358
System Performance Analysis (SPA) Software
System Performance Analysis Software
Sampling methods and data sorting
SPA provides a statistical summary of target system behavior over
time. The greater the number of samples, the more accurate the
statistics. Therefore, SPA should always be run in the Repetitive mode.
By doing this, the analyzer will continue to sample the data and update
the display until Stop is pressed or until a sampling variable is changed
on the display.
After SPA completes an acquisition, each sample in the current
acquisition is compared to the ranges or buckets for the current trace
mode. If the sample matches a range or bucket, it is sorted into that
range or bucket. This process is repeated for every sample in the
acquisition. After the entire acquisition has been sorted, the
histograms and displayed statistics are updated, and the analyzer is re-
armed for the next acquisition.
Refer to the following sections on the three trace modes for details on
sorting criteria and statistical computation.
Between each successive acquisition, there is blind time during which
the captured data is unloaded and sorted, the display is updated, and
the analyzer is restarted. The length of the blind time is a function of
the complexity of the SPA measurement.
Selecting and changing labels
A label in Agilent Technologies logic analy zers is defined in the Format
menu. A label is any named group of 32 or fewer data channels.
The State Overview and State Histogram modes monitor one label at a
time. Any labels that are defined in the Format Specification are
available. A typical example is the ADDR (address) label used in many
of the Agilent Technologies inverse assembler configurations. Often,
SPA measurements will use the ADDR label to monitor memory
activity.