Javascript power to weight ratio. What's yours? Click here, wait for the result and divide it by your device weight in kilos. E.g the Aigo MID with ff3.0 scores 18 which, with 0.35kg gives a jp/w of 51.4 http://bit.ly/v8v3benchmark - http://code.google.com/apis...
I just tried a few times to run this on my iPhone 3G 16GB. First time, it hung Safari. Second time, fresh after a reboot, it never completed, even though I let it run for about 10 minutes. Both times, it. Got to 57% complete, and the scores were: Richards 2, DeltaBlue 2, Crypto 5. Hope that helps, let me know if you want me to run it again! :-)
- Josh Bancroft
first run on my iPaq 614c with PocketIE : same as Josh - got to 57%, then 71% then device hanged ... Second try with Netfront 3.5 : stopped at 43% .. Third shot with Opera 9.5beta: completed ! score: 8 / .145 = 55 :)))
- jean-paul
it hung on an S60 phone for me in the same place. Good to know that Opera completes the task although for some reason, their browsers always seem to be poor at Javascript.
- Steve 'Chippy' Paine
Still haven been able to test Skyfire to see how well ajax works and how the latency is. I guess I could RDP to my server at home and get a similar 1000+ result though ;-) I'll post why I dont like the proxy solutions soon.
- Steve 'Chippy' Paine
N800 for shits and giggles: 3/0,206 = 14,6. funny enough it wondered if i wanted to kill the javascript around step 3 or 4.
- turn.self.off
and now multiply the figure by the online battery life of your device and you have an index of internet mobility. Mmmm. Maybe im on to something here!
- Steve 'Chippy' Paine
remind me, that archos is using cortex, yes? got to be somewhat bothersome that the older N800 scored as much as it did. or maybe that say something about the microb javascript, even before any kind of optimization...
- turn.self.off
The archos 605 is the old one. Not Cortex bassed but Opera, while accurate, isnt the best at javascript. I will contact Jenn to see if i can get a result from her.
- Steve 'Chippy' Paine
ah yes, i got my numbers crossed. that makes a bit more sense. and i recall having issues with javascript (to the point i left it off most of the time) back on my 770. but then that was also a weaker device then the N800, so...
- turn.self.off
Fujitsu Loox U/B50n 1.6ghz with chrome v1.0154 = 395 / 0.58 = 681 (looks like latest chrome is slower!)
- Steve 'Chippy' Paine
tried tear just now, on the old N800, interestingly enough that got me, 18,8 / 0.206 = 87,38! what a difference that is compared to the microb test i did some time ago...
- turn.self.off
Confirms the 3gs as the most powerful javascript engine on a smartphone.
- Steve 'Chippy' Paine
Kohjinsha SC3 standard, Vista and Chrome 422/0.798 = 529
- Paul Graham
I did some more javascript tests this morning. Android/ARM has made amazing steps forward in the last year but Chrome/X86 (intel Atom at same clockrate) is still 5-10 times faster at processing javascript. A Cortex core takes 0.5W at 800Mhz. An Atom 800Mhz takes 2W at 800Mhz. Based on that, Moorestown could be a more efficient javascript platform.
...which brings me to the question. Are javascript engines written on X86 and then simply ported to ARM or are there javascript engines out there that are specifically written for ARM. (E.g. Android browser)
- Steve 'Chippy' Paine
Chrome OS is both for ARM and X86. But I think the ARM version will be more Interesting. While the x86 version might provide only slightly cheaper netbooks (saving $30 Windows clicense for one), could be even cheaper at $199 if with 8GB flash instead of hard disk. But ARM Chrome OS will be about bringing fast and full browser experience on the cheapest ARM laptops as cheap as $80 or...
more...
- Charbax
from Android
@benz145 Original result would be good. Melted umpcs dont count :-)