Intro

The benchmarks were done using two test files (for mdf version 3 and 4) of around 170MB. The files contain 183 data groups and a total of 36424 channels.

asamdf 2.3.2 was compared against mdfreader 0.2.5. mdfreader seems to be the most used Python package to handle MDF files, and it also supports both version 3 and 4 of the standard.

The three benchmark cathegories are file open, file save and extracting the data for all channels inside the file(36424 calls). For each cathegory two aspect were noted: elapsed time and peak RAM usage.

Dependencies

You will need the following packages to be able to run the benchmark script

  • psutil
  • mdfreader

x64 Python results

The test environment used for 64 bit tests had:

  • 3.6.2 (v3.6.2:5fd33b5, Jul 8 2017, 04:57:36) [MSC v.1900 64 bit (AMD64)]
  • Windows-10-10.0.14393-SP0
  • Intel64 Family 6 Model 94 Stepping 3, GenuineIntel
  • 16GB installed RAM

Notations used in the results

  • nodata = MDF object created with load_measured_data=False (raw channel data not loaded into RAM)
  • compression = MDF object created with compression=blosc
  • compression bcolz 6 = MDF object created with compression=6
  • noDataLoading = MDF object read with noDataLoading=True

Files used for benchmark: * 183 groups * 36424 channels

Open file Time [ms] RAM [MB]
asammdf 2.3.2 mdfv3 831 371
asammdf 2.3.2 nodata mdfv3 609 190
mdfreader 0.2.5 mdfv3 3083 536
asammdf 2.3.2 mdfv4 1710 455
asammdf 2.3.2 nodata mdfv4 1349 260
mdfreader 0.2.5 mdfv4 30847 748
Save file Time [ms] RAM [MB]
asammdf 2.3.2 mdfv3 348 371
asammdf 2.3.2 nodata mdfv3 343 197
mdfreader 0.2.5 mdfv3 21244 1997
asammdf 2.3.2 mdfv4 530 462
asammdf 2.3.2 nodata mdfv4 522 272
mdfreader 0.2.5 mdfv4 19594 2795
Get all channels (36424 calls) Time [ms] RAM [MB]
asammdf 2.3.2 mdfv3 681 383
asammdf 2.3.2 nodata mdfv3 9175 209
mdfreader 0.2.5 mdfv3 29 537
asammdf 2.3.2 mdfv4 599 464
asammdf 2.3.2 nodata mdfv4 12191 273
mdfreader 0.2.5 mdfv4 38 748

x86 Python results

The test environment used for 32 bit tests had:

  • Python 3.6.1 (v3.6.1:69c0db5, Mar 21 2017, 17:54:52) [MSC v.1900 32 bit (Intel)]
  • Windows-7-6.1.7601-SP1
  • Intel64 Family 6 Model 94 Stepping 3, GenuineIntel (i7-6820Q)
  • 16GB installed RAM

The notations used in the results have the following meaning:

  • nodata = MDF object created with load_measured_data=False (raw channel data no loaded into RAM)
  • compression = MDF object created with compression=True (raw channel data loaded into RAM and compressed)
  • noconvert = MDF object created with convertAfterRead=False

Raw data

  • 3.6.1 (v3.6.1:69c0db5, Mar 21 2017, 17:54:52) [MSC v.1900 32 bit (Intel)]
  • Windows-10-10.0.14393-SP0
  • Intel64 Family 6 Model 94 Stepping 3, GenuineIntel
  • 16GB installed RAM

Notations used in the results

  • nodata = MDF object created with load_measured_data=False (raw channel data not loaded into RAM)
  • compression = MDF object created with compression=True/blosc
  • compression bcolz 6 = MDF object created with compression=6
  • noDataLoading = MDF object read with noDataLoading=True

Files used for benchmark: * 183 groups * 36424 channels

Open file Time [ms] RAM [MB]
asammdf 2.3.2 mdfv3 980 288
asammdf 2.3.2 nodata mdfv3 670 118
mdfreader 0.2.5 mdfv3 3776 455
asammdf 2.3.2 mdfv4 2071 342
asammdf 2.3.2 nodata mdfv4 1610 160
mdfreader 0.2.5 mdfv4 43559 578
Save file Time [ms] RAM [MB]
asammdf 2.3.2 mdfv3 406 291
asammdf 2.3.2 nodata mdfv3 432 125
mdfreader 0.2.5 mdfv3 19623 1224
asammdf 2.3.2 mdfv4 691 351
asammdf 2.3.2 nodata mdfv4 734 169
mdfreader 0.2.5 mdfv4 17657 1687
Get all channels (36424 calls) Time [ms] RAM [MB]
asammdf 2.3.2 mdfv3 963 298
asammdf 2.3.2 nodata mdfv3 19059 132
mdfreader 0.2.5 mdfv3 34 455
asammdf 2.3.2 mdfv4 868 349
asammdf 2.3.2 nodata mdfv4 20434 171
mdfreader 0.2.5 mdfv4 54 578