........And if 45nm can bring such amazing power savings to ARM9, quoted at 10-20% over ARM8, then the savings will be even greater if ARM8 were shrunk. There's nothing saying that Apple couldn't find someone to fab their own ARM8 SoC at 45nm. Apple could use the shrink to upclock the ARM8 a bit over that of the iPhone 3G S to offer more speed just like the 2nd gen iPod Touch, without splitting the ecosystem with ARM9. An ARM9 Apple SoC can come in iPhone 2011 using the experience picked up from an ARM8 SoC.
I'm wondering if one of us misread the bit about 10-20%. Because I read that as 10-20% WORST than A8. From other data I've seen I'd expect that to be true.
Oh please put away the garbage about splitting the ecosystem. The last thing we need is a stagnet ecosystem. My 3G simply is not fast enough to make me totally happy. {actually hoping 3.0 helps here}. That running Apples apps and nothing fancy from app store. The 's' may be better but I'm not convinced it is going to be that stunning once on the faster network connections. I fully expect and want a good update this time next year.
Well reason post Lt. Though I just have one problem
I really agree with Melgross in that Apple is not the 800lb Gorilla of smartphones. They sell a whole bunch but the competiton is fielding better and better phones (see Palm Pre)
If an ARM representative says to me that they are shipping Cortex A9 based smartphones I'm not really thinking "will Apple use this core?" I'm more inclined to think along the terms of "Apple better use this core because the competition is"
Granted I believe Apple will likely have a custom designed SoC it'll be based on A9 core[s].
With Apple's size and influence the question to ask is "why wouldn't they be the first company to bring us MPcore smartphones?"
I guess we're just going to have to wait and see but my money is on much beefier hardware with the next version and HD support with HD video output.
"Stop gap" may not be the right term to describe the jump from the 3G to the 3G S.
Dana, sorry about the pile-on. I didn't see that Mel replied just before me. \
In any case, it looks like we agree on just about everything. As long as the subsidy model is in place (and iPhone releases are a yearly thing), I aim to be on a two year cycle myself. I know it will be hard next June if some new awesomeness is released, but I have showed that I have will power--this one will be my first iPhone.
Well, that or I am just cheap. Heck, I am still using my 15" TiBook everyday...man I'd love a unibody to go with this shiny new iPhone...
In terms of version numbers, going from 2.1 to 3.1 may indicate a major change but doesn't have to call for a processor architecture change. For example, the first major version change in the iPhone/iPod Touch platform came on the 2nd Gen iPod Touch which used the same ARM11 processor just clocked higher to yield going from 1.1 to 2.1. A processor clock speed increase like I'm advocating with a 2010 Apple designed ARM8 SoC would already seem sufficient enough to go from 2.1 to 3.1, especially when combined with a likely external design refresh. The early leak of the iPhone 3.1 designation before iPhone 2.1 is even released may even support an upclocked ARM8 feature-set, since it'd better explain the quick turnaround time.
Hmm, a simple boost in clock speed doesn't seem quite up to par with what Apple has given us in past updates. Remember, a new case, GPS and 3G only warranted 1.2. Where as an major internal improvement in the 3GS led to a great change. Something tells me a small clock boost would not warrant the same change as the 3GS. The early 3.1 could also support the A9. While it hasn't yet been implemented as much specs have been available for Apple.
A simple clock boost doesn't seem to match what the 3GS delivered. Double the ram, improved GPU...etc. 2.2 would seem much more likely for a clock boost.
It's my belief that Apple had full intentions of implementing the A9 in 2010, and the 3GS was the result of needing some boost between the 3G and the 2010 phone. The gap between the A9 and the current A11 in the iPhone would be far to large a gap for developers to bridge with their Apps. Instead an incremental A8, would allow both generations of users to experience the more powerful apps.
I'm wondering if one of us misread the bit about 10-20%. Because I read that as 10-20% WORST than A8. From other data I've seen I'd expect that to be true.
Oh please put away the garbage about splitting the ecosystem. The last thing we need is a stagnet ecosystem. My 3G simply is not fast enough to make me totally happy. {actually hoping 3.0 helps here}. That running Apples apps and nothing fancy from app store. The 's' may be better but I'm not convinced it is going to be that stunning once on the faster network connections. I fully expect and want a good update this time next year.
Dave
Hmm. I did misread the article, ARM A9 is 10-20% worse in peak power consumption than ARM A8 but better in average consumption probably because it can idle one core.
Well it's not like I'm not supporting faster iPhones. Like John Cormack was saying, Apple could choose to get end-user speed increases 2 ways, either through hardware or through software. In this case, with iPhone 3G S they went with hardware from ARM11+MBX to ARM A8+SGX marketed as a 2 times speed increase. Cormack thinks that a similar 2 times speed increase could also be achieved on existing ARM11+MBX hardware if Apple optimized the OS and OpenGL stack. There would no doubt still be performance left on the table with ARM A8 in a years time that can be exploited. Really, software optimization is ideal for the user since it can benefit everyone, while hardware improvements may be faster to implement for Apple but require upgrades, which admittedly is how Apple can make money.
Quote:
Originally Posted by hmurchison
Well reason post Lt. Though I just have one problem
I really agree with Melgross in that Apple is not the 800lb Gorilla of smartphones. They sell a whole bunch but the competiton is fielding better and better phones (see Palm Pre)
If an ARM representative says to me that they are shipping Cortex A9 based smartphones I'm not really thinking "will Apple use this core?" I'm more inclined to think along the terms of "Apple better use this core because the competition is"
Granted I believe Apple will likely have a custom designed SoC it'll be based on A9 core[s].
With Apple's size and influence the question to ask is "why wouldn't they be the first company to bring us MPcore smartphones?"
I guess we're just going to have to wait and see but my money is on much beefier hardware with the next version and HD support with HD video output.
I guess I'm just thinking it's a matter of focus. Using the ARM A9 in itself probably won't cause people to jump to buy an iPhone. However, a focus on usability, portability, battery life, integration with your computer, etc. will more likely win people over regardless over whether an older ARM A8 CPU is used as long as it's not drastically slower. Of course, it doesn't have to be either or. The best case scenario will be killer software with bleeding edge hardware.
7.2 Mbps is pretty good too, hardly anything to scoff at.
I didn't mean to sound like I was belittling the radio, just implying that the current network and HW capabilites are well below that of LTE making it not feasible for next year.
While I expected HSUPA with an upload speed greater that the 384kbps HSDPA allows for, the 7.2mbps down was expected. I don't know of a single phone that has 14.4mbps down. Even Nokia's new N97 only has 3.6mbps HSDPA, which I find very surprising, though there is sone talk that a firmware update could bump it up to 7.2mbps.
Hmm, a simple boost in clock speed doesn't seem quite up to par with what Apple has given us in past updates. Remember, a new case, GPS and 3G only warranted 1.2. Where as an major internal improvement in the 3GS led to a great change. Something tells me a small clock boost would not warrant the same change as the 3GS. The early 3.1 could also support the A9. While it hasn't yet been implemented as much specs have been available for Apple.
A simple clock boost doesn't seem to match what the 3GS delivered. Double the ram, improved GPU...etc. 2.2 would seem much more likely for a clock boost.
It's my belief that Apple had full intentions of implementing the A9 in 2010, and the 3GS was the result of needing some boost between the 3G and the 2010 phone. The gap between the A9 and the current A11 in the iPhone would be far to large a gap for developers to bridge with their Apps. Instead an incremental A8, would allow both generations of users to experience the more powerful apps.
I agree and would also add that the feature people want most is multitasking. Even with the power and battery savings of the 3gs, Apple still thought that it couldn't handle it well enough. It would only stand to reason they think the A9 with twice the computing power and better battery life can. Especially when you consider PA Semi could get to add their own influence to increase battery life and performance. Chances are Apple will get a very early look at these chips because of their long relationship with ARM.
While others have said that it would be too soon of a move, Apple is going to have to make it sooner or later and there will be people upset by it. Apple has always shown that they will adopt new tech, even if it means leaving others behind. My iMac that I bought last year is not even on the supported card list for H.264 acceleration in Snow Leopard. As the Mac moves to Open CL, it would figure that they would do the same with the iPhone. This move to Open CL and the A9 will be one that lasts for awhile.
Contrary to what other think, I believe the 3gs is a stopgap move. The power right now is only being used for voice control, video, and to make apps load faster. It seems like it is the final product of this generation of hardware. Even the model name suggests that. Developers are not going to make apps that utilizes it's power anyway because they won't eliminate older iPhones and because Open CL will probably make it's appearance next year.
Don't know about you guys, but I'll go on record and say that the performance increase from the iPhone 3G to the iPhone 3G S will be the single greatest performance leap that the iPhone will ever have. You will never see an across the board 2x or greater increase in performance ever again.
The 3G to 3G S CPU performance increase is a single threaded performance increase and applies to all software. The next performance increase may come close when/if an out-of-order micro-architecture ships, but that comes at a cost of increased W/MHz compared to the in-order, so it is going to be power constrained. Other potential future increases will come from adding cores, and that really doesn't increase performance across the board. Throughput yes, single threaded performance no. So only applications that are easily multithreaded will see the performance increase. The 90% of the rest of the software will get by with measly 10-25% increases per year.
If Apple ships a dual-core ARM, it will be the form of a $500+ super-iPhone or super-iPod or media pad or whatever. It won't be in your regular $200 iPhone, which will likely be a 25% clock rate bump (and maybe a doubling in GPU performance).
I think the "shot across the bow" of iPhone developers was the announcement that Apple would be aggressive in promoting SDK 3.0 apps. Hell what am I talking about ..Apple requires them for the most port.
This sets the trend here. I expect that as Apple takes advantage of the more sophisticated ARM architecture their SDK will evolve quickly in support and developers will be moved up aggressively.
I think that we'll see push notifications for the most part but that eventually more 3rd party apps will be allowed to multitask. It'll be opt in so if your battery life sucks it'll be your fault for enabling multitasking for too many apps.
I do see the value of streaming some content while doing other things.
Well raw specs hasn't always been Apple's thing. The critics have long pointed out the many features that the iPhone has been missing to date like copy and paste, bluetooth headphones, MMS, 3G in the beginning, etc. yet that hasn't stopped the iPhone/iPod Touch from selling 40 million units in 2 years. Even if competitors adopt the ARM9 earlier Apple can still differentiate themselves on usability, style, and with a 45nm ARM8 shrink, better battery life than ARM9 phones.
What's more, there is plenty of power in existing architectures for Apple to exploit if needed. John Cormack has estimated that the ARM11+MBX in existing iPhones/iPod Touches can see a 2 times increase in speed for games if Apple only optimized their software. Even if 2 times is a bit much, it isn't hard to believe that there is quite a bit of potential to be exploited in either ARM11 or ARM8. Initial ARM9 implementations aren't likely to be the most efficient from either a software perspective or a hardware SoC design perspective, so a highly optimized design using the mature ARM8 won't necessarily be left in the dust. Especially, if Apple uses a 45nm shrink and clocks the ARM8 higher to close the gap.
Optimization isn't that easy. Carmack can say whatever he wants to about that, but he's very narrow minded in his outlook. Apple has done what they have with the OS in order to bring a much richer experience than just gaming to the table. They have to have though several year out with this. I imagine that they know where they're going, and that optimizations won't help them, as they need to add more features, which may depend on how the OS is now written.
This is why hardware is needed to complete the circle.
I don't understand why you and a few others would doubt the hardware path that seems pretty obvious. We can only assume that Apple's future path is to continue to strain the hardware as much as possible in order to give a better experience.
I agree with Tim here, and if he's correct, the more processing power, the better:
I'm wondering if one of us misread the bit about 10-20%. Because I read that as 10-20% WORST than A8. From other data I've seen I'd expect that to be true.
Dave
But it's only worse when the device is running flat out. It's more efficient when doing most other tasks. It's like Intel's new Nehalem line-up. They're not like Prescott of course, but their max power envelop isn't low. but the idle power draw is very low, and draw during average use is low. Overall, the average is lower than the previous generation while delivering much more computing power. More bang for the watt.
The same is true for the 9 vs the 8. Peak power usage is somewhat more, but overall, it's less.
From everything we've read so far, it does seem that the "s" will be more than a bit faster at many things. One of those things is rendering a web page. Downloading the page isn't the only aspect to the speed. If it takes as long as, or even longer to render a complex page, then everything is slow.
I guess I'm just thinking it's a matter of focus. Using the ARM A9 in itself probably won't cause people to jump to buy an iPhone. However, a focus on usability, portability, battery life, integration with your computer, etc. will more likely win people over regardless over whether an older ARM A8 CPU is used as long as it's not drastically slower. Of course, it doesn't have to be either or. The best case scenario will be killer software with bleeding edge hardware.
I know it's been stated earlier that the manufacturers are selling their new phone on the higher performance processing junk inside their phones, but its not true.
The manufacturers don't like taking about what's inside the phones at all. They want to talk about the features that customers can interact with. What's inside simply makes that possible. There are far fewer processors for smartphones than there are processors for computers, so manufacturers know that getting into that kind of battle is a losing one, as most of them use the same chips, or similar enough ones so as not to matter.
I don't agree with Carmack. It's not likely that he knows as much about the OS as does the team that designed it. He's got his priorities and Apple has theirs.
The iPhone OS is still pretty young, Apple has more important things to worry about than the nebulous concept of optimizing.
For the next couple of years at least, it will be raw power that they are concerned with, along with chip optimizations that will likely allow them to do in hardware what they now do in software, which will speed things up considerably.
I agree and would also add that the feature people want most is multitasking. Even with the power and battery savings of the 3gs, Apple still thought that it couldn't handle it well enough. It would only stand to reason they think the A9 with twice the computing power and better battery life can. Especially when you consider PA Semi could get to add their own influence to increase battery life and performance. Chances are Apple will get a very early look at these chips because of their long relationship with ARM.
While others have said that it would be too soon of a move, Apple is going to have to make it sooner or later and there will be people upset by it. Apple has always shown that they will adopt new tech, even if it means leaving others behind. My iMac that I bought last year is not even on the supported card list for H.264 acceleration in Snow Leopard. As the Mac moves to Open CL, it would figure that they would do the same with the iPhone. This move to Open CL and the A9 will be one that lasts for awhile.
Contrary to what other think, I believe the 3gs is a stopgap move. The power right now is only being used for voice control, video, and to make apps load faster. It seems like it is the final product of this generation of hardware. Even the model name suggests that. Developers are not going to make apps that utilizes it's power anyway because they won't eliminate older iPhones and because Open CL will probably make it's appearance next year.
I don't understand why you insist on using the word "stopgap". It doesn't make sense.
Every advance is a "stopgap" until the next one comes out, which is in itself a stopgap before the next.
It loses all meaning in that context.
The speed increases in the 3Gs is likely about the same as we will see with the 9 when it hopefully comes out next year. Which is to say, about 2+ times faster than the precious generation.
Those are both large advances in speed.
This enables video recording and editing, voice control, not just voice dialing, video MMS. The new compass has enabled much better, true turn by turn programs. Real 3D games. I could go on, but you know the new possibilities that the new hardware has enabled.
The 3G was the stopgap, if you insist in using that word. There was NO advance in any area of the phone hardware at all, except for the addition of 3G and GPS.
You have to explain how that minor advance wasn't the stopgap, and the 3Gs isn't the major advance.
Is it because, for some reason, you're not happy with what they've been able to do?
I look forward to this new computing horsepower. There are many times that the bottleneck of displaying bloated or complicated webpages on my iPhone 3G is the rendering speed rather than the downloading speed. I know that because the time it took to display those bloated webpages via Wi-Fi (no connection or signal issues involved) is pretty much the same as the one via 3G.
I don't understand why you insist on using the word "stopgap". It doesn't make sense.
Every advance is a "stopgap" until the next one comes out, which is in itself a stopgap before the next.
It loses all meaning in that context.
The speed increases in the 3Gs is likely about the same as we will see with the 9 when it hopefully comes out next year. Which is to say, about 2+ times faster than the precious generation.
Those are both large advances in speed.
This enables video recording and editing, voice control, not just voice dialing, video MMS. The new compass has enabled much better, true turn by turn programs. Real 3D games. I could go on, but you know the new possibilities that the new hardware has enabled.
The 3G was the stopgap, if you insist in using that word. There was NO advance in any area of the phone hardware at all, except for the addition of 3G and GPS.
You have to explain how that minor advance wasn't the stopgap, and the 3Gs isn't the major advance.
Is it because, for some reason, you're not happy with what they've been able to do?
If the word "stopgap" is too strong of a word for you, then I apologize. I'm not saying that the 3gs is not a powerful device. Nor am I saying it did not bring any features. But if you look in terms of a larger context it is clear that is not looking towards this iPhone as the one that signals the next-gen of iPhone hardware. This phone does not have any influence from PA Semi or the numerous graphics people they hired away from ATI. That alone should say that they are aiming for something more. Open CL is right around the corner and will soon move to it.
Don't say 3D games. I can count on one hand the amount of developers that will make the maximum use of this hardware at best. Developers are not going to throw away the potential cash cow of developing for all the past iPhone/ iPod touch users for the 3gs. The only way that will happen is when forces a break. As I said, has still not allowed multitasking even though the 3gs has better battery life and a far better processor. If they released only a minor upgrade with another A8 next year, there would still be no multitasking. With a dual processors and much better battery life in the A9, they would have it. It's hard for me to see that would go another year without multitasking.
You may say that the 3g is a stopgap and that is probably true. But I'm not looking at last year. I'm looking at today and what will soon be coming tomorrow. The hardware alone will be a significant release because it will be more powerful and have more battery life. This is almost like 's move to Intel's dual cores.
I'm quite fine with what 's done. They have as powerful a phone as any out there. I'd like to see multitasking and an API for integration between apps but it has some nice features that almost make me forget about them. I'm just being honest about it. Hell, I may even buy one. But I also know what is coming on the horizon. If anyone wanted to buy this phone, I'd say go for it.
If the word "stopgap" is too strong of a word for you, then I apologize. I'm not saying that the 3gs is not a powerful device. Nor am I saying it did not bring any features. But if you look in terms of a larger context it is clear that is not looking towards this iPhone as the one that signals the next-gen of iPhone hardware. This phone does not have any influence from PA Semi or the numerous graphics people they hired away from ATI. That alone should say that they are aiming for something more. Open CL is right around the corner and will soon move to it.
I understand what you're saying, but I don't know what that has to do with it. This is a much improved machine. It's not in any holding pattern, waiting for some miracle processor from Apple. It's an evolutionary product, and as such, it's done very well.
Next years' will be better.
Quote:
Don't say 3D games. I can count on one hand the amount of developers that will make the maximum use of this hardware at best. Developers are not going to throw away the potential cash cow of developing for all the past iPhone/ iPod touch users for the 3gs. The only way that will happen is when forces a break. As I said, has still not allowed multitasking even though the 3gs has better battery life and a far better processor. If they released only a minor upgrade with another A8 next year, there would still be no multitasking. With a dual processors and much better battery life in the A9, they would have it. It's hard for me to see that would go another year without multitasking.
You don't know that. I've already read about the 3D games that we'll be seeing, though not by name. It only takes a handful of cutting edge games to move the products to a new level. That's true of any gaming platform.
With this upgrade, Apple likely has enough oomp for multiprocessing if they want it. likely, they don't want it yet, because they want to come out with some framework that will enable it to be better than elsewhere. They have their priorities, as they had said about C/C/Paste last year.
Quote:
You may say that the 3g is a stopgap and that is probably true. But I'm not looking at last year. I'm looking at today and what will soon be coming tomorrow. The hardware alone will be a significant release because it will be more powerful and have more battery life. This is almost like 's move to Intel's dual cores.
Every new generation will be much faster and have more hardware features until the situation matures, when advances will slow down. We won't get double the processing power forever, perhaps for one or two more generations.
Quote:
I'm quite fine with what 's done. They have as powerful a phone as any out there. I'd like to see multitasking and an API for integration between apps but it has some nice features that almost make me forget about them. I'm just being honest about it. Hell, I may even buy one. But I also know what is coming on the horizon. If anyone wanted to buy this phone, I'd say go for it.
You can wait forever. I waited for two features. Two features I said I would wait for since we had out first discussions right after Apple announced the phone. That was 3G and supported native apps. GPS was a bonus I wanted, but wouldn't wait for.
The new 3Gs is an amazing machine. Much more so than the 3G, esp with the new OS.
Next year, you will hear of the possibility of even more amazing things with it, are you going to wait again? I don't recall now, did you say you have an older iPhone, or not? If so, then waiting makes more sense. We will likely wait, as we bought ours in late August, though my daughter will get the new one when she goes to school in London this summer (good excuse for her!).
If not, then unless you're waiting for a contract to end it just wastes another year.
The understandable reason for this processor upgrade is surely supporting VGA video at 30 fps. Besides, only competition remains. Just trying to be in line with achievements of Sunnyvale-based company.
I understand what you're saying, but I don't know what that has to do with it. This is a much improved machine. It's not in any holding pattern, waiting for some miracle processor from Apple. It's an evolutionary product, and as such, it's done very well.
Next years' will be better.
You don't know that. I've already read about the 3D games that we'll be seeing, though not by name. It only takes a handful of cutting edge games to move the products to a new level. That's true of any gaming platform.
With this upgrade, Apple likely has enough oomp for multiprocessing if they want it. likely, they don't want it yet, because they want to come out with some framework that will enable it to be better than elsewhere. They have their priorities, as they had said about C/C/Paste last year.
Every new generation will be much faster and have more hardware features until the situation matures, when advances will slow down. We won't get double the processing power forever, perhaps for one or two more generations.
You can wait forever. I waited for two features. Two features I said I would wait for since we had out first discussions right after Apple announced the phone. That was 3G and supported native apps. GPS was a bonus I wanted, but wouldn't wait for.
The new 3Gs is an amazing machine. Much more so than the 3G, esp with the new OS.
Next year, you will hear of the possibility of even more amazing things with it, are you going to wait again? I don't recall now, did you say you have an older iPhone, or not? If so, then waiting makes more sense. We will likely wait, as we bought ours in late August, though my daughter will get the new one when she goes to school in London this summer (good excuse for her!).
If not, then unless you're waiting for a contract to end it just wastes another year.
I think you're missing several points on the 3D games thing. First off, the developers would be alienating 40 million consumers in favor of 1-2 million. But, some developers will do it to simply put out a 3GS "only" app. It will be a novel hit since all the 3GS users will want to show off their phone. However there are even more roadblocks to such apps.
Money is a large issue when it comes to developing advanced games as the ones you're talking about. Games like that take almost a year to make (see PSP and DS games) and once they're made they cost $30-40. That's a huge jump over what we have in the app store now. We're so used to $5-15 games that $30-40 seems rather steep.
So here's what you have working against you when creating a 3GS app:
1. Development costs are much higher
2. Development time is much longer, new iPhone by the time the best ones come out?
3. Your price in the app store is much higher
4. Your consumers base is much, much smaller (1/40th the size!)
So here's what you have working against you when creating a 3GS app:
1. Development costs are much higher
2. Development time is much longer, new iPhone by the time the best ones come out?
3. Your price in the app store is much higher
4. Your consumers base is much, much smaller (1/40th the size!)
I don?t agree with any of that.
1. It?s still using the same Xcode.
2. It may be longer, but only because the developer has more options available to them. It may also be shorter as the developer no longer has to spend extra time shoehorning the app into older hardware to compete with newer apps that are utilizing the latest HW.
3.Still using the same Xcode. There are potential reasons why it could could more but nothing that states it will higher specifically for coding for the new HW. In fact, it?ll be cheaper than coding for both the new and old.
4. How you get 1/40 the size? Is that because there are 40M iPhone OS X devices out there now? how do you get one million as of today or as of one month from now. How about 6 months from now or one year from now when this new HW is the standard for future-forward app development?
Comments
........And if 45nm can bring such amazing power savings to ARM9, quoted at 10-20% over ARM8, then the savings will be even greater if ARM8 were shrunk. There's nothing saying that Apple couldn't find someone to fab their own ARM8 SoC at 45nm. Apple could use the shrink to upclock the ARM8 a bit over that of the iPhone 3G S to offer more speed just like the 2nd gen iPod Touch, without splitting the ecosystem with ARM9. An ARM9 Apple SoC can come in iPhone 2011 using the experience picked up from an ARM8 SoC.
I'm wondering if one of us misread the bit about 10-20%. Because I read that as 10-20% WORST than A8. From other data I've seen I'd expect that to be true.
Oh please put away the garbage about splitting the ecosystem. The last thing we need is a stagnet ecosystem. My 3G simply is not fast enough to make me totally happy. {actually hoping 3.0 helps here}. That running Apples apps and nothing fancy from app store. The 's' may be better but I'm not convinced it is going to be that stunning once on the faster network connections. I fully expect and want a good update this time next year.
Dave
I really agree with Melgross in that Apple is not the 800lb Gorilla of smartphones. They sell a whole bunch but the competiton is fielding better and better phones (see Palm Pre)
If an ARM representative says to me that they are shipping Cortex A9 based smartphones I'm not really thinking "will Apple use this core?" I'm more inclined to think along the terms of "Apple better use this core because the competition is"
Granted I believe Apple will likely have a custom designed SoC it'll be based on A9 core[s].
With Apple's size and influence the question to ask is "why wouldn't they be the first company to bring us MPcore smartphones?"
I guess we're just going to have to wait and see but my money is on much beefier hardware with the next version and HD support with HD video output.
Okay, okay folks? Point taken.
"Stop gap" may not be the right term to describe the jump from the 3G to the 3G S.
Dana, sorry about the pile-on. I didn't see that Mel replied just before me.
In any case, it looks like we agree on just about everything. As long as the subsidy model is in place (and iPhone releases are a yearly thing), I aim to be on a two year cycle myself. I know it will be hard next June if some new awesomeness is released, but I have showed that I have will power--this one will be my first iPhone.
Well, that or I am just cheap. Heck, I am still using my 15" TiBook everyday...man I'd love a unibody to go with this shiny new iPhone...
In terms of version numbers, going from 2.1 to 3.1 may indicate a major change but doesn't have to call for a processor architecture change. For example, the first major version change in the iPhone/iPod Touch platform came on the 2nd Gen iPod Touch which used the same ARM11 processor just clocked higher to yield going from 1.1 to 2.1. A processor clock speed increase like I'm advocating with a 2010 Apple designed ARM8 SoC would already seem sufficient enough to go from 2.1 to 3.1, especially when combined with a likely external design refresh. The early leak of the iPhone 3.1 designation before iPhone 2.1 is even released may even support an upclocked ARM8 feature-set, since it'd better explain the quick turnaround time.
Hmm, a simple boost in clock speed doesn't seem quite up to par with what Apple has given us in past updates. Remember, a new case, GPS and 3G only warranted 1.2. Where as an major internal improvement in the 3GS led to a great change. Something tells me a small clock boost would not warrant the same change as the 3GS. The early 3.1 could also support the A9. While it hasn't yet been implemented as much specs have been available for Apple.
A simple clock boost doesn't seem to match what the 3GS delivered. Double the ram, improved GPU...etc. 2.2 would seem much more likely for a clock boost.
It's my belief that Apple had full intentions of implementing the A9 in 2010, and the 3GS was the result of needing some boost between the 3G and the 2010 phone. The gap between the A9 and the current A11 in the iPhone would be far to large a gap for developers to bridge with their Apps. Instead an incremental A8, would allow both generations of users to experience the more powerful apps.
I'm wondering if one of us misread the bit about 10-20%. Because I read that as 10-20% WORST than A8. From other data I've seen I'd expect that to be true.
Oh please put away the garbage about splitting the ecosystem. The last thing we need is a stagnet ecosystem. My 3G simply is not fast enough to make me totally happy. {actually hoping 3.0 helps here}. That running Apples apps and nothing fancy from app store. The 's' may be better but I'm not convinced it is going to be that stunning once on the faster network connections. I fully expect and want a good update this time next year.
Dave
Hmm. I did misread the article, ARM A9 is 10-20% worse in peak power consumption than ARM A8 but better in average consumption probably because it can idle one core.
Well it's not like I'm not supporting faster iPhones. Like John Cormack was saying, Apple could choose to get end-user speed increases 2 ways, either through hardware or through software. In this case, with iPhone 3G S they went with hardware from ARM11+MBX to ARM A8+SGX marketed as a 2 times speed increase. Cormack thinks that a similar 2 times speed increase could also be achieved on existing ARM11+MBX hardware if Apple optimized the OS and OpenGL stack. There would no doubt still be performance left on the table with ARM A8 in a years time that can be exploited. Really, software optimization is ideal for the user since it can benefit everyone, while hardware improvements may be faster to implement for Apple but require upgrades, which admittedly is how Apple can make money.
Well reason post Lt. Though I just have one problem
I really agree with Melgross in that Apple is not the 800lb Gorilla of smartphones. They sell a whole bunch but the competiton is fielding better and better phones (see Palm Pre)
If an ARM representative says to me that they are shipping Cortex A9 based smartphones I'm not really thinking "will Apple use this core?" I'm more inclined to think along the terms of "Apple better use this core because the competition is"
Granted I believe Apple will likely have a custom designed SoC it'll be based on A9 core[s].
With Apple's size and influence the question to ask is "why wouldn't they be the first company to bring us MPcore smartphones?"
I guess we're just going to have to wait and see but my money is on much beefier hardware with the next version and HD support with HD video output.
I guess I'm just thinking it's a matter of focus. Using the ARM A9 in itself probably won't cause people to jump to buy an iPhone. However, a focus on usability, portability, battery life, integration with your computer, etc. will more likely win people over regardless over whether an older ARM A8 CPU is used as long as it's not drastically slower. Of course, it doesn't have to be either or. The best case scenario will be killer software with bleeding edge hardware.
7.2 Mbps is pretty good too, hardly anything to scoff at.
I didn't mean to sound like I was belittling the radio, just implying that the current network and HW capabilites are well below that of LTE making it not feasible for next year.
While I expected HSUPA with an upload speed greater that the 384kbps HSDPA allows for, the 7.2mbps down was expected. I don't know of a single phone that has 14.4mbps down. Even Nokia's new N97 only has 3.6mbps HSDPA, which I find very surprising, though there is sone talk that a firmware update could bump it up to 7.2mbps.
Hmm, a simple boost in clock speed doesn't seem quite up to par with what Apple has given us in past updates. Remember, a new case, GPS and 3G only warranted 1.2. Where as an major internal improvement in the 3GS led to a great change. Something tells me a small clock boost would not warrant the same change as the 3GS. The early 3.1 could also support the A9. While it hasn't yet been implemented as much specs have been available for Apple.
A simple clock boost doesn't seem to match what the 3GS delivered. Double the ram, improved GPU...etc. 2.2 would seem much more likely for a clock boost.
It's my belief that Apple had full intentions of implementing the A9 in 2010, and the 3GS was the result of needing some boost between the 3G and the 2010 phone. The gap between the A9 and the current A11 in the iPhone would be far to large a gap for developers to bridge with their Apps. Instead an incremental A8, would allow both generations of users to experience the more powerful apps.
I agree and would also add that the feature people want most is multitasking. Even with the power and battery savings of the 3gs, Apple still thought that it couldn't handle it well enough. It would only stand to reason they think the A9 with twice the computing power and better battery life can. Especially when you consider PA Semi could get to add their own influence to increase battery life and performance. Chances are Apple will get a very early look at these chips because of their long relationship with ARM.
While others have said that it would be too soon of a move, Apple is going to have to make it sooner or later and there will be people upset by it. Apple has always shown that they will adopt new tech, even if it means leaving others behind. My iMac that I bought last year is not even on the supported card list for H.264 acceleration in Snow Leopard. As the Mac moves to Open CL, it would figure that they would do the same with the iPhone. This move to Open CL and the A9 will be one that lasts for awhile.
Contrary to what other think, I believe the 3gs is a stopgap move. The power right now is only being used for voice control, video, and to make apps load faster. It seems like it is the final product of this generation of hardware. Even the model name suggests that. Developers are not going to make apps that utilizes it's power anyway because they won't eliminate older iPhones and because Open CL will probably make it's appearance next year.
The 3G to 3G S CPU performance increase is a single threaded performance increase and applies to all software. The next performance increase may come close when/if an out-of-order micro-architecture ships, but that comes at a cost of increased W/MHz compared to the in-order, so it is going to be power constrained. Other potential future increases will come from adding cores, and that really doesn't increase performance across the board. Throughput yes, single threaded performance no. So only applications that are easily multithreaded will see the performance increase. The 90% of the rest of the software will get by with measly 10-25% increases per year.
If Apple ships a dual-core ARM, it will be the form of a $500+ super-iPhone or super-iPod or media pad or whatever. It won't be in your regular $200 iPhone, which will likely be a 25% clock rate bump (and maybe a doubling in GPU performance).
This sets the trend here. I expect that as Apple takes advantage of the more sophisticated ARM architecture their SDK will evolve quickly in support and developers will be moved up aggressively.
I think that we'll see push notifications for the most part but that eventually more 3rd party apps will be allowed to multitask. It'll be opt in so if your battery life sucks it'll be your fault for enabling multitasking for too many apps.
I do see the value of streaming some content while doing other things.
Well raw specs hasn't always been Apple's thing. The critics have long pointed out the many features that the iPhone has been missing to date like copy and paste, bluetooth headphones, MMS, 3G in the beginning, etc. yet that hasn't stopped the iPhone/iPod Touch from selling 40 million units in 2 years. Even if competitors adopt the ARM9 earlier Apple can still differentiate themselves on usability, style, and with a 45nm ARM8 shrink, better battery life than ARM9 phones.
What's more, there is plenty of power in existing architectures for Apple to exploit if needed. John Cormack has estimated that the ARM11+MBX in existing iPhones/iPod Touches can see a 2 times increase in speed for games if Apple only optimized their software. Even if 2 times is a bit much, it isn't hard to believe that there is quite a bit of potential to be exploited in either ARM11 or ARM8. Initial ARM9 implementations aren't likely to be the most efficient from either a software perspective or a hardware SoC design perspective, so a highly optimized design using the mature ARM8 won't necessarily be left in the dust. Especially, if Apple uses a 45nm shrink and clocks the ARM8 higher to close the gap.
Optimization isn't that easy. Carmack can say whatever he wants to about that, but he's very narrow minded in his outlook. Apple has done what they have with the OS in order to bring a much richer experience than just gaming to the table. They have to have though several year out with this. I imagine that they know where they're going, and that optimizations won't help them, as they need to add more features, which may depend on how the OS is now written.
This is why hardware is needed to complete the circle.
I don't understand why you and a few others would doubt the hardware path that seems pretty obvious. We can only assume that Apple's future path is to continue to strain the hardware as much as possible in order to give a better experience.
I agree with Tim here, and if he's correct, the more processing power, the better:
http://www.pcmag.com/article2/0,2817,2348658,00.asp
I'm wondering if one of us misread the bit about 10-20%. Because I read that as 10-20% WORST than A8. From other data I've seen I'd expect that to be true.
Dave
But it's only worse when the device is running flat out. It's more efficient when doing most other tasks. It's like Intel's new Nehalem line-up. They're not like Prescott of course, but their max power envelop isn't low. but the idle power draw is very low, and draw during average use is low. Overall, the average is lower than the previous generation while delivering much more computing power. More bang for the watt.
The same is true for the 9 vs the 8. Peak power usage is somewhat more, but overall, it's less.
From everything we've read so far, it does seem that the "s" will be more than a bit faster at many things. One of those things is rendering a web page. Downloading the page isn't the only aspect to the speed. If it takes as long as, or even longer to render a complex page, then everything is slow.
I guess I'm just thinking it's a matter of focus. Using the ARM A9 in itself probably won't cause people to jump to buy an iPhone. However, a focus on usability, portability, battery life, integration with your computer, etc. will more likely win people over regardless over whether an older ARM A8 CPU is used as long as it's not drastically slower. Of course, it doesn't have to be either or. The best case scenario will be killer software with bleeding edge hardware.
I know it's been stated earlier that the manufacturers are selling their new phone on the higher performance processing junk inside their phones, but its not true.
The manufacturers don't like taking about what's inside the phones at all. They want to talk about the features that customers can interact with. What's inside simply makes that possible. There are far fewer processors for smartphones than there are processors for computers, so manufacturers know that getting into that kind of battle is a losing one, as most of them use the same chips, or similar enough ones so as not to matter.
I don't agree with Carmack. It's not likely that he knows as much about the OS as does the team that designed it. He's got his priorities and Apple has theirs.
The iPhone OS is still pretty young, Apple has more important things to worry about than the nebulous concept of optimizing.
For the next couple of years at least, it will be raw power that they are concerned with, along with chip optimizations that will likely allow them to do in hardware what they now do in software, which will speed things up considerably.
I agree and would also add that the feature people want most is multitasking. Even with the power and battery savings of the 3gs, Apple still thought that it couldn't handle it well enough. It would only stand to reason they think the A9 with twice the computing power and better battery life can. Especially when you consider PA Semi could get to add their own influence to increase battery life and performance. Chances are Apple will get a very early look at these chips because of their long relationship with ARM.
While others have said that it would be too soon of a move, Apple is going to have to make it sooner or later and there will be people upset by it. Apple has always shown that they will adopt new tech, even if it means leaving others behind. My iMac that I bought last year is not even on the supported card list for H.264 acceleration in Snow Leopard. As the Mac moves to Open CL, it would figure that they would do the same with the iPhone. This move to Open CL and the A9 will be one that lasts for awhile.
Contrary to what other think, I believe the 3gs is a stopgap move. The power right now is only being used for voice control, video, and to make apps load faster. It seems like it is the final product of this generation of hardware. Even the model name suggests that. Developers are not going to make apps that utilizes it's power anyway because they won't eliminate older iPhones and because Open CL will probably make it's appearance next year.
I don't understand why you insist on using the word "stopgap". It doesn't make sense.
Every advance is a "stopgap" until the next one comes out, which is in itself a stopgap before the next.
It loses all meaning in that context.
The speed increases in the 3Gs is likely about the same as we will see with the 9 when it hopefully comes out next year. Which is to say, about 2+ times faster than the precious generation.
Those are both large advances in speed.
This enables video recording and editing, voice control, not just voice dialing, video MMS. The new compass has enabled much better, true turn by turn programs. Real 3D games. I could go on, but you know the new possibilities that the new hardware has enabled.
The 3G was the stopgap, if you insist in using that word. There was NO advance in any area of the phone hardware at all, except for the addition of 3G and GPS.
You have to explain how that minor advance wasn't the stopgap, and the 3Gs isn't the major advance.
Is it because, for some reason, you're not happy with what they've been able to do?
I don't understand why you insist on using the word "stopgap". It doesn't make sense.
Every advance is a "stopgap" until the next one comes out, which is in itself a stopgap before the next.
It loses all meaning in that context.
The speed increases in the 3Gs is likely about the same as we will see with the 9 when it hopefully comes out next year. Which is to say, about 2+ times faster than the precious generation.
Those are both large advances in speed.
This enables video recording and editing, voice control, not just voice dialing, video MMS. The new compass has enabled much better, true turn by turn programs. Real 3D games. I could go on, but you know the new possibilities that the new hardware has enabled.
The 3G was the stopgap, if you insist in using that word. There was NO advance in any area of the phone hardware at all, except for the addition of 3G and GPS.
You have to explain how that minor advance wasn't the stopgap, and the 3Gs isn't the major advance.
Is it because, for some reason, you're not happy with what they've been able to do?
If the word "stopgap" is too strong of a word for you, then I apologize. I'm not saying that the 3gs is not a powerful device. Nor am I saying it did not bring any features. But if you look in terms of a larger context it is clear that is not looking towards this iPhone as the one that signals the next-gen of iPhone hardware. This phone does not have any influence from PA Semi or the numerous graphics people they hired away from ATI. That alone should say that they are aiming for something more. Open CL is right around the corner and will soon move to it.
Don't say 3D games. I can count on one hand the amount of developers that will make the maximum use of this hardware at best. Developers are not going to throw away the potential cash cow of developing for all the past iPhone/ iPod touch users for the 3gs. The only way that will happen is when forces a break. As I said, has still not allowed multitasking even though the 3gs has better battery life and a far better processor. If they released only a minor upgrade with another A8 next year, there would still be no multitasking. With a dual processors and much better battery life in the A9, they would have it. It's hard for me to see that would go another year without multitasking.
You may say that the 3g is a stopgap and that is probably true. But I'm not looking at last year. I'm looking at today and what will soon be coming tomorrow. The hardware alone will be a significant release because it will be more powerful and have more battery life. This is almost like 's move to Intel's dual cores.
I'm quite fine with what 's done. They have as powerful a phone as any out there. I'd like to see multitasking and an API for integration between apps but it has some nice features that almost make me forget about them. I'm just being honest about it. Hell, I may even buy one. But I also know what is coming on the horizon. If anyone wanted to buy this phone, I'd say go for it.
If the word "stopgap" is too strong of a word for you, then I apologize. I'm not saying that the 3gs is not a powerful device. Nor am I saying it did not bring any features. But if you look in terms of a larger context it is clear that is not looking towards this iPhone as the one that signals the next-gen of iPhone hardware. This phone does not have any influence from PA Semi or the numerous graphics people they hired away from ATI. That alone should say that they are aiming for something more. Open CL is right around the corner and will soon move to it.
I understand what you're saying, but I don't know what that has to do with it. This is a much improved machine. It's not in any holding pattern, waiting for some miracle processor from Apple. It's an evolutionary product, and as such, it's done very well.
Next years' will be better.
Don't say 3D games. I can count on one hand the amount of developers that will make the maximum use of this hardware at best. Developers are not going to throw away the potential cash cow of developing for all the past iPhone/ iPod touch users for the 3gs. The only way that will happen is when forces a break. As I said, has still not allowed multitasking even though the 3gs has better battery life and a far better processor. If they released only a minor upgrade with another A8 next year, there would still be no multitasking. With a dual processors and much better battery life in the A9, they would have it. It's hard for me to see that would go another year without multitasking.
You don't know that. I've already read about the 3D games that we'll be seeing, though not by name. It only takes a handful of cutting edge games to move the products to a new level. That's true of any gaming platform.
With this upgrade, Apple likely has enough oomp for multiprocessing if they want it. likely, they don't want it yet, because they want to come out with some framework that will enable it to be better than elsewhere. They have their priorities, as they had said about C/C/Paste last year.
You may say that the 3g is a stopgap and that is probably true. But I'm not looking at last year. I'm looking at today and what will soon be coming tomorrow. The hardware alone will be a significant release because it will be more powerful and have more battery life. This is almost like 's move to Intel's dual cores.
Every new generation will be much faster and have more hardware features until the situation matures, when advances will slow down. We won't get double the processing power forever, perhaps for one or two more generations.
I'm quite fine with what 's done. They have as powerful a phone as any out there. I'd like to see multitasking and an API for integration between apps but it has some nice features that almost make me forget about them. I'm just being honest about it. Hell, I may even buy one. But I also know what is coming on the horizon. If anyone wanted to buy this phone, I'd say go for it.
You can wait forever. I waited for two features. Two features I said I would wait for since we had out first discussions right after Apple announced the phone. That was 3G and supported native apps. GPS was a bonus I wanted, but wouldn't wait for.
The new 3Gs is an amazing machine. Much more so than the 3G, esp with the new OS.
Next year, you will hear of the possibility of even more amazing things with it, are you going to wait again? I don't recall now, did you say you have an older iPhone, or not? If so, then waiting makes more sense. We will likely wait, as we bought ours in late August, though my daughter will get the new one when she goes to school in London this summer (good excuse for her!
If not, then unless you're waiting for a contract to end it just wastes another year.
I can't wait .
9
I understand what you're saying, but I don't know what that has to do with it. This is a much improved machine. It's not in any holding pattern, waiting for some miracle processor from Apple. It's an evolutionary product, and as such, it's done very well.
Next years' will be better.
You don't know that. I've already read about the 3D games that we'll be seeing, though not by name. It only takes a handful of cutting edge games to move the products to a new level. That's true of any gaming platform.
With this upgrade, Apple likely has enough oomp for multiprocessing if they want it. likely, they don't want it yet, because they want to come out with some framework that will enable it to be better than elsewhere. They have their priorities, as they had said about C/C/Paste last year.
Every new generation will be much faster and have more hardware features until the situation matures, when advances will slow down. We won't get double the processing power forever, perhaps for one or two more generations.
You can wait forever. I waited for two features. Two features I said I would wait for since we had out first discussions right after Apple announced the phone. That was 3G and supported native apps. GPS was a bonus I wanted, but wouldn't wait for.
The new 3Gs is an amazing machine. Much more so than the 3G, esp with the new OS.
Next year, you will hear of the possibility of even more amazing things with it, are you going to wait again? I don't recall now, did you say you have an older iPhone, or not? If so, then waiting makes more sense. We will likely wait, as we bought ours in late August, though my daughter will get the new one when she goes to school in London this summer (good excuse for her!
If not, then unless you're waiting for a contract to end it just wastes another year.
I think you're missing several points on the 3D games thing. First off, the developers would be alienating 40 million consumers in favor of 1-2 million. But, some developers will do it to simply put out a 3GS "only" app. It will be a novel hit since all the 3GS users will want to show off their phone. However there are even more roadblocks to such apps.
Money is a large issue when it comes to developing advanced games as the ones you're talking about. Games like that take almost a year to make (see PSP and DS games) and once they're made they cost $30-40. That's a huge jump over what we have in the app store now. We're so used to $5-15 games that $30-40 seems rather steep.
So here's what you have working against you when creating a 3GS app:
1. Development costs are much higher
2. Development time is much longer, new iPhone by the time the best ones come out?
3. Your price in the app store is much higher
4. Your consumers base is much, much smaller (1/40th the size!)
So here's what you have working against you when creating a 3GS app:
1. Development costs are much higher
2. Development time is much longer, new iPhone by the time the best ones come out?
3. Your price in the app store is much higher
4. Your consumers base is much, much smaller (1/40th the size!)
I don?t agree with any of that.
1. It?s still using the same Xcode.
2. It may be longer, but only because the developer has more options available to them. It may also be shorter as the developer no longer has to spend extra time shoehorning the app into older hardware to compete with newer apps that are utilizing the latest HW.
3.Still using the same Xcode. There are potential reasons why it could could more but nothing that states it will higher specifically for coding for the new HW. In fact, it?ll be cheaper than coding for both the new and old.
4. How you get 1/40 the size? Is that because there are 40M iPhone OS X devices out there now? how do you get one million as of today or as of one month from now. How about 6 months from now or one year from now when this new HW is the standard for future-forward app development?