Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Thanks, this is very interesting! So even if the write is on average slower, it could potentially do that in the background as it dumps the memory, and then read it back in quickly later? Seems like it should be doable considering laptops can do this and the iPhone is getting pretty close to the lower-end of that performance, but I don't know how robust the NAND is and how much overhead that would cost. My guess is that it would require a significant re-write of the way iOS handles everything. That and it would probably consume more power.
The disks we've been used to seeing on recent Macbooks are at least an order of magnitude faster than the storage we have in the iPhones, so that is one reason the laptops can get get away with using them. Power utilization becomes a bigger concern on phones as well, because if you implement writing to disk for more tasks, you are keeping the CPU from going idle more often that it would otherwise be able to when operating with memory instead - that memory can accomplish the same task faster allows the CPU to sleep earlier (provides savings in power, keeping temperatures closer to ideal levels, etc.). Here is a crude scenario (e.g. off the top of my head, non-optimized, etc.) - implementing disk caching on browsers causes situations such as when I was attempting to transfer Xcode to my phone; depending on how many objects are being created so that a particular site can be rendered, you are potentially running disk I/O for long periods of time, which can be detrimental to battery life. You can also run the risk of queuing sufficiently large amounts of disk I/O operations to the point of causing processes to stall (again, due to the low speeds of disk I/O).

Writing to disk would not require a re-write to iOS, as APIs exist for those already (e.g. a game app stores its save data), but enabling swap files would, unless I'm mistaken. Keep in mind that everything I've been discussing so far is purely theoretical based on my computer architecture knowledge (I work with software, not hardware), so there may exist practical methods of implementing disk-based caching functionalities without hamstringing the phone's overall experience.
 
I was thinking a complete re-write would be necessary to mitigate some of the issues you mentioned, and would prioritize the writing of the swap file over everything else in the stack. I have a limited understanding as well, but would assume that iOS is written completely different than Mac OS in this regard. But years ago we also had swap files on old spinning drives which are likely much slower than the NAND in the iPhone, especially for writing many smaller random files to disk.

The disks we've been used to seeing on recent Macbooks are at least an order of magnitude faster than the storage we have in the iPhones, so that is one reason the laptops can get get away with using them. Power utilization becomes a bigger concern on phones as well, because if you implement writing to disk for more tasks, you are keeping the CPU from going idle more often that it would otherwise be able to when operating with memory instead - that memory can accomplish the same task faster allows the CPU to sleep earlier (provides savings in power, keeping temperatures closer to ideal levels, etc.). Here is a crude scenario (e.g. off the top of my head, non-optimized, etc.) - implementing disk caching on browsers causes situations such as when I was attempting to transfer Xcode to my phone; depending on how many objects are being created so that a particular site can be rendered, you are potentially running disk I/O for long periods of time, which can be detrimental to battery life. You can also run the risk of queuing sufficiently large amounts of disk I/O operations to the point of causing processes to stall (again, due to the low speeds of disk I/O).

Writing to disk would not require a re-write to iOS, as APIs exist for those already (e.g. a game app stores its save data), but enabling swap files would, unless I'm mistaken. Keep in mind that everything I've been discussing so far is purely theoretical based on my computer architecture knowledge (I work with software, not hardware), so there may exist practical methods of implementing disk-based caching functionalities without hamstringing the phone's overall experience.
 
I was thinking a complete re-write would be necessary to mitigate some of the issues you mentioned, and would prioritize the writing of the swap file over everything else in the stack. I have a limited understanding as well, but would assume that iOS is written completely different than Mac OS in this regard. But years ago we also had swap files on old spinning drives which are likely much slower than the NAND in the iPhone, especially for writing many smaller random files to disk.
Yes, iOS will definitely need to receive some optimizations if something similar to this were to be implemented. I'm rather skeptical as to how much we can truly "optimize away", however. And true, HDDs were much slower back then, which was why transfer speeds are also dramatically slower than what we have today. NAND types of disks mitigates this to an extent, so that's a welcome change.
 
Instead of focusing on a number, shouldn't we be more concerned about how Safari itself is coded? If people are saying that they same issues aren't the same on other mobile browsers installed, isn't the real culprit the app itself, and not the RAM size? Everyone gets so infatuated on numbers...
 
There are a hell of a lot of apologists here.
1GB of RAM on a late 2014 premium smartphone is PATHETIC, whether it reloads tabs or not. The fact that we've been experiencing tab reloads for *years* and that the problem has become MUCH worse since iOS7 seems to point to software becoming too powerful for the hardware. Now all these apologists bring out the techno diatribe as to why 1GB of RAM is perfect, but it obviously isn't. My Android tablet with 1.5GB of RAM doesn't reload tabs. Current premium Android smartphones have 3GB of RAM and they have the best battery life of any smartphone out there, slaying the i6 battery life comfortably. So if RAM drains batteries so much to make adding more unfeasible, how do Samsung et al manage to not only stay in the same ballpark as Apple when it comes to battery life, but to actually come out victorious?

Wow, way to go. There were 10 millions iphones sold in 3 days. That is a heck of a lot of "apologists" as you put it. Nobody cares, got it?

They couldn't keep the phone in stock. Nobody except the 10 people on this thread could care less about how much memory or battery life the iphone 6 has. Sorry to be the messenger of the news. The word obviously didn't get out that the iphone 6 battery life stinks and the memory is sub-par.

This is a big opportunity for Samsung to sell 10 million phones, since they can best the iphone 6 in every single hardware benchmark, at their next phone launch.
 
Ok, but keep in mind that android acrually does stuff, like things that come stock, so you really can't compare an OS that doesn't do crap on low specs, against a beast on beast specs
 
Why is that benchmark testing a 2013 A7 vs a 2012 S4 Pro?
Is it cause they knew that the 2013 SD800 would have spanked it to today's standards?

apple A7 also beats the 800/801 Snapdragon's, see Anandtechs article on that and alot of benchmarks she that too

Nvidia's own benchmark aswell

----------

Wow, way to go. There were 10 millions iphones sold in 3 days. That is a heck of a lot of "apologists" as you put it. Nobody cares, got it?

They couldn't keep the phone in stock. Nobody except the 10 people on this thread could care less about how much memory or battery life the iphone 6 has. Sorry to be the messenger of the news. The word obviously didn't get out that the iphone 6 battery life stinks and the memory is sub-par.

This is a big opportunity for Samsung to sell 10 million phones, since they can best the iphone 6 in every single hardware benchmark, at their next phone launch.

Lol Sunking is beyond wrong, Anandtech's test on performance and battery life shows the new iPhones coming out on top on performance and battery life

Also DisplayMate just named the i6/i6 plus best smartphone LCD displays ever
 
Instead of focusing on a number, shouldn't we be more concerned about how Safari itself is coded? If people are saying that they same issues aren't the same on other mobile browsers installed, isn't the real culprit the app itself, and not the RAM size? Everyone gets so infatuated on numbers...

Numbers are an indication of performance I'm not sure what's so hard to understand about that. They aren't the only factor but they are a big one, you can only optimize so much.

While android fanboys are blinded by thinking only specs matter you and other apple fanboys are blinded by thinking they don't matter at all.
 
Numbers are an indication of performance I'm not sure what's so hard to understand about that. They aren't the only factor but they are a big one, you can only optimize so much.

While android fanboys are blinded by thinking only specs matter you and other apple fanboys are blinded by thinking they don't matter at all.

Cool assumption buddy. All I was trying to say, is that the problem really isn't the amount of RAM, but the coding of the programs that have issues themselves. :rolleyes:
 
Cool assumption buddy. All I was trying to say, is that the problem really isn't the amount of RAM, but the coding of the programs that have issues themselves. :rolleyes:

There's nothing wrong with the coding. If there is explain how.

Safari stores web page data in ram. Ram gets full of web page data quickly because you have so little, safari dumbs web page data to free RAM.

That's how it works, explain to me the coding problem?
 
There's nothing wrong with the coding. If there is explain how.

Safari stores web page data in ram. Ram gets full of web page data quickly because you have so little, safari dumbs web page data to free RAM.

That's how it works, explain to me the coding problem?

You are assuming Safari ONLY blanks out a tab when it needs RAM. There could be several other factors involved, beyond RAM usage, such as timeout time or something similar.
 
You are assuming Safari ONLY blanks out a tab when it needs RAM. There could be several other factors involved, beyond RAM usage, such as timeout time or something similar.

What do you mean timeout time? It's not even loading new information in the background its simply storing the tab in the state you left it no? So what do you mean by that?
 
As promised.





Thank you. This is exactly what I was looking for.

That's a slight apples to oranges comparison as that's touchwiz skin. But it does show that software can address the issue as that doesn't happen with vanilla android with less lag. Now the question is why isn't Apple just fixing it? If touchwiz does it too I wonder if there is a reason for that behavior. On my Xoom I can load so many tabs the browser crashes (30+) maybe that's the reason.

Something else if you could try because my access to Samsung devices is limited, see if you lose data you typed. So for example, type a post on MR and make the tab reload and see if you lose everything you typed like in iOS if you have a chance.

My 4S can't keep 2 tabs open most of the time and to add insult to injury it Verizon 3G so it's unbearably slow to reload. And app reloading in ios 8 on it is nuts. My merely going to safari (not actually doing anything in it) will kill tapatalk. Then I have to dive back through the sections to get back here.

So while more RAM isn't a cure all by comparing to iOS devices with less there is a huge difference in user experience. I believe the same would apply by having more then 1 gb just like it does having more than 512 mb.
 
There's nothing wrong with the coding. If there is explain how.

Safari stores web page data in ram. Ram gets full of web page data quickly because you have so little, safari dumbs web page data to free RAM.

That's how it works, explain to me the coding problem?
Are you really this stupid? How many webpages are you visiting that have 512mb of data on them? Which is implied if more than two tabs causes reloads. The reloads are not caused by lack of RAM. Get it out of your thick head. My MacBook pro has 16gb and tabs reload all the time. It gives a great benefit to battery life amongst other things.
 
Are you really this stupid? How many webpages are you visiting that have 512mb of data on them? Which is implied if more than two tabs causes reloads. The reloads are not caused by lack of RAM. Get it out of your thick head. My MacBook pro has 16gb and tabs reload all the time. It gives a great benefit to battery life amongst other things.

Seems like you are the stupid one how dare you say that when you apparently have no clue what you're talking about.

Do you think the RAM is completely free when you're using safari? You do realize the OS constantly takes up RAM right? Jesus you idiots just need to shutup if you don't have a damn clue what you're talking about.
 
Here is why the iPhone 6/6 Plus didn't get more RAM.

Apple has been slowly losing me the past couple of years due part of the direction of the phone line. It's a slap in my face that they are trying to spin this "only having 1 gb of ram as a good thing "nonsense. It gives them room to improve for the next phone by not putting more in now. I don't upgrade every year or even every other year. If your happy with what you got in the iPhone 6 then good for you.
 
Seems like you are the stupid one how dare you say that when you apparently have no clue what you're talking about.

Do you think the RAM is completely free when you're using safari? You do realize the OS constantly takes up RAM right? Jesus you idiots just need to shutup if you don't have a damn clue what you're talking about.

The point is, and I can see you need help being shown this, is that two webpages do not take up enough RAM to be causing the tab reloads. Whether or not the full 1GB is available or not is irrelevant. A good portion is available and even half of a "good portion" is more than enough for 1 webpage in RAM. You also forgot to acknowledge the fact that my Macbook Pro with 16gb RAM reloads tabs constantly.
 
The point is, and I can see you need help being shown this, is that two webpages do not take up enough RAM to be causing the tab reloads. Whether or not the full 1GB is available or not is irrelevant. A good portion is available and even half of a "good portion" is more than enough for 1 webpage in RAM. You also forgot to acknowledge the fact that my Macbook Pro with 16gb RAM reloads tabs constantly.


I don't care about your macbook pro or your anecdotes, I never anywhere mentioned that 2 tabs cause reloads either, on my 4s 2 tabs can, I can even get crashes, I don't know about the new phones, I don't recall if I got reloads on a 5s with 2 tabs, was probably a few more.

You don't have a clue what you're talking about now you're trying to back pedal. You didn't even know the OS took up RAM. It takes up a lot of RAM by the way, a whole lot.
 
I don't care about your macbook pro or your anecdotes, I never anywhere mentioned that 2 tabs cause reloads either, on my 4s 2 tabs can, I can even get crashes, I don't know about the new phones, I don't recall if I got reloads on a 5s with 2 tabs, was probably a few more.

You don't have a clue what you're talking about now you're trying to back pedal. You didn't even know the OS took up RAM. It takes up a lot of RAM by the way, a whole lot.

Of course I know the OS takes up RAM. I was making my explanation simple for simpletons like yourself. Obviously it is still too complex for you though, so I apologise.
 
Of course I know the OS takes up RAM. I was making my explanation simple for simpletons like yourself. Obviously it is still too complex for you though, so I apologise.

Damage control won't work. You acted the fool, and now it's probably best you remain silent rather then removing all doubt you are indeed a fool.

You keep insulting me and in fact started out the insults, you accuse me of being a fool when you are the one spouting blatantly false information.

Apple won't be paying you by the way so don't even bother to be their attack dog.
 
The point is, and I can see you need help being shown this, is that two webpages do not take up enough RAM to be causing the tab reloads. Whether or not the full 1GB is available or not is irrelevant. A good portion is available and even half of a "good portion" is more than enough for 1 webpage in RAM. You also forgot to acknowledge the fact that my Macbook Pro with 16gb RAM reloads tabs constantly.

My MBA with 4GB RAM can keep 15 tabs without reload. Maybe you have fake macbook? lol
 
Damage control won't work. You acted the fool, and now it's probably best you remain silent rather then removing all doubt you are indeed a fool.



You keep insulting me and in fact started out the insults, you accuse me of being a fool when you are the one spouting blatantly false information.



Apple won't be paying you by the way so don't even bother to be their attack dog.


You really do have a way about you don't you. :God:
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.