The Drive Performance Consistency Test is a long-running and extremely demanding test with a heavy, continuous load for expert users. 

The test includes a degradation phase, a steady state phase, and a recovery phase. In-depth reporting shows how the performance of the drive varies under these conditions.

The typical run time is 10 to 20 hours.  


The expected useful life of some storage devices is based on the number of write-erase cycles they perform. Running this benchmark repeatedly will significantly shorten the life time of some drives.

Traces used

Result file labelDescription
aftUsing Adobe After Effects
indUsing Adobe InDesign
pshUsing Adobe Photoshop (heavy use)
cp1Copying 4 ISO image files, 20 GB in total, from an secondary drive to the target drive (write test)
cp2Making a copy of the ISO files (read-write test)
cps1Copying 339 JPEG files, 2.37 GB in total, in to the target drive (write test)
cps2Making a copy of the JPEG files (read-write test)

These traces are included in metrics.csv included in the result zip file.

Test method

The general approach of the Drive Performance Consistency Test is defined by five phases:

  1. Precondition phase
  2. Degradation phase
  3. Steady state phase
  4. Recovery phase
  5. Clean up phase

All the benchmarking I/Os are performed only on the files created by the benchmark. User data on the drive is not touched by the test. 

Precondition phase

To prepare the drive for testing, it is filled almost completely with temporary files.

  1. Write the drive sequentially up to the reported capacity with random data, write size of 256 × 512 = 131,072 bytes. See setting precondition_block_size
  2. Write it through a second time, to take care of overprovisioning. See setting precondition_passes

Degradation phase

The aim of the degradation phase is to overwhelm the drive with write operations so that it is not able to perform its general housekeeping and optimization routines. As the name implies, the performance of the drive is expected to degrade during this phase.

  1. Run writes of random size between 8 × 512 and 2048 × 512 bytes on random offsets for 1 minute or until 50 GB has been written, whichever comes first. See settings degrade_duration_init, degrade_data_size_init, degrade_min_block_size, degrade_alignment, degrade_max_block_size
  2. Run a performance test with one trace. 
  3. Repeat steps 1 and 2 with each trace. For each repetition, we increase the duration of random writes in step 1 by 1 minute or 50 GB. See settings degrade_duration_increment and degrade_data_size_increment
  4. Repeat steps 1-3 for 8 passes. See setting degrade_passes.

Steady-state phase

In the steady-state phase, the last and most demanding pass of the degradation phase is repeated a couple of times. 

  1. Run writes of random size on random offsets (as in degradation step 1) for the final duration and data size achieved in the degradation phase.  
  2. Run a performance test with one trace.  
  3. Repeat steps 1 and 2 with each trace.  
  4. Repeat steps 1-3 three times. See setting steady_passes

Recovery phase

The aim of the recovery phase is to see how quickly the drive is able to recover and optimize its performance.

  1. Idle for 5 minutes. See setting recovery_duration_init
  2. Run a performance test with all traces.  
  3. Repeat steps 1 and 2 five times. See setting recovery_passes. With the recovery_duration_increment setting, you can optionally increase the idle time for each repetition. The default is no increase.

Clean up

This phase removes the temporary test data from the drive. User data on the drive is left untouched. 

  1. Write the drive sequentially up to the reported capacity with zero data, write size of 256 × 512 = 131,072 bytes. See settings postcondition_passes, postcondition_block_size

Scoring

The overall score of the Drive Performance Consistency Test score is the performance score from the steady-state phase. It represents the worst-case scenario for the drive’s performance. 

The final result from the recovery phase can be considered the best-case scenario for the drive’s performance.

Export result data to Microsoft Excel

The Drive Performance Consistency Test generates a large amount of result data that is not shown on the in-app result screen. You can export the result data as an Excel file. 

  1. Load the result on the Results screen in the PCMark 10 app.
  2. Click on the OPTIONS button 
  3. Select “Save as" -> Excel

The exported Excel file contains graphs that provide insights into the performance of the drive during each phase. In the graph below, data from several devices have been exported. This graph shows the performance during the cp1 trace. The points of phases are displayed on the x-axis of the graph.

The Excel file contains one sheet for every storage result included in the export. This sheet contains the data from pcmark_storage_trace_metrics.csv and pcmark_storage_trace_statistics.csv converted to separate tables. 

The Excel file contains two additional summary sheets named Trace and Aggregated. The different tables on these summary sheets are for the different traces aft, cp1, psh and others. Each column in these tables shows data from a correspondingly named sheet. Formulas are used to reference data from the data sheets. 

You can configure the metrics data shown in the tables and graphs using the pulldown menu in cell B1. Examples of values to show are rw_bw and rw_aat.