Please demonstrate that the fps is limited due to lack of RAM, because I'm not seeing it in the iPhone 6 graphic benchmarks. The iPhone 6+ is pushing more pixels than the 6, which is a limitation of using the same graphic cores of the same processor.
Memory bandwidth impacts every SoC style APU made. This is why Intel designed in the in package cache chip.
Same RAM = unified platform (ala optical camera bump)?
So bump both machines to 2GB.
However, IMO, this RAM issue keeps coming up for a few basic reasons. For my use case (and as I anecdotally infer from many other comments) when I have more than a few tabs open, it constantly re-loads. The question is - is limited RAM really the culprit or it just the way Safari is design to behave no matter the size of RAM? Some might like it reload every time they hit the tab, not me, but others might.
The answer is yes to all of the above! Web sites can mark themselves as volatile and as such web browsers are suppose to reload those pages every time they reference them. In a case like this Safari does what it is suppose to do.
In a lot of other cases Safari is reloading apparently due to the lack of RAM.
Unless there is a ton of pics etc, what is the average web page size of say apple insider, or google news, yahoo news etc? I have no idea or know where to look, anyone, Buehler?
Web sites can become extremely complex so using the word average here is difficult at best. I suppose you could save a web page to get a rough idea of its size as text. The problem is sites will download resources from all over, you have CSS, JavaScript, pics as you note, sound tracks, ads (way to many of them) all of which require the in memory creation of data structures and the like.
In the end it is hard to tell. Then you have the reality that Safari has never been good at memory management. Over the last year they have done a huge amount of work modernizing Safari as a C++ code base. So memory management is much improved, I can see this on my iPad 3 running the iOS betas and now the GM. It still isn't perfect but Safari is snappier. ????????????????????
Yes we really do, if you don't want to use the feature then don't, simple,
Sure as long as it doesn't create UI complexity that we don't need.
I have absolutely zero doubt that you'll find a lot more who want the ability to view two apps at a time, especially on a screen so large then those who don't.
On a cell phone? Honestly I'm not sure why anyone would want to carry around one of these massive cell phones constantly. I won't, when I need a big screen I can bring along an iPad. Even on the iPad the few times that I wanted two "apps" on screen at the same time was when I needed to reference a calculator and frankly I'd rather see drop down apps that drop down like the Today view. Generally I'm happy with sweeping through apps to get stuff done.
On the iPad I'm a bit more open to additional functionality. But even there the big deal is the lack of RAM in current models.
Especially that it means Apple will not only increase the memory but finally allow more apps to be ran at the same time. I can't believe you would assume that the majority wouldn't want this.
Actually I'm on the side that thinks most users wouldn't want or need this in a cell phone. It will be a bit different for tablets though. Even in a tablet I don't see running apps side by side as the right solution, it just messes up one of the best features of iPad which is that the app gets all of the screen.
Safari performance seems much improved on my iPad Air with iOS 8.00
SafarI is much better in iOS 8. However it still runs into memory issues. There was a huge amount of work put into WebKit over the last year as such I would suggest reading the WebKit blog to get a flavor for the effort expended. Lots of improvements but also more than a few bugs. All in all though a big win for the user.
Sure as long as it doesn't create UI complexity that we don't need.
On a cell phone? Honestly I'm not sure why anyone would want to carry around one of these massive cell phones constantly. I won't, when I need a big screen I can bring along an iPad. Even on the iPad the few times that I wanted two "apps" on screen at the same time was when I needed to reference a calculator and frankly I'd rather see drop down apps that drop down like the Today view. Generally I'm happy with sweeping through apps to get stuff done.
On the iPad I'm a bit more open to additional functionality. But even there the big deal is the lack of RAM in current models.
Actually I'm on the side that thinks most users wouldn't want or need this in a cell phone. It will be a bit different for tablets though. Even in a tablet I don't see running apps side by side as the right solution, it just messes up one of the best features of iPad which is that the app gets all of the screen.
I was talking about tablets, who cares about side by side in a phone.
Memory bandwidth impacts every SoC style APU made. This is why Intel designed in the in package cache chip.
You are correct, and if I was speaking of memory bandwidth, an appropriate comment.
I was speaking of System RAM, i.e. 1 GB vs 2 GB, not cache memory, which is actually quite large in the A8 anyway, though not on the scale of a server or desktop processor.
I was talking about tablets, who cares about side by side in a phone.
Don't understand that request myself either: people want a larger display on their phone and then making that display even smaller than 'old school' 4" because of two apps side-by-side. Makes no sense - at all.
I would worry that an A8X on an iPad that big would give the performance of the iPad 3. Not good. Unless, of course, it is REALLY beefy! As for the name of the thing, no doubt now it will be the iPad Plus, not Pro.
Will there now be a MacPro Plus, Mac mini Plus, MacBook Plus, MacBook Air Plus, iMac Plus?
I fear they may call it the iPad Plus, but I wish they wouldn't. That moniker has become indelibly stained.
Wow in the 1.2 Unlimited - Physics department! Bottom 5 phones has 3 from Apple. No idea what this 'physics' relates to, but this one stands out.
Physics calculations are run on the CPU, the iPhone has a dual-core, the others tend to have quad-cores. If you have a racing game, your virtual car has to have speed, mass, collision calculations, direction vectors, rebound force, roll calculations and so on. The iPhone 6 CPU holds up pretty well in Geekbench though so it must be better at some kinds of calculations than others. There's an interesting article on the 3DMark site about this:
When they tested the 5S against the 5, they found no improvement in physics despite the chip being faster. Physics calculations have dependencies for future calculations that can slow it down. When they adjusted the code, they got a 17% speedup, making the 5S faster than the 5.
Geekbench tests are for raw performance without these dependencies and show a more accurate profile of the CPU's raw performance without software-specific bottlenecks. The iPhone 6 comes in the top 4 of all mobile devices tested for CPU performance.
If the physics calculations don't bottleneck the CPU, the graphics tests would be more important. When they aggregate the scores, they are just combining the score results and not running them together so the overall score isn't really an accurate representation of the performance. It's like scoring 9/10 in bed and 2/10 in cooking and then aggregating the score to 5.5/10. Nobody really cares how good you are at cooking, you can eat out or get a takeaway. The iPhone is good in bed and that's what counts. Plus, as we know, a game looking for the best performance would be developed using Metal. I don't know if it would be fair to optimize 3DMark in Metal though unless they do an equivalent optimization for competing devices. It depends on what's being measured really.
Physics calculations are run on the CPU, the iPhone has a dual-core, the others tend to have quad-cores. If you have a racing game, your virtual car has to have speed, mass, collision calculations, direction vectors, rebound force, roll calculations and so on. The iPhone 6 CPU holds up pretty well in Geekbench though so it must be better at some kinds of calculations than others. There's an interesting article on the 3DMark site about this:
When they tested the 5S against the 5, they found no improvement in physics despite the chip being faster. Physics calculations have dependencies for future calculations that can slow it down. When they adjusted the code, they got a 17% speedup, making the 5S faster than the 5.
Geekbench tests are for raw performance without these dependencies and show a more accurate profile of the CPU's raw performance without software-specific bottlenecks. The iPhone 6 comes in the top 4 of all mobile devices tested for CPU performance.
If the physics calculations don't bottleneck the CPU, the graphics tests would be more important. When they aggregate the scores, they are just combining the score results and not running them together so the overall score isn't really an accurate representation of the performance. It's like scoring 9/10 in bed and 2/10 in cooking and then aggregating the score to 5.5/10. Nobody really cares how good you are at cooking, you can eat out or get a takeaway. The iPhone is good in bed and that's what counts. Plus, as we know, a game looking for the best performance would be developed using Metal. I don't know if it would be fair to optimize 3DMark in Metal though unless they do an equivalent optimization for competing devices. It depends on what's being measured really.
That's a very informative post Marvin. And I haven't even begun with the article from the link, so thanks.
The rating on love making and cooking is a good one!
Physics calculations are run on the CPU, the iPhone has a dual-core, the others tend to have quad-cores. If you have a racing game, your virtual car has to have speed, mass, collision calculations, direction vectors, rebound force, roll calculations and so on. The iPhone 6 CPU holds up pretty well in Geekbench though so it must be better at some kinds of calculations than others. There's an interesting article on the 3DMark site about this:
When they tested the 5S against the 5, they found no improvement in physics despite the chip being faster. Physics calculations have dependencies for future calculations that can slow it down. When they adjusted the code, they got a 17% speedup, making the 5S faster than the 5.
Geekbench tests are for raw performance without these dependencies and show a more accurate profile of the CPU's raw performance without software-specific bottlenecks. The iPhone 6 comes in the top 4 of all mobile devices tested for CPU performance.
If the physics calculations don't bottleneck the CPU, the graphics tests would be more important. When they aggregate the scores, they are just combining the score results and not running them together so the overall score isn't really an accurate representation of the performance. It's like scoring 9/10 in bed and 2/10 in cooking and then aggregating the score to 5.5/10. Nobody really cares how good you are at cooking, you can eat out or get a takeaway. The iPhone is good in bed and that's what counts. Plus, as we know, a game looking for the best performance would be developed using Metal. I don't know if it would be fair to optimize 3DMark in Metal though unless they do an equivalent optimization for competing devices. It depends on what's being measured really.
That's a very informative post Marvin. And I haven't even begun with the article from the link, so thanks.
The rating on love making and cooking is a good one!
I disagree. If you've had good home cooking, you'd realise how important it is.
Physics calculations are run on the CPU, the iPhone has a dual-core, the others tend to have quad-cores. If you have a racing game, your virtual car has to have speed, mass, collision calculations, direction vectors, rebound force, roll calculations and so on. The iPhone 6 CPU holds up pretty well in Geekbench though so it must be better at some kinds of calculations than others. There's an interesting article on the 3DMark site about this:
When they tested the 5S against the 5, they found no improvement in physics despite the chip being faster. Physics calculations have dependencies for future calculations that can slow it down. When they adjusted the code, they got a 17% speedup, making the 5S faster than the 5.
Geekbench tests are for raw performance without these dependencies and show a more accurate profile of the CPU's raw performance without software-specific bottlenecks. The iPhone 6 comes in the top 4 of all mobile devices tested for CPU performance.
If the physics calculations don't bottleneck the CPU, the graphics tests would be more important. When they aggregate the scores, they are just combining the score results and not running them together so the overall score isn't really an accurate representation of the performance. It's like scoring 9/10 in bed and 2/10 in cooking and then aggregating the score to 5.5/10. Nobody really cares how good you are at cooking, you can eat out or get a takeaway. The iPhone is good in bed and that's what counts. Plus, as we know, a game looking for the best performance would be developed using Metal. I don't know if it would be fair to optimize 3DMark in Metal though unless they do an equivalent optimization for competing devices. It depends on what's being measured really.
Hmmm... I don't believe in compromising, plus cooking at home can lead to all kinds of fun outside of the bed :wow:
Considering your analogy to current popular "events"... I would say that the iPhone has been proven to be a very good partner in bed, regardless of CPU, GPU, or Metal speeds
Hmmm... I don't believe in compromising, plus cooking at home can lead to all kinds of fun outside of the bed :wow:
Considering your analogy to current popular "events"... I would say that the iPhone has been proven to be a very good partner in bed, regardless of CPU, GPU, or Metal speeds
Pour some sugar on me
Ooh, in the name of love
Pour some sugar on me
C'mon, fire me up
Pour your sugar on me
Oh, I can't get enough
I'm hot, sticky sweet
From my head to my feet, yeah
Comments
Memory bandwidth impacts every SoC style APU made. This is why Intel designed in the in package cache chip.
In a lot of other cases Safari is reloading apparently due to the lack of RAM.
Web sites can become extremely complex so using the word average here is difficult at best. I suppose you could save a web page to get a rough idea of its size as text. The problem is sites will download resources from all over, you have CSS, JavaScript, pics as you note, sound tracks, ads (way to many of them) all of which require the in memory creation of data structures and the like.
In the end it is hard to tell. Then you have the reality that Safari has never been good at memory management. Over the last year they have done a huge amount of work modernizing Safari as a C++ code base. So memory management is much improved, I can see this on my iPad 3 running the iOS betas and now the GM. It still isn't perfect but Safari is snappier. ????????????????????
On the iPad I'm a bit more open to additional functionality. But even there the big deal is the lack of RAM in current models. Actually I'm on the side that thinks most users wouldn't want or need this in a cell phone. It will be a bit different for tablets though. Even in a tablet I don't see running apps side by side as the right solution, it just messes up one of the best features of iPad which is that the app gets all of the screen.
SafarI is much better in iOS 8. However it still runs into memory issues. There was a huge amount of work put into WebKit over the last year as such I would suggest reading the WebKit blog to get a flavor for the effort expended. Lots of improvements but also more than a few bugs. All in all though a big win for the user.
Sure as long as it doesn't create UI complexity that we don't need.
On a cell phone? Honestly I'm not sure why anyone would want to carry around one of these massive cell phones constantly. I won't, when I need a big screen I can bring along an iPad. Even on the iPad the few times that I wanted two "apps" on screen at the same time was when I needed to reference a calculator and frankly I'd rather see drop down apps that drop down like the Today view. Generally I'm happy with sweeping through apps to get stuff done.
On the iPad I'm a bit more open to additional functionality. But even there the big deal is the lack of RAM in current models.
Actually I'm on the side that thinks most users wouldn't want or need this in a cell phone. It will be a bit different for tablets though. Even in a tablet I don't see running apps side by side as the right solution, it just messes up one of the best features of iPad which is that the app gets all of the screen.
I was talking about tablets, who cares about side by side in a phone.
Memory bandwidth impacts every SoC style APU made. This is why Intel designed in the in package cache chip.
You are correct, and if I was speaking of memory bandwidth, an appropriate comment.
I was speaking of System RAM, i.e. 1 GB vs 2 GB, not cache memory, which is actually quite large in the A8 anyway, though not on the scale of a server or desktop processor.
Don't understand that request myself either: people want a larger display on their phone and then making that display even smaller than 'old school' 4" because of two apps side-by-side. Makes no sense - at all.
Does it bend?
Will it bend (blend)?
I'm kind of leaning towards thinking 'iPad Plus', since that would follow iPhone 6 Plus' precedent.
I would love a larger iPad, but if it's an updated Air, I'd be very happy with that, too.
Please don't cock this launch up, Apple.
I fear they may call it the iPad Plus, but I wish they wouldn't. That moniker has become indelibly stained.
I don't especially care about the size. It's the weight that's important.
Physics calculations are run on the CPU, the iPhone has a dual-core, the others tend to have quad-cores. If you have a racing game, your virtual car has to have speed, mass, collision calculations, direction vectors, rebound force, roll calculations and so on. The iPhone 6 CPU holds up pretty well in Geekbench though so it must be better at some kinds of calculations than others. There's an interesting article on the 3DMark site about this:
http://www.futuremark.com/pressreleases/understanding-3dmark-results-from-the-apple-iphone-5s-and-ipad-air
When they tested the 5S against the 5, they found no improvement in physics despite the chip being faster. Physics calculations have dependencies for future calculations that can slow it down. When they adjusted the code, they got a 17% speedup, making the 5S faster than the 5.
Geekbench tests are for raw performance without these dependencies and show a more accurate profile of the CPU's raw performance without software-specific bottlenecks. The iPhone 6 comes in the top 4 of all mobile devices tested for CPU performance.
If the physics calculations don't bottleneck the CPU, the graphics tests would be more important. When they aggregate the scores, they are just combining the score results and not running them together so the overall score isn't really an accurate representation of the performance. It's like scoring 9/10 in bed and 2/10 in cooking and then aggregating the score to 5.5/10. Nobody really cares how good you are at cooking, you can eat out or get a takeaway. The iPhone is good in bed and that's what counts. Plus, as we know, a game looking for the best performance would be developed using Metal. I don't know if it would be fair to optimize 3DMark in Metal though unless they do an equivalent optimization for competing devices. It depends on what's being measured really.
I imagine the first-gen iPad Pro will be 1.5 lbs, like the original first-gen.
That's a very informative post Marvin. And I haven't even begun with the article from the link, so thanks.
The rating on love making and cooking is a good one!
That would be more than acceptable to me.
I disagree. If you've had good home cooking, you'd realise how important it is.
Hmmm... I don't believe in compromising, plus cooking at home can lead to all kinds of fun outside of the bed :wow:
Considering your analogy to current popular "events"... I would say that the iPhone has been proven to be a very good partner in bed, regardless of CPU, GPU, or Metal speeds
Pour some sugar on me
Ooh, in the name of love
Pour some sugar on me
C'mon, fire me up
Pour your sugar on me
Oh, I can't get enough
I'm hot, sticky sweet
From my head to my feet, yeah
Wooooohoooo....!!!! Coyote callin'.....:smokey:
[VIDEO]