I think the fact that not only does the Samsung's smaller chip NOT perform better, but actually performed 28% worse is the big story. Test 1000 phones, you'll see that the Samsung chip still performs contrary to what we would expect. Even if the chips did equally, the smaller chip would be a disappointment.
The difference is significant if CPU is under load, and that was repeated and verified by at least 4 groups.
Being a frequent mobile gamer, I would pick TSMC version any time of the day.
For non-gamers (with what Apple identifies as a "normal use") indeed it would not make a difference.
Basically the Samsung processor uses about 30% more power when it is under heavy load. Generally speaking, the processors in the phones spend most of their time at very low loads, so unless you're actively taxing the CPU for hours on end (e.g. by running benchmarking software), I'm not sure if you're likely to see a difference. Games might be a case where you do actually see a difference, especially in games that tax the CPU in addition to the GPU.
Dont be silly. There are small variations in all products.
I bet if they tested 2 TSMC phones they would test differently
Don't be silly, if they tested 2 TSMC chips and they performed differently, TSMC would be out of business. Six sigma originated in the electronics business.
So, the Lyrium app designer is forced to disable chip ID for these new phones IMMEDIATELY.
And we have a 28% difference showing. Has Apple ever looked more shifty? They are looking to be outdoing themselves on this one, which takes some doing!
So, the Lyrium app designer is forced to disable chip ID for these new phones IMMEDIATELY.
And we have a 28% difference showing. Has Apple ever looked more shifty? They are looking to be outdoing themselves on this one, which takes some doing!
I think Geekbench 3.4 (App Store) can ID the chip.
Then they will lose because both chips perform within spec.
Still wouldn't be good as Apple would have to reveal a whole lot of their testing methodology and results. Apple would want to avoid it getting to that point.
The way I look at it the Samsung chip is at Spec. It meets Apple's requirement for efficency for the A9.
The TSMC chip over performs Apple's requirement. What do you want Apple to do? Make the TSMC chip less efficent? Ridiculous.
What part of the post is wrong? :rolleyes: (Answer: None of it)
Doesn't matter one whit how you (or I or anyone else here) look at it. If a class action would happen to be certified it would only matter how the court looked at it, while considering the evidence that Apple would be ordered to begrudgingly submit during discovery. I can assure you they won't task you with sorting it, nor would Apple be happy with revealing it IMHO.
Strange that you raise the issue of whether or not the chips perform to specs and/or efficiency then use that to claim I was emphatically "WRONG" about something I wasn't discussing.
"As Anandtech explains, any chip manufacturing process—even within the same fab—will result in a range of performance, just as not every cookie baked from the same recipe will be exactly the same. By defining minimum tolerances, though, a customer like Apple can establish an acceptable range of performance. The chips that don’t live up to those minimums don’t make it into the devices. Those that do can still perform differently."
“This can only be reliably tested on a device, but if all else were equal, I could see single percentage points difference between Samsung and TSMC,” says Moorhead. Even testing on a single device, though, wouldn’t be enough, since a single system on chip doesn’t speak for all of them. You would need to test many, many devices to be sure of any significant differences.
This is true, we used to test each chip from a complete wafer, the die location on the wafer would be noted and then we would test them all. It was not unusually to have performance differences across the wafer, most time certain location on the wafer would yeild vastly better parts, knowing this you could cherry pick the fastest and slowest parts. Intel does this, they map the wafers to know which parts perform the best then mark them as higher speed or grade part and sell them for a lot more. Since Apple is not selling their chips and they do not make phones which run at different clock rates like the typical PC it would not surprise me to see some phones run better than others. Apple most likely is not cherry picking their processsor or only shipping processor would closely look the same as each other. I bet you could find a group of phones which behave completely different than another group.
So, the Lyrium app designer is forced to disable chip ID for these new phones IMMEDIATELY.
And we have a 28% difference showing. Has Apple ever looked more shifty? They are looking to be outdoing themselves on this one, which takes some doing!
They promise performance X, delivered X,. With TSMC you get X and a bit more, that's it, not conspiracy or drama.
You're fracking ignored because of your utter lack of logic from now on.
I agree with you that software can cause a lot of energy use. Last night I plugged the phone and went to bed. In the midnight I woke and checked the phone. It felt pretty warm. This morning I unplugged the phone. In one hour the battery went down to less than 80%. I checked the apps. They are the same apps I have been using all the time. I closed a few apps. The battery seems still going down fast. I fixed this by rebooting. It seems the problem is not the app. The problem is some process owned by iOS 9. Apple quality is getting worse. After I upgraded to iOS 9, the phone works worse here and there.
Try opening up the settings and check the battery usage stats. There's now a little clock icon that will show you how much everything has used by percentage and time both foreground and background over 24 hours or 7 days. Odds are somebody like FaceBook is being a bad little app while it's tucked away in the background. When I'm tight on power access, say when wandering around visiting a new city, I throw apps overboard rather than leave them lurk: a double click of the home button or that left side 3d Touch swipe (for S series phones) and I start flipping apps off to the top to shut them down.
OR you got a lemon and should replace it as is your right unless it's an old iPhone 3 or something since you didn't specify.
If you tested 1000 TSMC chips against 1000 Samsung chips, you will still see that the TSMC chips consistently score faster because of the extremely tight tolerances allotted to chip production.
Try opening up the settings and check the battery usage stats. There's now a little clock icon that will show you how much everything has used by percentage and time both foreground and background over 24 hours or 7 days. Odds are somebody like FaceBook is being a bad little app while it's tucked away in the background. When I'm tight on power access, say when wandering around visiting a new city, I throw apps overboard rather than leave them lurk: a double click of the home button or that left side 3d Touch swipe (for S series phones) and I start flipping apps off to the top to shut them down.
OR you got a lemon and should replace it as is your right unless it's an old iPhone 3 or something since you didn't specify.
I looked. I did not spend too much time looking for the app. I rebooted. After reboot, the phone seemed returned to normal.
The phone is not a lemon, It is a free replacement I got from Apple store of a three year old iPhone 5. That phone got a problem. When I press the home button it does not return to the home. Since the phone has a wake button replacement program. I thought it may be due to the wake button. Why not give it a try. I made an appointment with Apple store. The genius took a look. He showed me that the problem is not what I thought at all. The screen actually bulged a little near the center. He said this is due to the batter underneath. He replaced it free.
I don't know why Apple Keep giving Samsung a chance to screw up apple's. Seriously, if i am samsung, i would totally love to see how apple get into troubles for selling the new iPhones . If i am samsung, I would announce this to the press: "YES, it's true, apple buy cheap chips from us, because they wanted to make more money". "We leave the best for the S7.babba"
Don't buy a thing from your most against enemy, unless you are dumb. Cheaper is not always better.
Comments
The difference is significant if CPU is under load, and that was repeated and verified by at least 4 groups.
Being a frequent mobile gamer, I would pick TSMC version any time of the day.
For non-gamers (with what Apple identifies as a "normal use") indeed it would not make a difference.
Basically the Samsung processor uses about 30% more power when it is under heavy load. Generally speaking, the processors in the phones spend most of their time at very low loads, so unless you're actively taxing the CPU for hours on end (e.g. by running benchmarking software), I'm not sure if you're likely to see a difference. Games might be a case where you do actually see a difference, especially in games that tax the CPU in addition to the GPU.
Get over it.
Dont be silly. There are small variations in all products.
I bet if they tested 2 TSMC phones they would test differently
Don't be silly, if they tested 2 TSMC chips and they performed differently, TSMC would be out of business. Six sigma originated in the electronics business.
We're doomed I tell you. Dooooomed.
Then they will lose because both chips perform within spec.
And we have a 28% difference showing. Has Apple ever looked more shifty? They are looking to be outdoing themselves on this one, which takes some doing!
Doesn't matter one whit how you (or I or anyone else here) look at it. If a class action would happen to be certified it would only matter how the court looked at it, while considering the evidence that Apple would be ordered to begrudgingly submit during discovery. I can assure you they won't task you with sorting it, nor would Apple be happy with revealing it IMHO.
Strange that you raise the issue of whether or not the chips perform to specs and/or efficiency then use that to claim I was emphatically "WRONG" about something I wasn't discussing.
Stop making a mountain out of a molehill.
"As Anandtech explains, any chip manufacturing process—even within the same fab—will result in a range of performance, just as not every cookie baked from the same recipe will be exactly the same. By defining minimum tolerances, though, a customer like Apple can establish an acceptable range of performance. The chips that don’t live up to those minimums don’t make it into the devices. Those that do can still perform differently."
“This can only be reliably tested on a device, but if all else were equal, I could see single percentage points difference between Samsung and TSMC,” says Moorhead. Even testing on a single device, though, wouldn’t be enough, since a single system on chip doesn’t speak for all of them. You would need to test many, many devices to be sure of any significant differences.
Fortunately, someone has done just that: Apple.
http://www.wired.com/2015/10/iphone-6s-a9-battery-life/
This is true, we used to test each chip from a complete wafer, the die location on the wafer would be noted and then we would test them all. It was not unusually to have performance differences across the wafer, most time certain location on the wafer would yeild vastly better parts, knowing this you could cherry pick the fastest and slowest parts. Intel does this, they map the wafers to know which parts perform the best then mark them as higher speed or grade part and sell them for a lot more. Since Apple is not selling their chips and they do not make phones which run at different clock rates like the typical PC it would not surprise me to see some phones run better than others. Apple most likely is not cherry picking their processsor or only shipping processor would closely look the same as each other. I bet you could find a group of phones which behave completely different than another group.
So, the Lyrium app designer is forced to disable chip ID for these new phones IMMEDIATELY.
And we have a 28% difference showing. Has Apple ever looked more shifty? They are looking to be outdoing themselves on this one, which takes some doing!
They promise performance X, delivered X,. With TSMC you get X and a bit more, that's it, not conspiracy or drama.
You're fracking ignored because of your utter lack of logic from now on.
You still have an iPhone, get over it, Apple is always in the right, no matter what.Its a non issue yet again!
I agree with you that software can cause a lot of energy use. Last night I plugged the phone and went to bed. In the midnight I woke and checked the phone. It felt pretty warm. This morning I unplugged the phone. In one hour the battery went down to less than 80%. I checked the apps. They are the same apps I have been using all the time. I closed a few apps. The battery seems still going down fast. I fixed this by rebooting. It seems the problem is not the app. The problem is some process owned by iOS 9. Apple quality is getting worse. After I upgraded to iOS 9, the phone works worse here and there.
Try opening up the settings and check the battery usage stats. There's now a little clock icon that will show you how much everything has used by percentage and time both foreground and background over 24 hours or 7 days. Odds are somebody like FaceBook is being a bad little app while it's tucked away in the background. When I'm tight on power access, say when wandering around visiting a new city, I throw apps overboard rather than leave them lurk: a double click of the home button or that left side 3d Touch swipe (for S series phones) and I start flipping apps off to the top to shut them down.
OR you got a lemon and should replace it as is your right unless it's an old iPhone 3 or something since you didn't specify.
TSMC's version of the A9 is manufactured using a 16-nanometer FinFET process, whereas Samsung depends on a 14-nanometer technique.
Neither uses TSMC 16nm nor uses Samsung 14nm. That are just names for the process node.
Think of them as Charles and George and then ask yourself:
Is the Charles process faster than the George process because the name starts with a C?
Or is the George process more efficient than the Charles process because the name includes two 'e's instead of only one?
If you tested 1000 TSMC chips against 1000 Samsung chips, you will still see that the TSMC chips consistently score faster because of the extremely tight tolerances allotted to chip production.
This has been a great advertisement for TSMC.
Try opening up the settings and check the battery usage stats. There's now a little clock icon that will show you how much everything has used by percentage and time both foreground and background over 24 hours or 7 days. Odds are somebody like FaceBook is being a bad little app while it's tucked away in the background. When I'm tight on power access, say when wandering around visiting a new city, I throw apps overboard rather than leave them lurk: a double click of the home button or that left side 3d Touch swipe (for S series phones) and I start flipping apps off to the top to shut them down.
OR you got a lemon and should replace it as is your right unless it's an old iPhone 3 or something since you didn't specify.
I looked. I did not spend too much time looking for the app. I rebooted. After reboot, the phone seemed returned to normal.
The phone is not a lemon, It is a free replacement I got from Apple store of a three year old iPhone 5. That phone got a problem. When I press the home button it does not return to the home. Since the phone has a wake button replacement program. I thought it may be due to the wake button. Why not give it a try. I made an appointment with Apple store. The genius took a look. He showed me that the problem is not what I thought at all. The screen actually bulged a little near the center. He said this is due to the batter underneath. He replaced it free.
Seriously, if i am samsung, i would totally love to see how apple get into troubles for selling the new iPhones .
If i am samsung, I would announce this to the press: "YES, it's true, apple buy cheap chips from us, because they wanted to make more money". "We leave the best for the S7.babba"
Don't buy a thing from your most against enemy, unless you are dumb.
Cheaper is not always better.