Web browsing is no faster on Ipad2 - 4.3 is the main factor

[fertilized-egg]

...
I’m guessing you are not a scientist. If you were, I’m sure you would see the value of keeping variables in a comparison at a minimum. This is basic science, KEEP THE NUMBER OF VARIABLES TO A MINIUM. Try not to add too many exogenous factors that will lead to incorrect conclusions.
...
Now, if they were running the same code, sure it would be a good point. However, they are not.

While they don't run the same code, we already know the new Nitro Javascript engine performs about the same as Android's V8 engine under SunSpider. Since you love benchmark and facts, here is the link for that (http://arstechnica.com/apple/news/2...d-chrome-opera-still-have-javascript-edge.ars)

Thus then:

1) CNet posted SunSpider Javascript benchmark.
2) The new iPad 2 "only performed 1.5 times as fast." (Ars' wording)
3) Other A9 dual core chips performed 1.5 times as fast as A8 chips in SunSpider as I've showed you in the benchmark graph.
4) The new iPad 2's speed matches Tegra2's speed, both are using Javascript engines that perform similarly under SunSpider

So please tell me. Where do you see the evidence behind your assertion that the A5 isn's the Cortex A9? Android A9 dual core devices performed exactly the same in relation to Android A8 devices as the iPad 2 did against iPad 1. We know the Safari Nitro engine does about the same as the Google V8 in SunSpider, and indeed iPad 2 and Tegra2 devices gotten virtually identical results.

Now where's your proof against all these? I'd like to hear it since you're so insistent on being rational and relying only on empirical benchmark, not marketing. I'm all ears as to what you'll present as facts to prove your point as opposed to saying some rhetoric about proper modeling technique and being science-y.

 

[Coukos34]

While they don't run the same code, we already know the new Nitro Javascript engine performs about the same as Android's V8 engine under SunSpider. Since you love benchmark and facts, here is the link for that (http://arstechnica.com/apple/news/2...d-chrome-opera-still-have-javascript-edge.ars)

Thus then:

1) CNet posted SunSpider Javascript benchmark.
2) The new iPad 2 "only performed 1.5 times as fast." (Ars' wording)
3) Other A9 dual core chips performed 1.5 times as fast as A8 chips in SunSpider as I've showed you in the benchmark graph.
4) The new iPad 2's speed matches Tegra2's speed, both are using Javascript engines that perform similarly under SunSpider

So please tell me. Where do you see the evidence behind your assertion that the A5 isn's the Cortex A9? Android A9 dual core devices performed exactly the same in relation to Android A8 devices as the iPad 2 did against iPad 1. We know the Safari Nitro engine does about the same as the Google V8 in SunSpider, and indeed iPad 2 and Tegra2 devices gotten virtually identical results.

Now where's your proof against all these? I'd like to hear it since you're so insistent on being rational and relying only on empirical benchmark, not marketing. I'm all ears as to what you'll present as facts to prove your point as opposed to saying some rhetoric about proper modeling technique and being science-y.

Forget Full-of-win, he has an agenda and tries to say things that ruffle feathers. He still can't answer a simple question on why he considers it an iPad 1.5. His feeble attempt at interpreting data is not going to effect the facts. I'm sick of people like that on these forums (heck there was actually an 8 page thread about the iPad 2 not being thinner ). While he thinks he knows alot about science, he obviously has no idea about technology.

 

[ghostlyorb]

Well, it seems like iOS 4.3 (with its new javascript engine) will make the ipad 1 just as fast as the new ipad 2.

Check out this story from ars technica

So, now I'm begging to wonder if this upgrade is worth it. No ram increases, web browsing the same, a cheap toy webcam on front and back, and 2 ounces lighter...what do you all think?

I think it will be faster.. no matter what you say. =P

 

[yodaxl7]

Wow! This was a fun read. Can wait to get my new iPad 2!!!

 

[iMule]

If you can run your OS efficiently on 1MB, so be it. If 256 MB cripples the device and ruins it's speed and smoothness, then there's a problem. I honestly think some people assume that Apple owes them 512 MB for the sole purpose of upgrading the device. In my opinion though, it is amazing that a tablet device gets away with 256 MB.

 

[fertilized-egg]

If you can run your OS efficiently on 1MB, so be it. If 256 MB cripples the device and ruins it's speed and smoothness, then there's a problem. I honestly think some people assume that Apple owes them 512 MB for the sole purpose of upgrading the device. In my opinion though, it is amazing that a tablet device gets away with 256 MB.

The bad part is, I'm pretty darned sure it'll come with 512 MB. If I'm wrong I'll feel foolish and curse Apple but I'd be more than willing to do that. Those who complain based on the 256MB rumor..they'll just change their tune and will loudly complain that it's not 1G.

 

[vincenz]

That's a bummer because web browsing on the original iPad could be a tad faster.

 

[Sydde]

Are we arguing on a rumor site?

 

[admanimal]

That's a bummer because web browsing on the original iPad could be a tad faster.

Then you're in luck, because what this article actually shows is that one important aspect of web browsing is 50% faster on a iPad 2 than the original iPad. That counts as a tad, right? Of course this is strictly Javascript performance and does not take into consideration loading/rendering pages, which can certainly be multithreaded and take advantage of the faster graphics as well.

 

[MythicFrost]

I was comparing two devices, running the same version of an operating system (iOS 4.3), and one assumes the same code to run Java. When such a comparison is made, and the variable is the hardware running said software, the so-called iPad 2 (or iPad 1.5) is not much faster than the original iPad.

I’m guessing you are not a scientist. If you were, I’m sure you would see the value of keeping variables in a comparison at a minimum. This is basic science, KEEP THE NUMBER OF VARIABLES TO A MINIUM. Try not to add too many exogenous factors that will lead to incorrect conclusions. Do you not agree? When you bring in different OS’es and code bases, you are adding too many variables to be informational. Who is to say Goolge will not offer a refined version of their software tomorrow, making your position moot.

Now, if they were running the same code, sure it would be a good point. However, they are not.

You kidding? The iPad 2 is getting 60% more performance in an environment that isn't multi-threaded -- I.E, getting 60% more performance from one core than the iPad 1 was.

The CPU will be an ARM Cortex A9 because it offers more punch and less power consumption at the same clock speed, thus having an advantage over the A8 in the iPad. It certainly won't make up all of that additional 60% performance in that SunSpider benchmark, the rest would be from that same core not having to handle system/background tasks as well unlike in the iPad 1. And that's not including any multi-threaded capabilities programmed into Safari which are no doubt there.

 

[kultschar]

You kidding? The iPad 2 is getting 60% more performance in an environment that isn't multi-threaded -- I.E, getting 60% more performance from one core than the iPad 1 was.

The CPU will be an ARM Cortex A9 because it offers more punch and less power consumption at the same clock speed, thus having an advantage over the A8 in the iPad. It certainly won't make up all of that additional 60% performance in that SunSpider benchmark, the rest would be from that same core not having to handle system/background tasks as well unlike in the iPad 1. And that's not including any multi-threaded capabilities programmed into Safari which are no doubt there.

Hmm - do you think iPad 2 will have seperate firmware then the iPad 1 to take advantage of dual core / multithreading??

 

[BradClarke]

Wow I didnt realize that a devices speed was solely based on Javascript! Man do i feel dumb...

So it's only "1.5x" faster running javascript there is obviously a very legitimate reason why the iPad 2 can run the new iMovie App, and the original iPad cant.

 

[Piggie]

What do you think?

CNET ran the SunSpider JavaScript test on both an iPad 2 and an original iPad running last week's gold master build of iOS 4.3. Though the new iPad has a dual core processor, it only ran the JavaScript tests 1.5 times faster than the first iPad on average, leading Ars to speculate that the A5 is not based around Cortex A9 cores but rather two of the A8 cores found in original iPad's A4. Furthermore, CNET found that the original iPad running the 4.3 build generally ran faster than one running 4.2, suggesting that users of iPads both old and new could see a significant speed increase when 4.3 rolls out.

 

[fertilized-egg]

What do you think?
it only ran the JavaScript tests 1.5 times faster than the first iPad on average, leading Ars to speculate that the A5 is not based around Cortex A9 cores but rather two of the A8 cores found in original iPad's A4.

The comments in that Arstechnica article and postings in this thread, including mine, have been explaining endlessly why that was a ridiculous claim made by Ars. Some only pick and choose to read what fits their agenda, urg. Let me recap(again)

1) Tegra2 tests 1.5 times faster than the A8 using the same test. Is Tegra2 A8? (no)
2) Tegra2 and iPad2 test about the same. Is Tegra2 A8? (no)
3) Apple will have to make their own dualcore architecture if they were to use A8 dual core. It'll cost Apple more to make A8 dualcore rather than using A9 which has dualcore available readily. Why would they do that?

Hmm - do you think iPad 2 will have seperate firmware then the iPad 1 to take advantage of dual core / multithreading??

iOS had Grand Central Dispatch built-in since iOS4. It wasn't advertised big but it's been multicore ready for a while now. AFAIK, if you programmed multi threaded apps, it automatically takes advantage of dual core. (This is the kind of control over the platform improvement Jobs talked about earlier.)

 

[MythicFrost]

Hmm - do you think iPad 2 will have seperate firmware then the iPad 1 to take advantage of dual core / multithreading??

I don't really know. But you can already write applications with multiple threads... these should get much benefit from multi-core devices, etc.

 

[Scarpad]

That's why I'm waiting till a 3rd gen, or a smaller device or 128gb size increase, not enough of an update

 

[JulianL]

What do you think?

CNET ran the SunSpider JavaScript test on both an iPad 2 and an original iPad running last week's gold master build of iOS 4.3. Though the new iPad has a dual core processor, it only ran the JavaScript tests 1.5 times faster than the first iPad on average, leading Ars to speculate that the A5 is not based around Cortex A9 cores but rather two of the A8 cores found in original iPad's A4. Furthermore, CNET found that the original iPad running the 4.3 build generally ran faster than one running 4.2, suggesting that users of iPads both old and new could see a significant speed increase when 4.3 rolls out.

My opinion of Ars just plunged. As others have already pointed out, the work involved to make a multicore version of the Cortex-A8 is not insignificant so why on earth would Apple reinvent an inferior wheel when ARM have already done it for the significantly better Cortex-A9. It makes no sense whatsoever and I would be absolutely amazed if it (the bit in red) is true.

I think that MythicFrost's post about what is happening with the 60% increase is absolutely correct, I agree with every word of his explanation/opinion.

- Julian

 

[fertilized-egg]

That's why I'm waiting till a 3rd gen, or a smaller device or 128gb size increase, not enough of an update

*Facepalm..more*

That's a bummer because web browsing on the original iPad could be a tad faster.

Time to start a thread "iPad 2 not really faster" (because it doesn't look faster from a certain angle) vincenz.

 

[asleep]

Not going to noticeably increase browsing speed, correct.

 

[FLYING SE7EN]

Just wanted to try to ask this again. Is the nitro javascript part of iOS 4.3 (therefore it can be used by other 3rd party browsers) or is it part of Mobile Safari itself?

 

[fertilized-egg]

Just wanted to try to ask this again. Is the nitro javascript part of iOS 4.3 (therefore it can be used by other 3rd party browsers) or is it part of Mobile Safari itself?

Part of Mobile Safari. But all 3rd party browsers use the Safari rending engine and its Javascript engine so they all should benefit. edit: doh, it seems I'm could very well be wrong on this one!

 

[foodog]

I was comparing two devices, running the same version of an operating system (iOS 4.3), and one assumes the same code to run Java. When such a comparison is made, and the variable is the hardware running said software, the so-called iPad 2 (or iPad 1.5) is not much faster than the original iPad.

I’m guessing you are not a scientist. If you were, I’m sure you would see the value of keeping variables in a comparison at a minimum. This is basic science, KEEP THE NUMBER OF VARIABLES TO A MINIUM. Try not to add too many exogenous factors that will lead to incorrect conclusions. Do you not agree? When you bring in different OS’es and code bases, you are adding too many variables to be informational. Who is to say Goolge will not offer a refined version of their software tomorrow, making your position moot.

Now, if they were running the same code, sure it would be a good point. However, they are not.

In fact, I hear Goolge is releasing an updated version of Honeycobm tommorrow.

 

[gloss]

Part of Mobile Safari. But all 3rd party browsers use the Safari rending engine and its Javascript engine so they all should benefit. edit: doh, it seems I'm could very well be wrong on this one!

This is true, but I'm guessing 3rd party browsers will need to recompile or otherwise make some tweaks to take advantage of the new engine. I doubt it will be long before updates start popping up.

 

[JulianL]

Part of Mobile Safari. But all 3rd party browsers use the Safari rending engine and its Javascript engine so they all should benefit. edit: doh, it seems I'm could very well be wrong on this one!

I'm afraid that for now I think you are wrong. I just got some info from the iCab developer that I posted here: https://forums.macrumors.com/showthread.php?p=12110332

- Julian