What does sending pictures and rendering pages have to do with javascript? I'm confused.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
iPhone 3GS Tops T-Mobile G1 and Palm Pre in JavaScript Benchmarks
- Thread starter MacRumors
- Start date
- Sort by reaction score
What does sending pictures and rendering pages have to do with javascript? I'm confused.
Let me add iPod Touch 2G Benchmark
Thanks 3.0 Copy&Paste
http://www2.webkit.org/perf/sunspider-0.9/sunspider.html
Thanks 3.0 Copy&Paste
http://www2.webkit.org/perf/sunspider-0.9/sunspider.html
Code:
--------------------------------------------
Total: 31864.4ms +/- 2.3%
--------------------------------------------
3d: 5623.0ms +/- 3.1%
cube: 1569.6ms +/- 4.6%
morph: 2125.8ms +/- 9.3%
raytrace: 1927.6ms +/- 6.4%
access: 3324.6ms +/- 3.7%
binary-trees: 460.8ms +/- 11.6%
fannkuch: 712.2ms +/- 6.0%
nbody: 1834.8ms +/- 3.8%
nsieve: 316.8ms +/- 2.4%
bitops: 1573.0ms +/- 9.6%
3bit-bits-in-byte: 245.4ms +/- 1.8%
bits-in-byte: 318.2ms +/- 1.7%
bitwise-and: 327.8ms +/- 3.4%
nsieve-bits: 681.6ms +/- 23.2%
controlflow: 308.8ms +/- 52.3%
recursive: 308.8ms +/- 52.3%
crypto: 1702.2ms +/- 0.9%
aes: 564.4ms +/- 4.7%
md5: 589.0ms +/- 4.1%
sha1: 548.8ms +/- 5.3%
date: 3916.4ms +/- 6.2%
format-tofte: 1436.2ms +/- 5.7%
format-xparb: 2480.2ms +/- 8.8%
math: 3863.6ms +/- 3.2%
cordic: 1283.2ms +/- 7.8%
partial-sums: 1878.2ms +/- 3.9%
spectral-norm: 702.2ms +/- 9.7%
regexp: 2610.2ms +/- 0.7%
dna: 2610.2ms +/- 0.7%
string: 8942.6ms +/- 1.8%
base64: 852.6ms +/- 5.0%
fasta: 1421.8ms +/- 5.5%
tagcloud: 2210.0ms +/- 1.9%
unpack-code: 2527.8ms +/- 2.9%
validate-input: 1930.4ms +/- 1.9%Is anyone actually shocked by this? Of course the iPhone is faster.
Superior Software + Superior Hardware = Superior Product.
Superior hardware? It's got virtually the same graphics core and CPU (ARM Cortex A8) as the Pre. Obviously it's software. The Pre is brand new, whereas the iPhone is in its third iteration. I'm not arguing the iPhone isn't running Javascript faster, but if you're gonna be smugly superior (for some reason) at least be right about it.
The good thing is, contests keep all sides constantly improving. Love it.
One thing about this benchmark, though. It says:
I wonder why they didn't find a DOM test to run as well? It's crucial in AJAX apps.
One thing about this benchmark, though. It says:
I wonder why they didn't find a DOM test to run as well? It's crucial in AJAX apps.
Then going by that theory, shouldn't some of the high end Nokia phones blow away anything else? ...in theory at least
Symbian supports Opera mini which runs circles around any other custom webkit browser.
So the iPhone 3GS did the worst? It took the most time to complete. It says time completed in seconds. I'd assume lower would be better, as in fewer seconds, which means faster. I must be reading this graph way wrong.
I must be reading this graph way wrong.I don't completely doubt your speed claim, but do you have any documentation on that?
Regardless, Opera on Macs is so buggy with so many websites and so much slower than Firefox and Safari, it really makes that claim difficult for me to believe or worthy of interest. Even if its true, Opera still has so many website incompatibilities, it's almost not worth trying just for that reason alone.
I give Opera a try like every year or so, and I'm always impressed with several clever things, but ultimately disappointed in actual regular usage.
Maybe the Windows version is significantly better ???
So the iPhone 3GS did the worst? It took the most time to complete. It says time completed in seconds. I'd assume lower would be better, as in fewer seconds, which means faster.
3GS <17 sec
Pre <39 sec
yes, you are reading it way wrong.
Yup. I expect that the Pre will almost match the 3GS when/if they update the webbrowser. Currently it's based on Webkit 525, just like Safari in iPhone with OS 2.2.1. Between Webkit 525 and 528 (in iPhone OS 3.0) they updated the Javascript interpreter to the new Squirrelfish Extreme engine, which speeds up Javascript tremendously. On the iPhone 3G you can see that the software update alone gave it a boost of 2.75x on this benchmark with the same hardware. If the Pre gets a similar boost when they update to Webkit 528, they should be able to complete it in 17.7s, which is almost as fast as the 3GS. Which is to be expected, as the CPUs of the 3GS and the Pre are very similar, and this Javascript benchmark is very CPU dependent.
The G1 is a different story. The CPU is quite different, but they might also be using a custom Javascript interpreter as as well.
You're right, however, Pre is $300 - $100 mail-in rebate = $200 except when you get it from BestBuy which would be $100 instant rebate.
G1 is $100 on Amazon, but so is iPhone 3G
The Benchmarks clearly state they are rendering javascript that is already downloaded. In the wild you are seeing the download plus render. So Sprint's network is either faster than AT&T's (possible) or just less congested (highly possible.) Remember, a cell tower is a shared connection. The more devices hit the tower, the less pie everyone gets. AT&T has the most customers of the networks, and therefore you can assume more customers per tower.
Last night I tested my iPhone 2G (3.0) against my friend's 3GS (both iPhones on EDGE)
at a bar where reception was not so great - ie where I got 74 kbps on Speedtest.net.
We loaded MacRumors home page as well as a few other sites, and his phone rendered them
significantly faster, although it took perhaps 30 seconds for anything to appear due to the weak signal.
I don't think people will ever get that coverage and speeds varies by market.
Overall, VERIZON and ATT probably have the best coverage & speed in most major markets.
But basically, your mileage may vary depending on where you live.
In my area, SPRINT is already doing 4G so their phones would have a distinct advantage not at all shown in these numbers, certainly in overall speed, BUT ONLY WHERE I LIVE, not where anyone ELSE LIVES!
Funny thing is, I've always found Sprint coverage around here to be pretty shabby, so what good is 4G if you can't get a signal? LOL
Ah haha!
Still ... quite impressive performance otherwise. Now next year's model:
... might be twice as fast ...
... might have a front facing camera ...
... might have a matte case ...
... might come with 64GB memory and more RAM ...
... might be less expensive than the 3GS ...
... might have Voice Activation (not Voice Control ... I KNOW what that is) support for BT devices ...
... might have better battery life ...
Hello Daniel...having trouble reading the posts the preceded your own? Or did you just skip right to bellyaching?
All this tests does is show that the iPhone's ability to render javascript is three times faster than your dad's pré. It does not measure anything else, such as how the AT&T cell towers in your neighborhood may be loaded heavily with digital traffic compared to Sprint's traffic. Don't forget AT&T is handling millions and millions of iPhone traffic compared to Sprint's thousands and thousands.
Finally, three bars to five bars doesn't speed up things. That's just a measure of signal strength. Once a phone has two or three bars, more won't speed things up.
Your accusation of media bias in this case is based on your own ignorance of the facts.
So the iPhone 3GS did the worst? It took the most time to complete. It says time completed in seconds. I'd assume lower would be better, as in fewer seconds, which means faster.
You are not only reading the graph way wrong, but you're not even reading the text that went with it.
