Testing

[AirPower4ever]

Not sure if this belongs here, but it does deal with drives. My question is I have an SSD and I want to test it using ATTO. Final got an SSD and great so far.

Looking at ATTO shows that there are different settings so I started reading up. Seems like there is no standard settings, read that people use the "Total Length" size to make their numbers look good. Some use the "Total Length" set to 8MB, some use 32MB, others use 512KB, etc. Saw others use the "Transfer Size" .5 to 8192 while others used .8 to 1024, etc.

Is there a standard setting that I should set ATTO at to test the SSD? What settings should I set the following to do the test

Transfer Size: to and from
Total Length
Queue Depth: leave at 4?

When you reply, the settings anyone provides, can those be used for platter hard drives? Are platter drive tested with different settings or the same will be okay to get data?

Thanks for your help

Note: Sorry, I forgot to add that ATTO version is 2.46. I noticed that the highest setting for Total Length is 2GB, not sure if this is needed for an answer.

 

[ghost6303]

search for "benchmarking SSDs" and the difference between compresable and incompressable data, and the difference between sandforce controller-based drives, and intel controller-based drives.

long story short is that "real world usage" (whatever that means to you) is normally vastly different than the single number you see in benchmarking, and is therefore difficult to quantify. that doesnt mean that benchmarks are necessarily wrong or inaccurate, but they only test for a single set of circumstances while in real life there is a mixture of many. you should also know that if you have a sandforce based drive (like OCZ vertex), running repeated benchmarks will cause the drive to artificially throttle its speeds, which even further separates your benchmark number from real world usage.

what SSD do you actually have?

 

[AirPower4ever]

My brother gave me two Corsair Force 3 60GB SSD's. I was researching which SSD's I wanted for my 2 HTPC builds (one is now running), but he bought several and then just gave me 2.

I checked on these and they are not the bad serial numbers that I have read on.

The one I have running seems really fast. These forums are right, once you get a taste of these, you are not going back. I am already planning on getting more for my other systems which are for gaming and work.

I did a test using ATTO with .5 to 1024 at 32MB and she is hitting 535on reads and around 467 on the writes. The only thing I am concerned with is that these are Asynchronous which I read kyle's review between the difference between Async and Sync. But how can I complain, these were free.