Most mainstream users will probably use the stock browser, but probably not most Anandtech readers. So, why are you not testing with Chrome? If you did we'd be able to compare the performance against our own current devices.
If they can watch youtube and install GL Benchmark and Linpack, they can install Chrome just the same. Plus, they don't run benchmarks on Chrome even on review units. This is simply unacceptable, especially without any reaction to probably the most requested benchmark improvement...
Chrome would be great but I'd also like to see Firefox. Also, consider adding the v8 and kraken benchmarks. You complain about how old the sunspider benchmark is but then neglect to run modern ones. I'd be especially curious to see how the new iphone does with the newer browser tests.
I concur, especially since the clear move is towards Chrome <i>being</i> the stock browser (see: Nexus 7). Since you'd be able to use the same, unadulterated Javascript and rendering engine on every Android device, it would actually make the numbers meaningful.
I also recommend testing with Chrome on iOS. You know, for fairness., and totally not because Apple's anti-competitive lockout of JIT for non-stock browsers would cause scores to tank.
For a lot of phones it's probably that the review units were returned/resold. Retesting for new OS versions is only possible if someone at AT still has one.
Don't worry ! I've already tried a htc test rom on jelly bean. The stock browser scored a whooping 1196.x ms.. The browser has really improved. It would be my stock rom, but sadly (as far as i'm informed) radio is currently not working due to older hboot.
While they're(LG) working on optimizing the software, they can also put some efforts on efficiency. I think Android is more than powerful enough with a lot of neat features, I think we should focus on battery life and efficiency now.
I concur, most tech savvy people will not be using the stock browser so testing on other browsers that are more in popular use makes more sense. Pretty much everyone prefers Chrome so why not test on that?
Besides, what are you exactly trying to test, the latest software optimisations or the SoC? If the latter, then using a third party browser that's not modified by manufacturers and actually is the same browser on different devices would make a lot more sense than the OEM-modified one.
Also, please give Firefox a new chance and perhaps include its scores as well. I find recent versions of FF (after FF15) to be a very capable browser on Android which actually gives Chrome a good run for its money.
I'm not sure why people are expecting to see a linear improvement going from 2 cores to 4. Twice as many cores does not equal twice the performance in any real-life scenario. Cyanogen mods are not going to be able to change that (though they could improve the threading somewhat, possibly).
When do you think we get something like Q-slide inside the US? Do you think Google will add a similar feature to future Android Operating Systems? How can we download an application or feature like this to our existing Android phones?
hi brian why don't you consider V8 benchmark (version7) suite from google (http://v8.googlecode.com/svn/data/benchmarks/v7/ru... or the earliest octane benchmark? The test documentation is also very clear and it test more complex javascript compare to sunspider. but unfortunately there are some compatibilty issue with google latest javascript benchmark i.e:octane, that will skip older browser or older OS
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
21 Comments
Back to Article
JaPeL - Thursday, September 20, 2012 - link
Can you ask If they are going to use ISDB-t for japan and south america? that would be SO awesomej85 - Thursday, September 20, 2012 - link
Most mainstream users will probably use the stock browser, but probably not most Anandtech readers. So, why are you not testing with Chrome? If you did we'd be able to compare the performance against our own current devices.ratte - Thursday, September 20, 2012 - link
Because they don't have a review unit ?MantasPakenas - Thursday, September 20, 2012 - link
If they can watch youtube and install GL Benchmark and Linpack, they can install Chrome just the same. Plus, they don't run benchmarks on Chrome even on review units. This is simply unacceptable, especially without any reaction to probably the most requested benchmark improvement...ssj4Gogeta - Thursday, September 20, 2012 - link
+1 for testing with Chrome.@j85: I wonder how fast Lt. Commander Data would run SunSpider :P
tuxRoller - Thursday, September 20, 2012 - link
Chrome would be great but I'd also like to see Firefox.Also, consider adding the v8 and kraken benchmarks. You complain about how old the sunspider benchmark is but then neglect to run modern ones.
I'd be especially curious to see how the new iphone does with the newer browser tests.
lowlymarine - Thursday, September 20, 2012 - link
I concur, especially since the clear move is towards Chrome <i>being</i> the stock browser (see: Nexus 7). Since you'd be able to use the same, unadulterated Javascript and rendering engine on every Android device, it would actually make the numbers meaningful.I also recommend testing with Chrome on iOS. You know, for fairness., and totally not because Apple's anti-competitive lockout of JIT for non-stock browsers would cause scores to tank.
powerarmour - Friday, September 21, 2012 - link
I agree, using the Chrome browser would be a much fairer test.powerarmour - Thursday, September 20, 2012 - link
Shame there are a few phones on that list that haven't been updated with their Jellybean stock browser scores, namely the Nexus devices.Galaxy Nexus gets 122k on Browsermark and ~1500 on Sunspider now :)
DanNeely - Thursday, September 20, 2012 - link
For a lot of phones it's probably that the review units were returned/resold. Retesting for new OS versions is only possible if someone at AT still has one.Azurael - Thursday, September 20, 2012 - link
Damn, they beat the One X running HTC 2.17/Android 4.0.4. 1475.4ms... Still, we've got 4.1 to come yet :)shadeh - Sunday, September 23, 2012 - link
Don't worry ! I've already tried a htc test rom on jelly bean. The stock browser scored a whooping 1196.x ms.. The browser has really improved. It would be my stock rom, but sadly (as far as i'm informed) radio is currently not working due to older hboot.Can't wait to see another leak.
toyotabedzrock - Thursday, September 20, 2012 - link
What about Chrome performance and an upgrade to 4.1?amdwilliam1985 - Thursday, September 20, 2012 - link
While they're(LG) working on optimizing the software, they can also put some efforts on efficiency.I think Android is more than powerful enough with a lot of neat features, I think we should focus on battery life and efficiency now.
snoozemode - Thursday, September 20, 2012 - link
Seems to me that Q-slide is the answer to multitasking on small screen devices.aryonoco - Thursday, September 20, 2012 - link
I concur, most tech savvy people will not be using the stock browser so testing on other browsers that are more in popular use makes more sense. Pretty much everyone prefers Chrome so why not test on that?Besides, what are you exactly trying to test, the latest software optimisations or the SoC? If the latter, then using a third party browser that's not modified by manufacturers and actually is the same browser on different devices would make a lot more sense than the OEM-modified one.
Also, please give Firefox a new chance and perhaps include its scores as well. I find recent versions of FF (after FF15) to be a very capable browser on Android which actually gives Chrome a good run for its money.
owned66 - Friday, September 21, 2012 - link
it should be two times faster than the S3 !!i guess we need to wait for cyanogenMod devs to take advantage of it
kyuu - Friday, September 21, 2012 - link
I'm not sure why people are expecting to see a linear improvement going from 2 cores to 4. Twice as many cores does not equal twice the performance in any real-life scenario. Cyanogen mods are not going to be able to change that (though they could improve the threading somewhat, possibly).Or am I misunderstanding?
1008anan - Saturday, September 22, 2012 - link
I learnt something about Java script performance.When do you think we get something like Q-slide inside the US? Do you think Google will add a similar feature to future Android Operating Systems? How can we download an application or feature like this to our existing Android phones?
evan_hit - Wednesday, October 3, 2012 - link
hi brianwhy don't you consider V8 benchmark (version7) suite from google (http://v8.googlecode.com/svn/data/benchmarks/v7/ru... or the earliest octane benchmark? The test documentation is also very clear and it test more complex javascript compare to sunspider. but unfortunately there are some compatibilty issue with google latest javascript benchmark i.e:octane, that will skip older browser or older OS
evan_hit - Wednesday, October 3, 2012 - link
sorry i mean the latest javascript benchmark from google : octane (http://octane-benchmark.googlecode.com/svn/latest/...