/2011

A quick summary of testing.

Intel 320 -40GB
81TB Host Writes.
58MWI
Reallocated Sectors- 5

Kingston SSD Now 40GB (Intel X25V rebranded)
61.3TB Host writes
MWI 66
Reallocated Sectors- 5

The Vertex2- 40GB also being tested has Vertex "DuraClass" firmware throttling.
This "saves" the drive from wearing out to quickly as to extend life years of service.
It activated after approximately one week of usage.
Writes dropped from 175MB/s to 7MB/s.
It essentialy quit working at 28TB of Host Writes.
The operator and others on the forum have been trying to resurrect the Vertex for the past two weeks to no avail.

Good read. 15 pages but mainly focus on Anvil, One Hertz and Ao1 postings.

 

[Dave76]

Have been following that for a couple weeks.

See my last post, others ran similar tests, one got to 789.9TB host writes, and it appears they just stopped - no SSD failure.

 

[Hopalong X]

Edited my above post.

I was expanding on your posting.
A summary of their testing.


I can't imagine writing 80TB in my lifetime let alone 789TB. :shock:

 

[Dave76]

That is an amazing amount of writes, guess the manufacturers are being very conservative.

Looks like we can all relax about controlling the number of writes.

 

[whs]

That is an amazing amount of writes, guess the manufacturers are being very conservative.

Looks like we can all relax about controlling the number of writes.

You are absoplutely right. I never understood why people stand on their heads to control the SSD activity - and for a device that they will probably replace in the next 3 years. Must be the excitement over such an amazing piece of hardware.

 

[Miranova23]

Hey, I've been wondering now just how good this SSD actually is. Glad I found the tests to run here

OCZ Vertex2 3.5" 120GB SSD

 

[A Guy]

Only a lousy 789TB

A Guy

 

[essenbe]

Hey, I've been wondering now just how good this SSD actually is. Glad I found the tests to run here

OCZ Vertex2 3.5" 120GB SSD

You can also use As SSD, and ATTO. But don't run too many tests as it will slow the SSD down. OCZ also recommends to let the computer idle at the login screen for 2 hours a week for 'garbage collection'. It works too.

 

[fseal]

I gotta say that the vertex2 issue is a little disturbing though

It's "good enough" but no where near as good as the others.

THough one problem is that these are /extremely/ small samples. You would need to run this test 50 times on 50 copies of the same device to get a more meaningful picture. It may be that 50% of the devices last to 900TB and 10% die at 10TB...

 

[whs]

Hey, I've been wondering now just how good this SSD actually is. Glad I found the tests to run here

OCZ Vertex2 3.5" 120GB SSD

The transfer rate should be above 200MB/sec and the access time should be 0.1ms. But I had the same problem with HD Tune - the Vertex2 results were all wrong whilst the Intel results were OK. Atto is a better measuring tool.