Similar in concept to head switch time,
cylinder switch time is the time that elapses when the drive finishes reading (or
writing) all the data on a given cylinder and needs to switch to the next one. This
normally only occurs during fairly long reads or writes, since the drive will read all the
tracks in a cylinder before switching
cylinders. Cylinder switch time is slower than head switch time because it involves a
mechanical process: moving the actuator assembly. It is usually somewhere around 2 to 3
milliseconds.
Note: You might think
that cylinder switch time would be the same as track-to-track seek
time, after all, it's the same thing, isn't it? They aren't the same however, because
cylinder switch time includes all of the overhead time that passes from the time the read
stops on one track until it starts again on the next one. This is why cylinder switch
times are typically double those of track-to-track seeks.
Cylinder switch time is another specification that is fairly obscure and not commonly
discussed, but is an important component of sustained
transfer rate, since STR measures transfer rate over larger reads or writes that can
encompass more than one cylinder. See the discussion of sustained transfer rate for more details. You may also want to read about head and cylinder skew here.
Cylinder switch time is influenced by the characteristics of the hard disk's controller as well as its actuator mechanics. It does not vary greatly from
drive model to model or between manufacturers.
Next: Internal Sustained Transfer Rate (STR)
