[Coco] megaread benchmarking

[Roger Taylor]

At 12:14 PM 5/20/2010, you wrote:


>If you're done playing tit-for-tat, then let's get back to the 
>business at hand:  benchmarks, performance, and throughput of our 
>respective products.
>
>You stated that you would be posting new performance numbers on 
>Drive Pak.  I'm anxious to see them.  And since you are so adept at 
>playing games, I'll lay out the rules for you:
>
>- Your numbers should reflect the rate at which your software 
>drivers read data from your device in some unit (bytes or bits) per second
>- The source for any software that you write to do the benchmarking 
>should be made available
>- Results should be demonstrably repeatable by other parties for verification
>
>According to my calculations, the maximum throughput that your pak 
>can attain is 45.51 seconds per megabyte of data read.  It will be 
>interesting to see how close you can get to that number.
>Good luck!



It doesn't matter how close I get to that figure... you guys have it 
out for me, and I'm really ultimately ok with that because it makes 
me grin more than I normally do.
As for me, I really don't waste time with worrying about what Cloud-9 
is actually doing or how it's not affecting my own services at all.

I'm working on a new NitrOS-9 driver set now for the native sector 
disk format but I won't rush anything for you.  I have customers to 
take care of first.

Btw, what are your megawrite numbers?   ;)

Benchmarking only one aspect of a pak's usage?




-- 
~ Roger Taylor