Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

cynics

macrumors G4
Jan 8, 2012
11,959
2,156
I think a great innovation would be LCD quality with amoled power consumption. I think LCD has seen its limit, hopefully Samsung will perfect it.

Too my eyes it's really close as is but some will complain until specs and instruments say otherwise.
 

thekev

macrumors 604
Aug 5, 2010
7,005
3,343
should be excessive? I think it isexcessive. RAM is useful for speeding up the use of applications off a hard drive, but in this case (on both the iPhone and android phones, whatever the model) its coming off flash memory, so it doesnt need to be as much as all that.

The net speeds of the NAND storage on your iphone are much slower than say a macbook air or even an HDD based computer.
 

notkodysmith

macrumors newbie
Jun 20, 2012
20
0
The net speeds of the NAND storage on your iphone are much slower than say a macbook air or even an HDD based computer.

I understand that computers are faster then phones, that's not the point i'm making. I am instead saying that flash memory is faster than hdd, and that's the reason that ram is flash memory, is to increase the speed at wich you can open and switch between apps. So it just seems kind of redundant for there to be a high ammount of. flash-based ram paired with a flash based memory unit.

wouldnt it be crazy if the iPhone 5 has a SSD? that would be crazy.
 

thekev

macrumors 604
Aug 5, 2010
7,005
3,343
I understand that computers are faster then phones, that's not the point i'm making. I am instead saying that flash memory is faster than hdd, and that's the reason that ram is flash memory, is to increase the speed at wich you can open and switch between apps. So it just seems kind of redundant for there to be a high ammount of. flash-based ram paired with a flash based memory unit.

wouldnt it be crazy if the iPhone 5 has a SSD? that would be crazy.

I'll try to explain again. Do not take this the wrong way, but you're completely unaware of how memory systems work. The OS requires a certain amount of memory address space to be reserved for itself. It reserves this space in volatile random access memory, which is quite fast. In notebooks and desktops it can also reserve more space in the form of virtual memory, although the way this is used is a bit different. Regardless of the presence of disk space used as virtual memory, it requires a minimum amount of contiguous address space in ram. Actually it goes further in that some values are simultaneously stored to the cpu cache. It accesses the values from ram on a cache miss. Both of these things are much faster than reading from disks, including ssds.

In the case of the iphone/ipad, they use very slow storage that is really designed more for energy efficiency than performance. Unless I'm grossly mistaken, iOS doesn't even have a system for virtual memory, so whatever is loaded must be addressable within available ram. Even if it did have the ability to cache memory values within storage space, this would result in very slow behavior on these devices as the storage performance isn't there. Not all solid state storage is the same in terms of performance, so it's not overkill having this. It's also necessary as there's no suitable method of turning storage space into addressable memory. Once again the volatile random access memory is not the same thing as NAND in terms of performance or usability. The one thing I forgot to mention earlier in the explanation is that NAND has a limited number of write cycles before it goes bad. SSDs have a number of replacement cells on reserve to compensate for this, but it would be totally inappropriate for something like ram.

I hope this helps:p.
 

notkodysmith

macrumors newbie
Jun 20, 2012
20
0
thank you!!!

Thank you very much for explaining this to me. I am glad that someone from the community was able to explain this to me, because i dont know very much about computer internals. i knew that ram increased performance and up graded my mac by myeslf, but thank you again for correcting me, without putting me down!
 

KnightWRX

macrumors Pentium
Jan 28, 2009
15,046
4
Quebec, Canada
I'll try to explain again. Do not take this the wrong way, but you're completely unaware of how memory systems work. The OS requires a certain amount of memory address space to be reserved for itself. It reserves this space in volatile random access memory, which is quite fast. In notebooks and desktops it can also reserve more space in the form of virtual memory, although the way this is used is a bit different. Regardless of the presence of disk space used as virtual memory, it requires a minimum amount of contiguous address space in ram. Actually it goes further in that some values are simultaneously stored to the cpu cache. It accesses the values from ram on a cache miss. Both of these things are much faster than reading from disks, including ssds.

Actually, Virtual memory is simply the memory space your process sees. The OS kernel translates the virtual memory addresses your program reads/writes to into page addresses. A memory page can reside in any kind of storage medium, be it random access memory or in what is called a swap area. It can also be CPU registers or CPU caches or other hardware registers. It can be in what is called shared memory or in your process' own pool of ressources.

If the memory page is in the swap area, the OS then tries to find dirty pages in random access memory to swap out to swap area so it can swap your page back in.

Virtual memory was erroneously used by Microsoft to indicate a swap area. They have since corrected that mislabelling by calling the swap area the "Paging file".
 

thekev

macrumors 604
Aug 5, 2010
7,005
3,343
Actually, Virtual memory is simply the memory space your process sees. The OS kernel translates the virtual memory addresses your program reads/writes to into page addresses. A memory page can reside in any kind of storage medium, be it random access memory or in what is called a swap area. It can also be CPU registers or CPU caches or other hardware registers. It can be in what is called shared memory or in your process' own pool of ressources.

If the memory page is in the swap area, the OS then tries to find dirty pages in random access memory to swap out to swap area so it can swap your page back in.

Virtual memory was erroneously used by Microsoft to indicate a swap area. They have since corrected that mislabelling by calling the swap area the "Paging file".

You always explain things so much better (and more accurately) than me:cool:. My explanation wasn't that great when i re-read it. I just wanted to indicate that none of the prior statements were correct and that you can't attempt to use iphone NAND storage as a means of ram. There are some things that require a certain amount of real contiguous ram unless I've somehow been grossly misled by many many pages of various API documentation.
 

KnightWRX

macrumors Pentium
Jan 28, 2009
15,046
4
Quebec, Canada
You always explain things so much better (and more accurately) than me:cool:. My explanation wasn't that great when i re-read it. I just wanted to indicate that none of the prior statements were correct and that you can't attempt to use iphone NAND storage as a means of ram. There are some things that require a certain amount of real contiguous ram unless I've somehow been grossly misled by many many pages of various API documentation.

Your explanation was top notch, just wanted to correct the terminology about virtual memory since every modern OS today uses that old Unix concept (one of the reason Unix required a MMU back in the 80s and couldn't be ported to x86 until they added such a feature, circa the 286).

You're right about NAND used in storage being way too removed from the CPU (too far on the bus) and way too slow in actual performance compared to even older DRAM models.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.