Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Seeds iOS 5.0.1 Beta 2 to Developers
- Thread starter MacRumors
- Start date
- Sort by reaction score
Well there are two things you need to figure on mobile hardware design. One is that the screen is nearly always, by far, the most power hungry component. Turning the backlight off can save you 50% of the power consumption on most systems. Behind that, the RAM, any storage, wireless and then things like CPUs (in that order) are what starts to reduce your battery life.
Apple have done an incredible job in basically turning your phone off completely when you are not using it; when the screen is black and you have to hit the top button to get to the slide-to-unlock display, you are waking it up from full standby where nearly everything is off. That's how you get decent battery life on phones, make sure they are not turned on all the time, as with everything running even with decent screen management they'd all barely last 2 hours of runtime.
The second thing is that clocking things down aggressively and turning things off is not always the best solution for situations where you need to DO something. While it's sitting on a desk, nothing on the network locking wireless, screen off, power management is great, but otherwise you are nearly always better off in a "race to idle" scenario. This is where you simply run as fast as you can to get the work off the table so that you can go back to a state where you can actively power manage. Since you can run at 800MHz and get "800 million things done a second" (not realistic, but bear with me) doing 100 million things will take 1/8th of a second. At 400MHz it will take 1/4 of a second. At 166MHz it will take 2/3 of a second. It is usually much better to run at full power for 0.125 seconds than at 20% of the speed (and probably only 50% of the power consumption of full) for 0.7 seconds, or incur the time taken to switch.
So what Apple tend to do (and what most mobile manufacturers, CPU schedulers and power management libraries do as well) is keep the CPU running as fast as possible for as long a time as is necessary plus a little bit more. The moment you take an input, you ramp up the CPU, and until you are dead sure that the user is idle, you stay there. This works on mobile devices without impacting power management too much because as above the CPU is by far the least impact on the battery life.
That, and the Cortex-A9 class dual-core CPUs (Apple A5) are more efficient than the Cortex-A8 (Apple A4) when one core is idle and not greatly more power hungry when both cores are active at the same clock speed. They are also more efficient from an instruction scheduling point of view. So bumping from an A4 to an A5 from iPhone 4 to iPhone 4S wouldn't actually make a great deal of difference. Graphics cores are the same way.
For wireless, Bluetooth 2.1 has a cute feature in that it can support something like 24Mbps data transfers compared to older standards which can only do around 10Mbps. The same race to idle scenario applies here; it only powers the transmitter to send data (reception is tantamount to passive) so to send 24Mbit of data in 1 second uses less power than sending 24Mbit of data at a lower speed over more than 2 seconds. Holding that transmitter powered is what gives you the better power consumption even though actually sending at 24Mbps using the EDR rate uses more power at any sampled point, it is active for less time when you add it all together.
The 4S also does have a smaller battery, which is supposed to be made up for by more efficient components..
All in all, it is more likely that Apple simply are not using a more aggressive power saving strategy when they can on certain parts of the iPhone 4S that are brand new to the 4S compared to the 4. Probably the new GSM/CDMA chip (putting it into an idle state instead of a nap or sleep state, this is my personal opinion of the culprit based on the talk time numbers for the 4 vs. 4S - they are too significantly different to explain it being a battery size issue, and while on the phone with it to your ear, the proximity sensor has allowed the screen and much of everything else to be turned off, and all it needs to do is route the audio to the GSM chip, so the GSM chip must be using a ton of extra power in comparison. Driving two antennas, and boosting power for better signal would account for it)
.. or Bluetooth 4.0 (keeping the Low Energy link active accidentally) or even using NAND with a different page size and having to do more aggressive or time consuming bad block handling or page buffering.
... or something else. But it's a lot more complicated than "there are two CPUs instead of one so it will take more power". Dual core (say, a Core i5) laptops these days use a lot less power than single core ones (say, a Pentium M) at much higher clocks.
OK, will report back when I've recalibrated. Still, last time after I installed Beta v1 it only stayed on 1% for about five mins whilst being used. That was after 2hrs use and 18hrs standby. I honestly think my battery is just faulty, but we'll see how it looks after a recharge from 0%.
By the way, took a shower, cost me 1% battery life (aka standby for about 15 mins cost me 1%).
Here's a whole thread on it (one of many here and elsewhere).
https://forums.macrumors.com/threads/1254178/
And a list I posted on a previous thread, I'm sure there are more...
It's not just that: the iPad version has been changed by-design and certain features removed. It's way beyond just bugs, unfortunately.
Michael
I really wish I could get more than ~1.5 Mbps with AT&T in St. Louis.
Old 3GS phone? 1.5 Mbps. New 4S phone? 1.5 Mbps. It's silly for AT&T to request a "4G" symbol when they know it's both not 4G, and their network is not that fast in all locations.
If anything, I'd rather have an option to disable 3G. I'm always on WiFi, and would happily take EDGE for better battery life.
That would be nice. In Columbus, we get about the same. However, when I was in Orlando a couple months ago for training, I got 6Mbps down on my 4! That's faster than alot of spots for LTE around here. I can only imagine what the 4S would get down there!
"open in..." not fixed
Anyone noticed that since the iOS 5.0 GM seed the "open in..."-option only allows about 500mb to be transfered to another app? However, it's still not working for me on 5.0.1 beta2. I really hope this isn't a silent restriction by Apple, but a bug... Orherwise, opening a movie or another bigger file has to be done via itunes again - a step back in my oppinion.
Anyone noticed that since the iOS 5.0 GM seed the "open in..."-option only allows about 500mb to be transfered to another app? However, it's still not working for me on 5.0.1 beta2. I really hope this isn't a silent restriction by Apple, but a bug... Orherwise, opening a movie or another bigger file has to be done via itunes again - a step back in my oppinion.
Hey quit whining about your down/up. 
In the big office complex in Hollywood, CA, I just got 0.47 down and 0.05 up!
On second attempt I got 0.90 down, 0.62 up.
ATT still has work to do in high-traffic areas at lunch/dinner time...
Edit: At least on 3rd attempt, I got 3.32 down...
In the big office complex in Hollywood, CA, I just got 0.47 down and 0.05 up!
On second attempt I got 0.90 down, 0.62 up.
ATT still has work to do in high-traffic areas at lunch/dinner time...
Edit: At least on 3rd attempt, I got 3.32 down...
By any chance did Siri type this for you?
Gotta watch those recall rates!
Michael
Dammit Siri!!!!
Code:
Unable to Verify Update
iOS 5.0.1 beta failed verification because you are
no longer connected to the Internet.I am browsing the Internet and listening to Pandora with no problems. What might cause the above issue?
Hey quit whining about your down/up.
In the big office complex in Hollywood, CA, I just got 0.47 down and 0.05 up!
On second attempt I got 0.90 down, 0.62 up.
ATT still has work to do in high-traffic areas at lunch/dinner time...
Edit: At least on 3rd attempt, I got 3.32 down...
Remember that you live in a county where there are nearly 10 million people.
A lot of good points in your post. This one caught my eye though b/c when I turn off the cellular radio...my battery life increases three-fold. If I sit at the office all day in airplane mode with all other radios on, I drop 1% - 2% an hour. With it on...it goes up to about 6% an hour. That might just be the nature of the beast but it does make a difference.
So I think I have a theory that explains the size of the update for the 4S.
The 4S uses a new CPU type (A5) as opposed to other versions of the iPhone, the massive size of the package may indicate that they made a change to the compiler to take advantage of the properties of the CPU.
This would mean that all binaries were rebuilt with the optimizations, and therefore even a small change would require a full package.
The 4S uses a new CPU type (A5) as opposed to other versions of the iPhone, the massive size of the package may indicate that they made a change to the compiler to take advantage of the properties of the CPU.
This would mean that all binaries were rebuilt with the optimizations, and therefore even a small change would require a full package.
Sounds legit. I just installed 5.0.1 beta 1 onto my iPhone without any issues and now I'm OTA updating to beta 2. I never noticed too much of an issue with battery life, but wasn't there a rumor about them improving data speeds on Sprint?
Right, but it's a county that's bigger than a few small states. In addition, the population density in Hollywood isn't really that high.
Buildings are (mostly) capped at about 20 floors for earthquake purposes, and I still get bad reception. I think for some reason LA just needs better equipment in place, or a few more towers.
Buildings are (mostly) capped at about 20 floors for earthquake purposes, and I still get bad reception. I think for some reason LA just needs better equipment in place, or a few more towers.
I'm having problems with compilations, worked fine under iOS 4 and screwed up on iOS 5. There's also a page on the Apple discussion forums with 25k views about the iOS 5 music app problems:
https://discussions.apple.com/thread/3375697?start=0&tstart=0
Really hope this is fixed soon, might have to go back to iOS 4 with my iPod Touch!