Apple's A4 Processor is Based on Cortex A8, Not A9?

[deconstruct60]

You can take a decent SSD and put it on very old computer hardware and still get snappy performance, as Runcore did with old Windows laptops

Snappy performance doing what? Running IE , solitaire, and an old copy of Word ? The launch times are snappy? Those micro, synthetic benchmarks ... the disk ones don't stress anything other than the disk (not memory or virtual memory ). So yeah it is faster, but that is a moot point unless that was the sole limiting factor.

In that on the iPhoneOS will quit/start apps to switch between that aspect is snappy. However, if have large memory and/or computational workload isn't as likely to help as much.

 

[holmesf]

This is quite a glob of jibberfish. I ran yes > /dev/null & to give it some utilization. So the CPU is near 100%, the battery is't exactly flatlining even after 20 minutes of this. And you'll notice it thinks it has 814M of VM.

I can have streaming radio playing in the media server (7%) while yes is running (81%) while playing Wurdle (4%) and sending an email. There is no problem multitasking.

It's an iPod touch 2g. (not an A8 I know.) I didn't want to risk my iPhone.

/k

Processes: 25 total, 2 running, 23 sleeping... 95 threads 17:51:33
Load Avg: 1.12, 1.09, 0.78 CPU usage: 94.74% user, 5.26% sys, 0.00% idle
SharedLibs: num = 0, resident = 0 code, 0 data, 0 linkedit.
MemRegions: num = 1796, resident = 29M + 0 private, 15M shared.
PhysMem: 28M wired, 11M active, 8068K inactive, 81M used, 35M free.
VM: 814M + 0 36215(0) pageins, 277(0) pageouts

PID COMMAND %CPU TIME #TH #PRTS #MREGS RPRVT RSHRD RSIZE VSIZE
294 yes 93.5% 1:53.00 1 14 25 184K 6536K 352K 8888K
295 top 6.1% 0:07.07 1 21 34 364K 6708K 844K 17M
272 bash 0.0% 0:00.14 1 15 49 252K 6720K 856K 19M
242 sshd 0.0% 0:01.22 1 15 34 300K 6492K 576K 18M
237 MobileSafa 0.0% 0:03.98 10 124 150 4352K 7644K 6084K 75M
170 MobileMusi 0.0% 0:01.69 3 105 84 2860K 7320K 3468K 69M
63 MobileMail 0.0% 1:06.37 5 146 107 2984K 7316K 3884K 69M
60 apsd 0.0% 0:08.23 2 46 36 464K 6492K 708K 17M
56 BTServer 0.0% 0:19.59 5 93 69 948K 6492K 1248K 31M
36 aosnotifyd 0.0% 0:14.28 3 96 66 1068K 6696K 1424K 41M
35 CommCenter 0.0% 0:00.95 4 78 58 676K 6696K 900K 42M
31 SpringBoar 0.0% 2:53.21 10 321 484 7716K 7648K 18M 86M
30 accessoryd 0.0% 0:02.01 1 53 29 336K 6484K 552K 17M
29 dataaccess 0.0% 0:05.85 6 121 60 1500K 6688K 1652K 42M
28 fairplayd 0.0% 0:00.78 1 34 38 384K 6480K 536K 18M
​

Not sure why that shows up, possibly they are memory mapped files. In any case, you're not right about this: the iPhone does not back memory with disk storage.

http://developer.apple.com/iphone/l...tual/ManagingMemory/Articles/AboutMemory.html

"In Mac OS X, the virtual memory system often writes pages to the backing store. The backing store is a disk-based repository containing a copy of the memory pages used by a given process. Moving data from physical memory to the backing store is called paging out (or “swapping out”); moving data from the backing store back in to physical memory is called paging in (or “swapping in”). In iPhone OS, there is no backing store and so pages are are never paged out to disk, but read-only pages are still be paged in from disk as needed."

Second, I never said the CPU had trouble multitasking. I said it was the OS memory system and battery life.

Third, 20 minutes? How about some more rigorous testing.

 

[cmaier]

Snappy performance doing what? Running IE , solitaire, and an old copy of Word ? The launch times are snappy? Those micro, synthetic benchmarks ... the disk ones don't stress anything other than the disk (not memory or virtual memory ). So yeah it is faster, but that is a moot point unless that was the sole limiting factor.

In that on the iPhoneOS will quit/start apps to switch between that aspect is snappy. However, if have large memory and/or computational workload isn't as likely to help as much.

SSDs do greatly improve the speed of virtual memory paging.

 

[HyperZboy]

Originally Posted by HyperZboy...

A) No FLASH support on a device that is supposed to be a computer, NOT a HUGE iPOD.
B) No multitasking. And this processor rumor just adds insult to injury there too, no doubt.
C) No Mac or PC software support & requires developers to create all new apps that are basically iPhone/iPOD Touch apps.

Link please where any ranking Apple official has ever claimed the iPad is "supposed to be a computer." If you can't provide one, which I'm pretty sure you cannot, then your supposition B & C is logically flawed as well.

In fact, Jobs said the iPad is intended to fit in between an iPod/Touch and a laptop. By that very definition it's not meant to be a computer in the lay use of the concept. The iPad is a new platform. Developers will support it or not based on whether they can make money doing so, just like any other platform.

Steve Jobs specifically said the device would be much better than a netbook, which is technically a cheap laptop computer that does the things I mentioned above while the iPAD does not. If you don't believe that is true, then don't bother reading any further, the Reality Distortion Field has already kicked into high gear! LOL

As an ebook reader and iPhone/iPOD app player, the iPAD rocks, but I already have those. I can't imagine carrying around a laptop or netbook, and an iPhone or iPOD + an iPAD.

This device doesn't replace any of them! LOL
This device doesn't even fully work with the internet like a netbook, let alone run real programs like a netbook. It's like the coolest device without a purpose I've ever seen.

A netbook runs a REAL computer operating system, NOT A PHONE OS and is potentially cheaper and can more likely replace a laptop for many things while traveling.

I imagine if you own none of the devices I mentioned above or have beaucoup dollars lying around, you might want this, even I'm intrigued & fascinated by it & would love to try one out, but I'm predicting this is a Steve Jobs failure of G4 Cube proportions, tons of hype, was incredibly cool, but nobody bought it.
The iPAD seems similarly & eerily incredibly cool, but I just can't imagine what I'd use it for at the price.

There was no Flash support showed in the Keynote video. Did you even watch it?


Maybe every website YOU visit requires Flash, that is not true for me. Around 1 in 10 I visit need Flash. Apple is being brave in taking on Adobe regarding Flash, face it, it's a bad technology and it needs someone with the guts of Steve Jobs to eradicate it.




Like it or not, you're caught in the 'Reality distortion field' that you are criticizing. You are either a hypocrite or a flamer. Which is it?

Even MacRumors reported that Apple posted pics on their website of the New York Times on an iPAD showing Flash Video.

Read back and then get back to me. Geez, the fanboys just refuse to believe the truth no matter what!
LOL

And btw, a large number of major newspapers, TV news stations ALL use FLASH video.

Just hating FLASH will not make it go away and I don't think Steve Jobs has the ability to make it go away even at this point, but it's clear he'd definitely like it to.

I'll be the 1st to admit I'm mad as hell at Adobe for making Flash so buggy on Macs.

But that still doesn't take away the limited nature of the iPAD because of the lack of it.
On a phone, it's ok really, I don't really mind, but on something that's supposed to better than a netbook... "Houston, we have a problem."

 

[admanimal]

I imagine if you own none of the devices I mentioned above or have beaucoup dollars lying around, you might want this, even I'm intrigued & fascinated by it & would love to try one out, but I'm predicting this is a Steve Jobs failure of G4 Cube proportions, tons of hype, was incredibly cool, but nobody bought it.
The iPAD seems similarly & eerily incredibly cool, but I just can't imagine what I'd use it for at the price.

I think your assessment as a whole is fair enough, but the difference between the iPad and G4 cube is that the cube really did nothing that any other Apple computer could do at the time (AFAIK...I don't think there was anything super special about its specs). In any case, it was basically just another Apple computer in a fancy case.

The iPad really is a different device that enables a new kind of computing and new kinds of apps. Of course people say with varying levels of sarcasm that it's just a big iPod Touch, but beyond the hardware that's really not true, as the iWork apps demonstrate as an example. For better or worse, Apple has hidden a lot of the stuff that you normally see in a desktop OS, while the larger screen allows for more seriously useful apps than its smaller cousins can handle. Only time will tell how the population at large reacts to this. I think its success will also be tied to whether developers can come up with some killer apps that really make people say I have to have one. Again this could never be the case with the Cube, since anything that you could do on a Cube you can do in the exact same way on any other Apple computer.

 

[kernkraft]

Long and winding road

What you forget is that the Netbook is trying to be a computer. The iPad is not.

Well, it would be pretty stupid to try when it is basically an oversized and over-clocked iPod Touch.

There's no way this thing has been under development since 2003.

Everything about it screams "rush job".

As an oversized iPod Touch, it could have been under development since 2003. The technology was already there then.

No. Once again: A8 has a 13 stages. A9 has 8. A8 has more pipe stages than A9.

PowerVR
SGX 535 = 2 ALUs (USSE1), 2 TMUs
SGX 540/545 = 4 ALUs (USSE1), 2 TMUs
SGX 543 = 4 ALUs (USSE2), 2 TMUs

Is it an A8 or an A9? no it's a A4 !!!

Wonder what the reason was for the name???

Here's what happens on each processor:

A8
---
----Adder 1 --- Adder 2
1) A=B+C
2) D=A+E
3) F=G+H


A9
---
----Adder 1 --- Adder2
1) A=B+C --- F=G+H
2) D=A+E

BUT THEY DON'T GO UP TO ELEVEN!!!

I studied really primitive ones in college in the late 1970's. It's all about clock signals and stuff going in and coming out and being shoved to and fro.

It sounds like there is a chip-mogul among us. Sir, can you confirm that it doesn't go up to eleven? You appear to have all the know-how.

Its like a contagious cancer, no one is immune from it.
Not even you.

You might want to do an upgrade on your medical knowledge. As far as I am aware, there is no human cancer that is contagious. Some illnesses that can cause cancer are, but that's not the same.

Yuck. You're getting a case of feature-itis. See Apple User Experience for a cure.

Again, I'm not sure that there is such a thing. I mean that so-called "Apple User Experience". It is a myth, isn't it?

Wow. No one believed me when I said PA didn't have time to do a full design. Until I start getting some respect around here I ain't spilling any beans

These kinds of postings and over-posting in general will never get you respect. When you unload a dozen posts within half an hour, you don't give people much of a chance to like you. At least with this one, it is just the case of skipping it. You don't want to read it? You don't have to, it's one post only.

 

[admanimal]

These kinds of postings and over-posting in general will never get you respect. When you unload a dozen posts within half an hour, you don't give people much of a chance to like you.

Yes, shame on cmaier for trying to share some actual facts with us rather than just carrying on with the ignorant supposition that is the hallmark of a good macrumors thread.

 

[gnasher729]

I studied really primitive ones in college in the late 1970's. It's all about clock signals and stuff going in and coming out and being shoved to and fro.

For a low power design, it is all about not doing anything useless and not spending energy on doing nothing. And on getting the performance out of the lowest possible frequency. So the Cortex A9's speculative execution means that sometimes it does something useless (speculating going wrong). Out-of-order execution takes complex hardware that costs energy.

The most power efficient design would be something like 2 core + hyper threading, low clock speed, no out of order execution, no speculative execution. I am sure the A9 has higher performance than the A8. I would think that Apple can get more power efficiency out of the A8.

hey, that was actually pretty insightful! i'm a performance-sensitive software engineer but i had never really stopped to think until now what the costs may be from mispredicted branches in an out-of-order design. heck, i guess that with sufficiently bad luck in OoOE you could kill your carefully gathered dcaches, and watch everything crumble down.. ok, now i'm scared.

There are cases where OoO kills performance on Intel processors, because it performs memory accesses speculatively to the wrong addresses. Something like

int index = i1;
    if (index is wildly wrong) index = i2;
    x = array [index];

will often try to read array [i1], which may cost significant time.

 

[MorphingDragon]

Well, it would be pretty stupid to try when it is basically an oversized and over-clocked iPod Touch.



As an oversized iPod Touch, it could have been under development since 2003. The technology was already there then.










BUT THEY DON'T GO UP TO ELEVEN!!!



It sounds like there is a chip-mogul among us. Sir, can you confirm that it doesn't go up to eleven? You appear to have all the know-how.



You might want to do an upgrade on your medical knowledge. As far as I am aware, there is no human cancer that is contagious. Some illnesses that can cause cancer are, but that's not the same.




Again, I'm not sure that there is such a thing. I mean that so-called "Apple User Experience". It is a myth, isn't it?



These kinds of postings and over-posting in general will never get you respect. When you unload a dozen posts within half an hour, you don't give people much of a chance to like you. At least with this one, it is just the case of skipping it. You don't want to read it? You don't have to, it's one post only.

Mr literal forgot that if it doesn't make sense its either a fanboy or a joke.

There are cases where OoO kills performance on Intel processors, because it performs memory accesses speculatively to the wrong addresses. Something like

int index = i1;
    if (index is wildly wrong) index = i2;
    x = array [index];

will often try to read array [i1], which may cost significant time.

Wait, Intel or X86 in general? If its just intel that would explain the crap memory bandwidth. I swear I saw in a benchmark that the new AM3 semprons beat a lot of the C2Ds in memory bandwidth, and the Athlon II beating everything up until the Nehalem chips using Triple Channel.

 

[blackcrayon]

that's why they got rid of the "our most advanced technology" from the website and just left the part about the magical device

You realize it still says that on the Apple Store page, right?

 

[pmz]

Originally Posted by HyperZboy...

A) No FLASH support on a device that is supposed to be a computer, NOT a HUGE iPOD.
B) No multitasking. And this processor rumor just adds insult to injury there too, no doubt.
C) No Mac or PC software support & requires developers to create all new apps that are basically iPhone/iPOD Touch apps.



Steve Jobs specifically said the device would be much better than a netbook, which is technically a cheap laptop computer that does the things I mentioned above while the iPAD does not. If you don't believe that is true, then don't bother reading any further, the Reality Distortion Field has already kicked into high gear! LOL

As an ebook reader and iPhone/iPOD app player, the iPAD rocks, but I already have those. I can't imagine carrying around a laptop or netbook, and an iPhone or iPOD + an iPAD.

This device doesn't replace any of them! LOL
This device doesn't even fully work with the internet like a netbook, let alone run real programs like a netbook. It's like the coolest device without a purpose I've ever seen.

A netbook runs a REAL computer operating system, NOT A PHONE OS and is potentially cheaper and can more likely replace a laptop for many things while traveling.

I imagine if you own none of the devices I mentioned above or have beaucoup dollars lying around, you might want this, even I'm intrigued & fascinated by it & would love to try one out, but I'm predicting this is a Steve Jobs failure of G4 Cube proportions, tons of hype, was incredibly cool, but nobody bought it.
The iPAD seems similarly & eerily incredibly cool, but I just can't imagine what I'd use it for at the price.



Even MacRumors reported that Apple posted pics on their website of the New York Times on an iPAD showing Flash Video.

Read back and then get back to me. Geez, the fanboys just refuse to believe the truth no matter what!
LOL

And btw, a large number of major newspapers, TV news stations ALL use FLASH video.

Just hating FLASH will not make it go away and I don't think Steve Jobs has the ability to make it go away even at this point, but it's clear he'd definitely like it to.

I'll be the 1st to admit I'm mad as hell at Adobe for making Flash so buggy on Macs.

But that still doesn't take away the limited nature of the iPAD because of the lack of it.
On a phone, it's ok really, I don't really mind, but on something that's supposed to better than a netbook... "Houston, we have a problem."

Oh for crap's sake, its not an issue on 75 million iOS devices. A few more million won't make a difference either.

MOVE ON!

 

[kemal]

Not sure why that shows up, possibly they are memory mapped files. In any case, you're not right about this: the iPhone does not back memory with disk storage.

http://developer.apple.com/iphone/l...tual/ManagingMemory/Articles/AboutMemory.html

"In Mac OS X, the virtual memory system often writes pages to the backing store. The backing store is a disk-based repository containing a copy of the memory pages used by a given process. Moving data from physical memory to the backing store is called paging out (or “swapping out”); moving data from the backing store back in to physical memory is called paging in (or “swapping in”). In iPhone OS, there is no backing store and so pages are are never paged out to disk, but read-only pages are still be paged in from disk as needed."

Second, I never said the CPU had trouble multitasking. I said it was the OS memory system and battery life.

Third, 20 minutes? How about some more rigorous testing.

Thanks for the correction on the VM.
Yes 20 minutes is not a good test. However, consider the only things that weren't full on were bluetooth and flash writes. WiFi, backlight and processor at 100% barely warmed the case up.

I don't buy battery life as a reason for not having apps stay active.

How bad is using MLC nand flash in Apple's vm implementation? Swap outs could be a real problem for the 128MB devices. Maybe Apple will make a multitasking OS only for the devices with RAM >=256MB(?)

Thanks /kemal

 

[840quadra]

Wow. No one believed me when I said PA didn't have time to do a full design. Until I start getting some respect around here I ain't spilling any beans

Haha you have respect, Don't worry about it!

These kinds of postings and over-posting in general will never get you respect. When you unload a dozen posts within half an hour, you don't give people much of a chance to like you. At least with this one, it is just the case of skipping it. You don't want to read it? You don't have to, it's one post only.

Sorry, who are you again?

 

[AidenShaw]

How bad is using MLC nand flash in Apple's vm implementation? Swap outs could be a real problem for the 128MB devices.

See the Apple quote above:

In iPhone OS, there is no backing store and so pages are are never paged out to disk​

 

[diamond.g]

See the Apple quote above:

In iPhone OS, there is no backing store and so pages are are never paged out to disk​

So when you run out of physical ram all you can do is (hopefully gracefully) die?

I wonder if that is why there haven't been any games that really push iPhone OS devices. It appears that the (later) hardware is capable of Gears of War type graphics (purely on a technical level) but RAM shortage can be problematic.

 

[kernkraft]

Sorry, who are you again?

I'm the guy on your ignore list. Or should be...

For you, the emphasis was on "At least with this one, it is just the case of skipping it. You don't want to read it? You don't have to, it's one post only."

My point was that there is no need to post around 25 posts in a few hours in one thread. As I just read them, I can see that a number of them were even merged by the administrator as they followed one another. So, it's not just me...

 

[cmaier]

Yes, shame on cmaier for trying to share some actual facts with us rather than just carrying on with the ignorant supposition that is the hallmark of a good macrumors thread.

If I understood what he was trying to say I'd probably cry.

Now I'm looking forward to the apple HTC patent lawsuit thread - for the last few years I've been a patent attorney so I can share my genius there, too

 

[840quadra]

So when you run out of physical ram all you can do is (hopefully gracefully) die?

I wonder if that is why there haven't been any games that really push iPhone OS devices. It appears that the (later) hardware is capable of Gears of War type graphics (purely on a technical level) but RAM shortage can be problematic.

If an application is using too much memory (under the "mobile" user), it is automatically killed in order to save "root" processes. (from my understanding and experience).

I'm the guy on your ignore list. Or should be...

<Rambling Diatribe removed>

If you say so, but I can make decisions like that without recommendations.

Unlike some (perhaps many), I value all opinions, and appreciate (or respect), people who take time to answer questions posed to them. If you actually read the posts you are complaining about, you would see that cmaier was responding (on topic) to many people. If you don't like seeing people help others, you are on the wrong website and discussion forums.

 

[kernkraft]

Unlike some (perhaps many), I value all opinions, and appreciate (or respect), people who take time to answer questions posed to them.
<Rambling Diatribe removed>

I'm sorry, but who are you again? (Not my words...)

 

[firewood]

Originally Posted by HyperZboy...

A) No FLASH support on a device that is supposed to be a computer...

A computer doesn't have to run flash to be called a computer. For example: Ascii White and Blue Gene, which are some of the most powerful computers on the planet.

 

[kernkraft]

It's not a PC and not a Mac...

A computer doesn't have to run flash to be called a computer. For example: Ascii White and Blue Gene, which are some of the most powerful computers on the planet.

Oh yeah, those! That's what my neighbour uses for web-browsing (I suppose, it's not porn and video).

 

[gnasher729]

Wait, Intel or X86 in general? If its just intel that would explain the crap memory bandwidth. I swear I saw in a benchmark that the new AM3 semprons beat a lot of the C2Ds in memory bandwidth, and the Athlon II beating everything up until the Nehalem chips using Triple Channel.

Since I don't have a Mac with an AMD chip, I can only talk about Intel. But this is very, very, very specific code. Say you write a loop

for (i = 0; i < 100; ++i) sum += a ;

then it is quite possible that due to speculation the processor will try to read a [100]. That's no problem and will generally have very, very little effect on performance. Now look at this code:

 bool whichIndexArray [1000000];
   long indexArray1 [1000000];
   long indexArray2 [1000000];
   double data [100];

   for (i = 0; i < 1000000; ++i) {
    long index = whichIndexArray [i] ? indexArray1 [i] : indexArray2 [i]; 
    sum += data [index];
  }

Lets say whichIndexArray is mostly filled with TRUEs and 10% FALSE, quite randomly. Then due to speculation, the processor will _always_ try to read data [indexArray1 ]. Now assume that when whichIndexArray contains FALSE, indexArray1 contains large random numbers. Then the processor will try to read a random location. If the unused locations in indexArray1 all contained a zero, it would be fine. It is an unusual situation, usually you'll have no problems.

You can try to benchmark this on an AMD chip if you like, I wouldn't bet that it would do any better than Intel. And I wouldn't bet that it would do better than an Intel chip in any memory bandwidth benchmark.

 

[cmaier]

Since I don't have a Mac with an AMD chip, I can only talk about Intel. But this is very, very, very specific code. Say you write a loop

for (i = 0; i < 100; ++i) sum += a ;

then it is quite possible that due to speculation the processor will try to read a [100]. That's no problem and will generally have very, very little effect on performance. Now look at this code:

 bool whichIndexArray [1000000];
   long indexArray1 [1000000];
   long indexArray2 [1000000];
   double data [100];

   for (i = 0; i < 1000000; ++i) {
    long index = whichIndexArray [i] ? indexArray1 [i] : indexArray2 [i]; 
    sum += data [index];
  }

Lets say whichIndexArray is mostly filled with TRUEs and 10% FALSE, quite randomly. Then due to speculation, the processor will _always_ try to read data [indexArray1 ]. Now assume that when whichIndexArray contains FALSE, indexArray1 contains large random numbers. Then the processor will try to read a random location. If the unused locations in indexArray1 all contained a zero, it would be fine. It is an unusual situation, usually you'll have no problems.

You can try to benchmark this on an AMD chip if you like, I wouldn't bet that it would do any better than Intel. And I wouldn't bet that it would do better than an Intel chip in any memory bandwidth benchmark.

I would guess that AMD would do a little better. I designed the execution unit on athlon 64 and opteron, including the out-of-order issue hardware. We were less aggressive than Intel (though I'm not sure what recent Intel hardware does).

 

[Jimmy James]

It looked snappy in the presentation. It looked like it had flash in the marketing photos. We'll see.

 

[lilo777]

A computer doesn't have to run flash to be called a computer. For example: Ascii White and Blue Gene, which are some of the most powerful computers on the planet.

But... The computer has to run flash to be useful