Are seek times "built-in" to STR tests?

[SlipperyPete]

So I always had this conception that sustained transfer rates were determined by the benchmarking program writing a bunch of data in sequence which would minimize seek times. Then I realized I was only assuming this, and that STR tests may write data randomly, taking seek times into account. So, say when HDTach says a drive can do a 130MB/sec sustained transfer rate, with an access time of 13.5ms, is this better than a drive that can push 110MB/sec with an access time of 5.6ms (15k)?

 

[drescherjm]

No the 110MB/sec with an access time of 5.6ms (15k) drive will be faster in the real world then the 130MB/s drive with a 13.5ms access time. The 130MB/s drive would only be faster if you loaded a large (many MB) file sequentially. This is not an access pattern that happens in most applications.