As ssd Benchmark giving radially different results. Why?

voodooking

Distinguished
Mar 2, 2012
298
0
18,810
Every time I run AS SSD I get a performance swing of 100 MB/s on both the read and the write sections for the sequential test. I am assuming that it is not the tests fault. I am thinking that it could be my programs running in the background. Is this the problem? I have Norton and several others. Do I need to shut all of these programs down when I am testing?

If it is not the background programs then what is it that is slowing my SSD down. I am running the Vertex 3 at about 400 or so read and 100 or so write. I bought this thing because Tom's said that It was going to be running over 500 for both read and write. What gives? I did everything in the speed up your ssd post. What do I do?

Thanks,

Jim

BTW. Should I post in my specs that I run Norton? Is it that big of a deal?
 
Solution
AS-SSD writes 5GB of incompressible test data per run.
CrystalDiskMark writes 20GB of incompressible test data per run.

How many time have your run AS-SSD? If you are running these benchmarks more than twice, then yes, your results will get lower and lower.
You're writing data in minutes to your drive that you probably don't write in weeks.

Also, advertised AS-SSD specs for your drive are "up to" 500MB/s Read and 155MB/s Write.

If you are looking for the advertised 550/500 Read/Write speeds then you have to benchmark with ATTO software, which uses compressible test data.
AS-SSD writes 5GB of incompressible test data per run.
CrystalDiskMark writes 20GB of incompressible test data per run.

How many time have your run AS-SSD? If you are running these benchmarks more than twice, then yes, your results will get lower and lower.
You're writing data in minutes to your drive that you probably don't write in weeks.

Also, advertised AS-SSD specs for your drive are "up to" 500MB/s Read and 155MB/s Write.

If you are looking for the advertised 550/500 Read/Write speeds then you have to benchmark with ATTO software, which uses compressible test data.
 
Solution

voodooking

Distinguished
Mar 2, 2012
298
0
18,810
a BIG TEXAS Thanks! to the both of you... I love asking questions here. There are always the best responces.

As for number of times. There have been several maybe 6 or 7. But never back to back.

So I was wrong? about the background programs? i don't need to turn them off?

Rock on. Thanks for your help. I have no idea who the best answer is..

Jim
 
Background programs use CPU and RAM resources which should not drastically affect benchmark results.

If you're running a Full (or Quick) virus scan then let it finish before running a benchmark.

When you finish your benchmarking Log Off (not Shut Down) overnight (3 or 4 nights) to allow GC (idle Garbage Collection) to restore drive performance.
 

Meganano

Distinguished
Feb 22, 2011
121
0
18,680

Is it really necessary to log off for garbage collection to run, or can I just turn off sleep and let it idle? I'm not worried about security. Should I also turn off the screen saver?
 


With my OCZ Vertex 2 SSDs I have to be at the login screen or I have to be in Safe Mode in order for GC to run.
In any other mode there are too many Windows Services loaded for GC to run.

 

Meganano

Distinguished
Feb 22, 2011
121
0
18,680

How do you know when GC is running? This brings up the question, how do you know when GC is complete. I was of the understanding that the state of GC was not accessible to the user. My system goes to sleep after 30 minutes as I'm trying to conserve power.