IPC@CHIP® RTOS-PPC – API Documentation

Header image

Main page


Performance comparison

RTOS task switch time

The following table shows information about the @CHIP-RTOS-PPC task switch times.

 

T1 [us]

T2 [us]

T3 [us]

T4 [us]

SC12

104

80

100

60

SC13/SC11

35

35

36

26

SC13-LF/SC11-LF

36

40

42

29

SC123/SC143/SC23/SC24

8,2

8,2

9,0

6,4

SC243/SC223

1,1

1,7

2,0

2,3

Table 1) Task switch times

The following diagramm and C code shows how the measured values were calculated.

taskswitch.gif
Example:
void taskA(void)
{
   RTX_Sleep_Request();
}

void taskB(void)
{
   RTX_Restart_Task (taskAID);
   RTX_Wakeup (taskAID);
}

void main (void)
{
   //...
   result = RTX_Create_Task_Without_Run(&taskAID , &taskAdefblock);
   RTX_Sleep_Time(10);
   result = RTX_Create_Task(&taskBID , &taskBdefblock);
   //...
}

Interrupt latency

The interrupt handler latency is the typical time from the processor's first response to an interrupt request signal through to the first useful instruction inside of the user interrupt service procedure.

The user interrupt service procedure can be of type HW API or RTX.

 

HW API handler latency [us]

RTX handler latency [us]

SC12

84

90

SC13/SC11

21

25

SC13-LF/SC11-LF

22

26

SC143/SC123/SC23/SC24

6,5

7,3

SC243/SC223

1,2

1,6

Table 2) Interrupt latency

TCP echo

A TCP Echo client application, running on Windows 7 (Intel Pentium IV, 2,8 GHz) establishes a TCP connection to an IPC@CHIP® TCP Echo server application. The TCP server echoes the incoming data back to the client. The client application measures the response time and calculates the amount of databytes in KBytes/sec, which the IPC@CHIP® is able to echo back to the client.

The tests were made with 10 MBytes (300 MBytes for SC243) of data.

This test was made under special test conditions. We cannot guarantee that the measured performance results are achievable on other test environment and conditions.

 

Echo
transfer rate
KByte/s

Receive only
transfer rate
KByte/s

Send only
transfer rate
KByte/s

SC12
10 Mbit/s, RTOS V1.25

98

184

165

SC13-LF
100 Mbit/s, RTOS V1.25

263

574

443

SC1x3/SC2x
100 Mbit/s, RTOS V1.20

697

1395

1220

SC243
100 Mbit/s, RTOS V1.05 Beta

6211

11295

10587

Table 3) TCP performance
tcp_benchmark.gif

UDP echo

An UDP Echo client application, running on WinXP (Intel Pentium IV, 2,8 GHz) sends UDP datagrams of different data sizes to an IPC@CHIP® UDP Echo server application. The UDP server echoes the incoming datagramms back to the client. The client application measures the response time and calculates the amount of databytes in KBytes/sec, which the IPC@CHIP® is able to echo back to the client in dependency of the datagram size. The communication peers (WinXP PC and IPC@CHIP®) are connected with a "Twisted pair crossover cable" to avoid disturbtion by other network traffic.

This test was made under special test conditions. We cannot guarantee that the measured performance results are achievable on other test environment and conditions.

 

UDP Echo
1472 bytes
KByte/s

SC12
10 Mbit/s, RTOS V1.25

108

SC13-LF
100 Mbit/s, RTOS V1.25

371

SC1x3/SC2x
100 Mbit/s, RTOS V1.20

919

SC243
100 Mbit/s, RTOS V0.80

3071

udp_benchmark.gif





Top of page | Main page

Copyright © 2017 Beck IPC GmbH
Generated on Thu Jan 26 16:21:35 2017 by Doxygen 1.6.1