Benchmarks

This test simulates high-level use cases where the user browses common websites with a web browser application. The test uses the following website archetypes and use cases: social media, online shopping, map, video, and static web page.

Implementation

The Web Browsing test uses two browsers: Firefox and Google Chromium. Any other browsers possibly installed in the system will not affect the benchmark. 

The content is served with a local lightweight web server that is embedded into the benchmark. The content is custom made for the benchmark and represents common web sites.

The web pages are shown using both browsers, except the video page that is only run on Chromium. All the pages are run 2 times in both browsers.

Workloads

Social media

The social media workload simulates the usage of social media platforms and includes the following tasks:

• Navigates to and load a social media site.
• The page updates the news feed with new content.
• The page updates the feed again.

The workload measures the loading time of the page content and of the feed update.

Social media page load    =    geomean(M_1 / 1000, M_2 / 1000)

Where: 
M_1    =    dbg_pcm10_web_chromium_some_rendertime
M_2    =    dbg_pcm10_web_firefox_some_rendertime 

Social media feed update    =    geomean(M_3 / 1000, M_4 / 1000)

Where: 
M_3    =    dbg_pcm10_web_chromium_some_updatetime
M_4    =    dbg_pcm10_web_firefox_some_updatetime 

Online shopping

The online shopping workload simulates an online store. The workload performs the following tasks:

• View and zoom in on high-resolution images of shopping items.
• View 3D models of items.

The workload measures the time to view an image, load a 3D object, and animate a 3D object.

Shop view image    =    (Min(60,M_5))

Where: 
M_5    =    dbg_pcm10_web_chromium_webshop_zoom_fps

Shop load 3D object    =    geomean(M_6 / 1000, M_7 / 1000)

Where: 
M_6    =    dbg_pcm10_web_chromium_shop_modelparsetime
M_7    =    dbg_pcm10_web_firefox_shop_modelparsetime

Shop animate 3D object    =    geomean(Min(300,M_8),Min(300,M_9))

Where: 
M_8    =    dbg_pcm10_web_chromium_shop_fps
M_9    =    dbg_pcm10_web_firefox_shop_fps

Map

The map workload simulates the visualization of information on a map. The workload includes the following tasks:

• Navigate to and load a map site.
• The page adds useful graphics such as traffic information.
• Zoom in on the map.

The workload measures the time it takes to update the information on the map and the time to zoom in.

Map infographics update    =    geomean(M_10 / 1000, M_11 / 1000, M_12 / 1000, M_13 / 1000)

Where: 
M_10    =    dbg_pcm10_web_chromium_realestate_heatmapLoad
M_11    =    dbg_pcm10_web_firefox_realestate_heatmapLoad
M_12    =    dbg_pcm10_web_chromium_realestate_markers
M_13    =    dbg_pcm10_web_firefox_realestate_markers

Map zooming    =    geomean(M_14 / 1000, M_15 / 1000, M_16 / 1000, M_17 / 1000, M_18 / 1000, M_19 / 1000)

Where: 
M_14    =    dbg_pcm10_web_chromium_realestate_tileLoad01
M_15    =    dbg_pcm10_web_chromium_realestate_tileLoad02
M_16    =    dbg_pcm10_web_chromium_realestate_tileLoad03
M_17    =    dbg_pcm10_web_firefox_realestate_tileLoad01
M_18    =    dbg_pcm10_web_firefox_realestate_tileLoad02
M_19    =    dbg_pcm10_web_firefox_realestate_tileLoad03 

Video

The video workload simulates online video playback. The workload views a selection of HD and 4K UHD video clips using two codecs. The video workload measures the frame rate of the video playback.

Video H.264 1920 × 1080    =    M_20

Where: 
M_20    =    dbg_pcm10_web_chromium_video_video_fhd_mp4

Video H.264 3840 × 2160    =    M_21

Where: 
M_21    =    dbg_pcm10_web_chromium_video_video_uhd_mp4

Video VP9 1920 × 1080    =    M_22

Where: 
M_22    =    dbg_pcm10_web_chromium_video_video_fhd_webm

Video VP9 3840 × 2160    =    M_23

Where: 
M_23    =    dbg_pcm10_web_chromium_video_video_uhd_webm

Web Browsing score

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

Web Browsing score    =    
K × geomean(1 / R_1, 1 / R_2, R_3, 1 / R_4, R_5, 1 / R_6, 1 / R_7, R_8, R_9, R_10, R_11)

Where:
K             =     scoring coefficient     =     419