Benchmarks

The Writing test models common use cases with text processing applications.

Implementation

The test uses the LibreOffice Writer application and is implemented using AutoIt3 scripts.

In the copy and cut tests, the operation is repeated ten times to reduce random error. The secondary scores described in the Workload sub-chapter are then based on the geometric mean of the ten repeats.

Workloads

The Writing test simulates the work with documents. The workloads perform the following tasks:

1. Load Document 1, display in a window
2. Load Document 2, display in a window
3. Copy a large part of Document 1 and paste it into Document 2
4. Save As with Document 2
5. Resize Document 2 window 
6. Cut and paste parts of Document 2 around within the document
7. Save Document 2
8. Type some text in Document 2
9. Save Document 2
10. Insert some pictures from a local drive into Document 2
11. Save Document 2

The workloads measure the time it takes to load the documents, save the file, add pictures, and edit the document.

Load document    =    geomean(M_1,M_2)

Where: 
M_1    =    dbg_pcm10_writing_display_document_source
M_2    =    dbg_pcm10_writing_display_document_destination

Save document    =    geomean(M_3,M_4 ,M_5,M_6)

Where: 
M_3    =    dbg_pcm10_writing_save_document_as
M_4    =    dbg_pcm10_writing_save_document_1
M_5    =    dbg_pcm10_writing_save_document_2
M_6    =    dbg_pcm10_writing_save_document_3

Add pictures to document    =    geomean(M_7,M_8,M_9,M_10,M_11)

Where: 
M_7     =    dbg_pcm10_writing_editing_add_image_0
M_8     =    dbg_pcm10_writing_editing_add_image_1
M_9     =    dbg_pcm10_writing_editing_add_image_2
M_10    =    dbg_pcm10_writing_editing_add_image_3
M_11    =    dbg_pcm10_writing_editing_add_image_3

Copy and paste    
=    Max(0.12,geomean(M_12,M_13,M_14,M_15,M_16,M_17,M_18,M_19,M_20,M_21))

Where: 
M_12    =    dbg_pcm10_writing_editing_src_copy_0
M_13    =    dbg_pcm10_writing_editing_src_copy_1
M_14    =    dbg_pcm10_writing_editing_src_copy_2
M_15    =    dbg_pcm10_writing_editing_src_copy_3
M_16    =    dbg_pcm10_writing_editing_src_copy_4
M_17    =    dbg_pcm10_writing_editing_src_copy_5
M_18    =    dbg_pcm10_writing_editing_src_copy_6
M_19    =    dbg_pcm10_writing_editing_src_copy_7
M_20    =    dbg_pcm10_writing_editing_src_copy_8
M_21    =    dbg_pcm10_writing_editing_src_copy_9

Cut and paste    
=    Max(0.3,geomean(M_22,M_23,M_24,M_25,M_26,M_27,M_28,M_29,M_30,M_31))

Where: 
M_22    =    dbg_pcm10_writing_editing_dest_cut_0
M_23    =    dbg_pcm10_writing_editing_dest_cut_1
M_24    =    dbg_pcm10_writing_editing_dest_cut_2
M_25    =    dbg_pcm10_writing_editing_dest_cut_3
M_26    =    dbg_pcm10_writing_editing_dest_cut_4
M_27    =    dbg_pcm10_writing_editing_dest_cut_5
M_28    =    dbg_pcm10_writing_editing_dest_cut_6
M_29    =    dbg_pcm10_writing_editing_dest_cut_7
M_30    =    dbg_pcm10_writing_editing_dest_cut_8
M_31    =    dbg_pcm10_writing_editing_dest_cut_9

Writing score

We use a weighted harmonic mean to calculate the Writing score from the workload scores.

Writing score    =    K × 1 / geomean(R_1,R_2,R_3,R_4)

Where: 
R_3    =     Overall copy and cut   =     geomean(R_5,R_6)

Where:
K    =       scoring coefficient    =    4093

Where: