System Performance
Original instructions
About us
Overview
The formulas that apply to typical M580 system for computing the minimum cycle time of tasks and the application response time can be used when you integrate an M580 system to an open Ethernet network.
Calculating a Minimum MAST Cycle
If only the MAST task is configured, the minimum MAST cycle time (in ms) can be calculated as follows:
The minimum cycle time for other tasks can similarly be estimated:
If multiple tasks need to be configured, satisfy the following conditions (where all cycle times are measured in ms):
(number of drops using MAST task) / (MAST cycle time) + (number of drops using FAST task) / (FAST cycle time) + (number of drops using AUX0 task) / (AUX0 cycle time) + (number of drops using AUX1 task) / (AUX1 cycle time) < 1.5
If DIO devices are configured, the minimum cycle time needs to be increased.
Application Response Time Example
Application response time (ART) is the time a CPU application takes to react to an input, starting when the input signal triggers a write command from the CPU and ending when the corresponding output module changes state.
The ART example values (provided in the table below) are calculated from the below open M580 system associated with the MAST task: 
1
M580 CPU with RIO scanner service in the local rack 
2
Industrial Ethernet managed switch 
3
X80 RIO drop with BM•CRA312•0 X80 EIO adapter module.
Recall that the ART formula is:
ART= (2*CRA_Drop_Process) + (RPI) + (Network_In_Time) + (Network_In_Jitter) + (CPU_In_Jitter) + (2*CPU_Scan) + (CPU_Out_Jitter) + (Network_Out_Time) + Network_Out_Jitter)
For a detailed explanation of each parameter, refer to the ART Computation Parameters topic.
The maximum values calculated for the ART computation example with CPU scan time of 50 ms and RPI of 25 ms are:
Parameter
Description and formula
Maximum value (ms) for the M580 open architecture example
CRA_Drop_Process
BM•CRA312•0 processing time
CRA_Drop_Process = The sum of BM•CRA312•0 input scan time and queue delay.
CRA_Drop_Process = 4.4
RPI
BM•CRA312•0 input RPI value is depending to the defined CPU task
Default values are:
  • 0.5 * CPU period if the MAST task is in periodic mode.
  • watchdog/4 if the MAST task is in cyclic mode.
RPI = 0.5 * 50 = 25
Network_In_Time
Network input time
Network_In_Time = (network delay based on I/O packet size) * (the number of hops(1) the packet travels)
Network_In_Time = (0.078 * 4) = 0.312
  • 0.078 s is the estimated network delay based on I/O packet size of 800 bytes. The network delay values are estimated and given in the table below.
  • 4 is the number of hops the packet travels
I/O packet size (bytes):
Estimated network delay (μs):
128
26
256
35
400
46
800
78
1200
110
1400
127
Network_In_Jitter
Network input jitter is the delay occasioned by devices
Network_In_Jitter = [(network delay) * (number of RIO drops)] + [(network delay) * (number of distributed equipment hops(1))]
Network_In_Jitter = (0.078 * 3) = 0.234
  • 0.078 s is the estimated network delay based on I/O packet size of 800 bytes.
  • 3 is the number of RIO drops
  • There is no distributed equipment
CPU_In_Jitter
CPU Ethernet I/O scanner service input jitter: CPU input queue delay (owing to RIO drops and DIO traffic) to read packet.
CPU_In_Jitter = (1 + (0.07 * (number of RIO drops))
CPU_In_Jitter = (1 + (0.07 * 3)) = 1.21
  • 3 is the number of RIO drops.
CPU_Scan
CPU scan time
User defined, based on application.
CPU_Scan = 50
CPU_Out_Jitter
CPU Ethernet I/O scanner service output jitter: CPU output queue delay (owing to RIO drops).
CPU_Out_Jitter = (1 + (0.07 * (number of RIO drops))
CPU_Out_Jitter = (1 + (0.07 * 3)) = 1.21
  • 3 is the number of RIO drops.
Network_Out_Time
Same formula as network input time
Network_Out_Time = (network delay) * (the number of hops(1) the packet travels)
Network_OUT_Time = (0.078 * 4) = 0.312
  • 0.078 s is the estimated network delay based on I/O packet size of 800 bytes.
  • 4 is the number of hops the packet travels
Network_Out_Jitter
Same formula as Network input jitter without I/O frames from RIO drops
Network_Out_Jitter = (network delay) * (number of distributed equipment hops(1))
Network_Out_Jitter = 0
Does not apply. No distributed equipment is connected to the RIO network.
CRA_Drop_Process
BM•CRA312•0 process time
CRA_Drop_Process = The sum of the BM•CRA312•0 X80 EIO adapter module output scan time and queue delay
CRA_Drop_Process = 4.4 
(1)
The maximum potential hop count, which represents the maximum number of switches a packet might pass through from the adapter module BM•CRA312•0 to the CPU. This count includes the switches in both the BM•CRA312•0 X80 EIO adapter module and the CPU with Ethernet I/O scanner service.
Thus for the M580 open network architecture example, the max ART = (2*4.4) + 25 + 0.312 + 0.234 + 1.21 + (2*50) + 1.21 + 0.312 + 0 = 137.078 ms