BenchmarkXPRT Blog banner

Category: HTML5

Interesting questions

We’ve had a couple of interesting questions about WebXPRT this week.

The first question was about the Face detect test in WebXPRT. One person, having noticed that changing the version of Firefox affected the WebXPRT score on a particular device, asked whether the test used the JavaScript Canvas element. The answer is yes, the Face detection test does use the Canvas element. It is based on the JavaScript library by Dr. Liu Liu.

As we have discussed in the past, the software stack on a device affects the benchmark scores. WebXPRT is a HTML5 benchmark and uses elements in the HTML5 specification, such as Canvas. Browsers implement HTML in their JavaScript engines, whose performance depends on the OS and the underlying platform.  So, WebXPRT scores are influenced by the browser and OS, as well as the platform.

The second question was whether it is possible to run WebXPRT without an Internet connection. Generally speaking, the answer to that is no. WebXPRT is a hosted application, and to run the official version, you must be able to connect to the WebXPRT servers.

However, community members can download the WebXPRT source and configure local servers that will run WebXPRT, if they desire. Note: As we discussed in Sources, any published results must be from the version hosted at

Thanks for the questions and keep experimenting!


Comment on this post in the forums

Ending the year with a bang!

As we promised in the blog post The newest member of the family, we made the WebXPRT 2013 community preview available this week. It has already been used in a review! The AnandTech review of the Acer Iconia W510 includes results from the WebXPRT 2013 community preview for that device and for the Microsoft Surface RT and the Apple iPad 4. The review has results from the TouchXPRT 2013 community preview for the Acer Iconia W510, Microsoft Surface RT, and the ASUS VivoTab RT as well.

Obviously, we’ve been doing some testing ourselves. Here’s a sampling of the devices on which we’ve successfully run WebXPRT:

Device Processor Operating system Browser Score Confidence interval
HP Envy 2 1.8 GHz Intel Atom Z2760 Windows 8 Internet Explorer 10.0.92 201 +/- 6
Asus VivoTab RT 1.2 GHz Tegra 3 T30L Windows RT Internet Explorer 10.0.92 160 +/- 5
Kindle Fire 1.2 GHz ARM Cortex-A9 Android OS 2.3 (customized: 6.3.1_user_4107720) Safari 5 92 +/- 2
ASUS-made Google Nexus 7 1.2 GHz Tegra 3 T30L Android 4.2 Chrome 18 201 +/- 4
Motorola DroidX phone 1 GHz TI OMAP3630-1000 Android 4.5.621 Browser version 2.3.4 26 +/- 1
iPhone 5 1.3 GHz Apple A6 iOS 6.0.2 Safari 6 168 +/- 2
iPad mini 1GHz Apple A5 iOS 6.0.2 Safari 6 110 +/- 1
iPad 4 1.4 GHz Apple A6X iOS 6.0.1 Safari 6 180 +/- 2


As the results above show, WebXPRT can run on a wide range of devices. We are working to get results on lots of different devices and would like your help. We’ll set up a forum thread for results that starts with these. We’ll then add additional ones we produce. Please respond in the thread with results you get.

In addition to performance results, the WebXPRT 2013 community preview also provides a report on the HTML 5 capabilities of your device. For those who want to know more about the capabilities of HTML 5, there’s more good news. The W3C community released the feature-complete spec for HTML 5 and Canvas 2D this week.

You can find an explanation of scenarios in the WebXPRT 2013 community preview, and an explanation of how it calculates its results in the WebXPRT 2013 CP1 Overview. Let us know what you think. There’s still time to help us shape the final version of both WebXPRT 2013 and TouchXPRT 2013.


Comment on this post in the forums

The newest member of the family

In his blog post TouchXPRT Web test update, Bill mentioned that we would be releasing Web-based workloads for the community to try out. Although we developed them as part of TouchXPRT, the cross-platform nature of these tests suggested to us that they should stand on their own.

After seeking community input, we have decided to make them a separate benchmark. So, we are proud to announce WebXPRT. The first community preview will be available mid-next week. WebXPRT CP1 contains four workloads: Photo Effects, Face Detect, Stocks Dashboard, and Offline Notes. Because the workloads are all HMTL5 based, they run on a wide variety of devices and operating systems—from iPad tablets to Android phones to Windows computers.

As with all community previews, we are very interested in your opinion. Tell us what you like or don’t like about the workloads. Are there other use cases you’d like to see?

Now that we have three benchmarks, the old HDXPRT-centric model of the community needs updating. Earlier this week, a message went out to the community announcing that we will be reorganizing the benchmarks under the umbrella of BenchmarkXPRT. This reorganization will touch all aspects of the community, from the Web site to Facebook, Twitter, and even the memberships themselves. We’ll be rolling out these changes over the next few weeks, and we’ll keep you informed every step of the way.


Comment on this post in the forums

TouchXPRT Web test update

On October 22, we released TouchXPRT CP1 to the community. We took the unprecedented step of releasing CP1 without any restrictions on publishing results, and since then reviews of the Microsoft Slate and the Sony Duo 11 Convertible Laptop have used TouchXPRT.

The five scenarios in CP1 focus on media manipulation. While this is an important activity on touch devices, we know this is not all people do.

Next week, we plan to release Web-based scenarios. They use HTML 5 for a variety of activities.  Unlike the original scenarios in TouchXPRT CP1, there will be nothing to download. You simply browse to a URL and run the tests online. There’s nothing to set up, just browse and run.

That means that there is nothing preventing you from running these tests on pretty much any system-browser combination that supports HTML5, not just on touch-based, Windows 8 devices like the rest of TouchXPRT. That started us wondering whether these Web-based activities should be thought of as a different benchmark entirely.  When these tests are available, please try them out and let us know what you think. Do you think they are worthwhile for a broader range of devices? Do you think their scenario-based emphasis is a good alternative to existing lower-level Web-based tests?

Please keep in mind that it’s not too late to give feedback on TouchXPRT CP1. Let us know how you like the scenarios on CP1 as well as what other activities you would like to see.

-Bill Catchings

Comment on this post in the forums

Check out the other XPRTs:

Forgot your password?