DAPL Command Benchmarks
Technical Note TN-164 Version 2.4
Benchmarks are valuable tools for evaluating the maximum sampling speed of certain types of applications and for comparing the performance of different Data Acquisition ProcessorsTM. Simple applications were written with a selected set of DAPL commands to provide performance benchmarks. These benchmarks can be helpful when deciding which Data Acquisition Processor is best for an application.
Following is an overview of the benchmark procedures, the hardware and software configuration used during testing, and a benchmark summary. A table presents benchmark results for various Data Acquisition Processors.
Test Procedures
Each DAPL application used to benchmark a command consisted of an input procedure sampling one channel. A 100 Hz square wave was applied to the inputs.
The benchmarks were run by setting the TIME command in the input procedures as low as possible without causing input channel pipe overflow. The DAPL command DISPLAY MEM was used to check memory use on the Data Acquisition Processor. Stable memory use indicated that the processing procedure could keep up with the input procedure sample rate. There were some normal fluctuations in memory use as the Data Acquisition Processor filled and emptied buffers. However, the maximum memory use reported by DISPLAY MEM did not increase over time.
Most applications transfer data to the PC. PC transfer tasks were replaced with PCOUNT tasks during the benchmark tests. This enables comparisons between Data Acquisition Processors without generating large volumes of data output. Since data values are not sent to the PC, the benchmarks are independent of computer performance.
Test Accuracy
The results of the benchmark tests are accurate to within five percent. TIME commands greater than 20 microseconds were changed in 1.0 microsecond increments. TIME commands less than 20 microseconds were changed by the smallest increment permitted by each Data Acquisition Processor (0.1 microseconds for DAP 4200a/526; 0.05 microseconds for DAP 4000a, DAP 840/103, DAP 5000a/526, DAP 5016a/527, and DAP 5200a/626; and 0.02 microseconds for DAP 5400a).
Table: Benchmark Results - Sampling Rate in kS/s
| Description | 4000a & 840/103 | 4200a/526 | 5000a/526 & 5016a/527 | 5200a/626 | 5400a/627 |
|---|---|---|---|---|---|
| AVERAGE 100 | 1111* | 1667* | 1667* | 1667* | 7547 |
| AVERAGE 10 | 833* | 1667* | 1667* | 1667* | 6667 |
| LIMIT & WAIT | 1000* | 1667* | 1667* | 1667* | 7143 |
| INTEGRATE | 189 | 500 | 1667* | 1667* | 3478 |
| RMS (100) | 526 | 1250* | 1667* | 1667* | 5940 |
| SCALE | 204 | 526 | 1428* | 1667* | 2963 |
| FIRFILTER 21 taps decm. 5 |
476 | 1000* | 1667* | 1667* | 4706 |
| FIRFILTER 101 taps decm. 5 |
179 | 357 | 909* | 1667* | 2667 |
| FIRFILTER 21 taps | 132 | 286 | 769** | 1667* | 2000 |
| FIRFILTER 101 taps | 40.3 | 80.0 | 208 | 869* | 740 |
| FFT 1024 pts. | 90.9 | 213 | 1052* | 1667* | 2222 |
| FFT 16384 pts. | 68.9 | 178 | 556** | 1428* | 1081 |
| THERMO | 66.7 | 163 | 909* | 1052* | 1739 |
| PID | 42.6 | 188 | 555* | 1000** | 941 |
|
NOTE: Numbers that appear in BOLD represent the maximum DAP sample rate. For the DAP 5400a, the sample rates are obtained by performing the same command on each of the eight simultaneously sampled channels. |
|||||
View other Technical Notes.