Post your Internet Browser Benchmark

[Maxxwire]

xXSevenXx- I would suggest that you try Peacemaker once again with Opera Turbo completely off and see if Opera 10.62 does any better. Your last Opera text parsing score was even higher than mine so it may be that Turbo is holding back the other scores due to delays caused by compression.

~Maxx~
.

 

[xXSevenXx]

xXSevenXx- I would suggest that you try Peacemaker once again with Opera Turbo completely off and see if Opera 10.62 does any better. Your last Opera text parsing score was even higher than mine so it may be that Turbo is holding back the other scores due to delays caused by compression.

~Maxx~
.

Here you go.

 

[Hopalong X]

These were done back to back.
Opera Turbo won't even load a web page. Too much compression.

 

[Paddyy]

opera and chrome <3

 

[yowanvista]

IE9 is not so fast

 

[Hopalong X]

New IE9 Benchmark

This will do other browsers but which ones I don't know yet.

WebVizBench

 

[Maxxwire]

WebVizBench is for HTML 5 isn't it?

~Maxx~
.

 

[Hopalong X]

Maxxwire

It is.
The way I read it the first time I tried it would run others but not in HTML5.
Tried Chrome and no go.

Oh well. You can test your IE9 and your GPU.
Mike

 

[Maxxwire]

Opera will run through the HTML 5 test, but I think that the Opera developers who are now working on 2 builds simultaneously are concentrating more on their new Presto 2.6.35 rendering engine than HTML 5.

~Maxx~
.

 

[royal tyrant]

Here's mine again.

 

[cluberti]

The PeaceKeeper benchmark isn't really an accurate description of javascript performance - they're using js timers to do the timing for starters, which given the accuracy of a js timer in milliseconds (anyone who's ever tried to use one and have it be anywhere near accurate knows the horrors I speak of) means you could have a browser score 9,000 when it's nowhere close to that score, or the converse - a browser that doesn't update date.now quickly will have a poor score. Also, the fact that the test turns some of the operations into no-ops as the test runs (but still counts the no-op times.... wtf???) means a browser that can handle a no-op that you will never see in real life the fastest, or that updates (specifically) Date.Now() will have a higher score than ones that do not. Neither of these are real-world areas of pain, but they play a very large part in the PeaceKeeper score. Top it off with running inside a Java VM (and the browser's ability to run Java code faster will have an impact on the ultimate score) means it's a crap benchmark, because it has nothing to do with the real world nor is it laid out with actual performance testing in mind - it just tests some "things" and the fastest browser at those "things" that are in no way real-world tests will get a huge score.

The test, frankly, is a disaster and not very accurate. Heck, sunspider is more accurate at least at timing the execution of certain js functions, and we already know that test is wildly inaccurate because some browser makers have been tweaking their js engine load-times so that these tests look faster too. It is what it is, and using a multitude of benchmarks are nice, but what is really a good idea is to clear your cache and visit the sites you normally visit and see which browser is fastest at doing what *you* do with a browser, not testing random js functions or HTML4 or 5, or some CSS or DOM functions, and determining which is faster at a "test". The best test is which one works the way you work, and runs the sites you use without getting in your way.

Really, that is the best benchmark.

 

[Maxxwire]

......................Firefox..................................Opera.......

This is a very simple and yet practical test that I developed using my Bandwidth Meter Pro where each bar represents 1 second of time to measure how quickly different browsers could download 5 sets (pages) of 64-80 Mb pictures that I have posted at a site I regularly visit doing what I normally do with a browser.

~Maxx~
.

 

[saakeman]

lunascapes scriped broke halfway !
opera 10 is doing great !
firefox beta crashed !

 

[Maxxwire]

opera 10 is doing great !

I've been using Opera for almost 2 years and its been a very exciting ride especially now since they have both v10.62 and v10.70 in development!

~Maxx~
.

 

[saakeman]

opera 10 is doing great !

I've been using Opera for almost 2 years and its been a very exciting ride especially now since they have both v10.62 and v10.70 in development!

~Maxx~
.

opera is a very cool browser, best looking interface, just sorry they have such a competitor like Google chrome
have you tried opera on any Linux systems
it bowled me away!
just like safari on mac

 

[Maxxwire]

I've run Opera on Win Vista and Win 7 and on both its so fast that I have measured download speeds that are only 0.6 Mbps slower than the 37 Mbps bandwidth of my modem! And with over 5,000 skins Opera runs great and looks good doing it!

~Maxx~
.

 

[Seven Eleven]

lunascapes scriped broke halfway !
opera 10 is doing great !
firefox beta crashed !

Peacekeeper used benchmark!
It was super effective!

 

[Hopalong X]

The PeaceKeeper benchmark isn't really an accurate description of javascript performance - they're using js timers to do the timing for starters, which given the accuracy of a js timer in milliseconds (anyone who's ever tried to use one and have it be anywhere near accurate knows the horrors I speak of) means you could have a browser score 9,000 when it's nowhere close to that score, or the converse - a browser that doesn't update date.now quickly will have a poor score. Also, the fact that the test turns some of the operations into no-ops as the test runs (but still counts the no-op times.... wtf???) means a browser that can handle a no-op that you will never see in real life the fastest, or that updates (specifically) Date.Now() will have a higher score than ones that do not. Neither of these are real-world areas of pain, but they play a very large part in the PeaceKeeper score. Top it off with running inside a Java VM (and the browser's ability to run Java code faster will have an impact on the ultimate score) means it's a crap benchmark, because it has nothing to do with the real world nor is it laid out with actual performance testing in mind - it just tests some "things" and the fastest browser at those "things" that are in no way real-world tests will get a huge score.

The test, frankly, is a disaster and not very accurate. Heck, sunspider is more accurate at least at timing the execution of certain js functions, and we already know that test is wildly inaccurate because some browser makers have been tweaking their js engine load-times so that these tests look faster too. It is what it is, and using a multitude of benchmarks are nice, but what is really a good idea is to clear your cache and visit the sites you normally visit and see which browser is fastest at doing what *you* do with a browser, not testing random js functions or HTML4 or 5, or some CSS or DOM functions, and determining which is faster at a "test". The best test is which one works the way you work, and runs the sites you use without getting in your way.

Really, that is the best benchmark.

No argument just a curiosity.
If Peacemaker is JAVA based then why do I have some of the highest scores posted with Chrome and Opera when I have no JAVA installed on my PC.
No JAVA/Sun software of any kind.:shock:

Things that make you go Hmmm???
Mike

 

[Dave76]

Everyone's browser choice is the best browser.

All benchmarks are tainted and not real world.

This is an intersting thread, it's meant to be just a bit of fun.

 

[cluberti]

No argument just a curiosity.
If Peacemaker is JAVA based then why do I have some of the highest scores posted with Chrome and Opera when I have no JAVA installed on my PC.
No JAVA/Sun software of any kind.:shock:

Things that make you go Hmmm???
Mike

You can run it without java, but on (for example) internet explorer or Firefox, you'll get a java window and part of the page runs from a java-based applet. Supposedly this doesn't affect the score, but in brief testing, machines where I did *not* run the java applet score consistently higher than the same test on the same machine in the same browser where I did run the java applet.

Very odd indeed.