they are a mystery to everyone, because GB is THE ONLY test showing a significative difference between the two chips.
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.
According to a PostDoc's paper, TSMC is much better than SAMSUNG 6s' battery efficiency
- Thread starter Prince134
- Start date
- Sort by reaction score
Uh, the Samsung runs cooler than tsmc? Still way better testing than just running geek bench.
Samsung for the win.
Look at the apple logo color, the article cheated, Apple paid for you? You guess ha ha
Nothing you just said makes any sense FYI.
The Apple logo is plastic and not a good heat conductor, the rest of the case is metal, it's perfectly possible for it to have still been cool especially if the rest of the case wasn't massively hot either. There may be internal differences to the layout between the two main boards that contribute to this.
If they were going to cheat the test, why be running a fan or anything like that, why not simply take the picture when the phone isn't doing very much? If they were running a fan, why would that have such a dramatically greater effect on the plastic anyway - in short how does a fan explain anything?
Finally - they chose to publish that picture to the entire world, on their site, which lives or dies by the credibility of their hardware testing - do you honestly think they would do that if it's such the giveaway you've decided it is?
Looks like there's another thread to discuss about Tom's Hardware test,
https://forums.macrumors.com/thread...is-slightly-better-overall-lol.1928358/unread
https://forums.macrumors.com/thread...is-slightly-better-overall-lol.1928358/unread
oh this is still going 
well why not, here goes
okay so people are still using the geekbench battery benchmark
remember every single other bench mark shows them to be nearly identical, video's have shown them to perform the same in real life, in anandtech's test the samsung was a tiny bit ahead. in tomshardware's testing the samsung was a bit ahead. all other results can be put down to manufacturing variances if you accept Apple's 2-3% difference between the two chips
so from this we can conclude the geekbench battery test is an anomaly but it is a constant anomaly so now we have to look at this benchmark specifically, so we go look at the chart for the test over here
https://browser.primatelabs.com/battery-benchmarks
there are strange results all over the thing. the Galaxy S6 is a glaring one but even if you want to only look at the iPhones look at the difference between the 6 Plus and the 6S Plus, remember what Apple was boasting about. smaller battery, same battery life. yeah right everyone was saying. notice the result for the 6S Plus is 30%+ better then then the 6 Plus. considering every review and reading the battery life threads on here (TSMC or Samsung) the majority compare their battery life as about the same as last generation, this is strange. surely someone would have noticed a difference that big.
if we check Rene Ritchie's review he has the TSMC chip and claims the same battery life as last gen, Leo Laporte and Marques Brownlee have the Samsung chip and also claim their battery life is the same as last gen, reading through the battery life threads here there are a couple of people who have reported up to a hour less, an hour more with both chips but most have described theirs as the same with both chips
there is no indication of anyone getting 2 hours screen on time more the the 6 Plus that the geek bench battery test shows. none at all, combining that with previous evidence it provides a very strong argument the geek bench battery test doesn't in any way reflect real world battery life and may even be flawed.
of course I'm sure someone is going to mention gaming again at some point so let me include the 1st post in the thread below
https://forums.macrumors.com/threads/iphone-6s-plus-battery-performance.1926699/
there is someone with the 6S Plus and the apparently better chip, one and a half hours of intensive gaming has resulted in 17 hours of battery life on a 6S Plus!. for the sake of your battery's long term health I suggest just plugging the phone in if you want to play hours of games
this turned out longer then I thought, sorry
well why not, here goes
okay so people are still using the geekbench battery benchmark
remember every single other bench mark shows them to be nearly identical, video's have shown them to perform the same in real life, in anandtech's test the samsung was a tiny bit ahead. in tomshardware's testing the samsung was a bit ahead. all other results can be put down to manufacturing variances if you accept Apple's 2-3% difference between the two chips
so from this we can conclude the geekbench battery test is an anomaly but it is a constant anomaly so now we have to look at this benchmark specifically, so we go look at the chart for the test over here
https://browser.primatelabs.com/battery-benchmarks
there are strange results all over the thing. the Galaxy S6 is a glaring one but even if you want to only look at the iPhones look at the difference between the 6 Plus and the 6S Plus, remember what Apple was boasting about. smaller battery, same battery life. yeah right everyone was saying. notice the result for the 6S Plus is 30%+ better then then the 6 Plus. considering every review and reading the battery life threads on here (TSMC or Samsung) the majority compare their battery life as about the same as last generation, this is strange. surely someone would have noticed a difference that big.
if we check Rene Ritchie's review he has the TSMC chip and claims the same battery life as last gen, Leo Laporte and Marques Brownlee have the Samsung chip and also claim their battery life is the same as last gen, reading through the battery life threads here there are a couple of people who have reported up to a hour less, an hour more with both chips but most have described theirs as the same with both chips
there is no indication of anyone getting 2 hours screen on time more the the 6 Plus that the geek bench battery test shows. none at all, combining that with previous evidence it provides a very strong argument the geek bench battery test doesn't in any way reflect real world battery life and may even be flawed.
of course I'm sure someone is going to mention gaming again at some point so let me include the 1st post in the thread below
https://forums.macrumors.com/threads/iphone-6s-plus-battery-performance.1926699/
there is someone with the 6S Plus and the apparently better chip, one and a half hours of intensive gaming has resulted in 17 hours of battery life on a 6S Plus!. for the sake of your battery's long term health I suggest just plugging the phone in if you want to play hours of games
this turned out longer then I thought, sorry
Last edited:
oh this is still going
well why not, here goes
okay so people are still using the geekbench battery benchmark
remember every single other bench mark shows them to be nearly identical, video's have shown them to perform the same in real life, in anandtech's test the samsung was a tiny bit ahead. in tomshardware's testing the samsung was a bit ahead. all other results can be put down to manufacturing variances if you accept Apple's 2-3% difference between the two chips
so from this we can conclude the geekbench battery test is an anomaly but it is a constant anomaly so now we have to look at this benchmark specifically, so we go look at the chart for the test over here
https://browser.primatelabs.com/battery-benchmarks
there are strange results all over the thing. the Galaxy S6 is a glaring one but even if you want to only look at the iPhones look at the difference between the 6 Plus and the 6S Plus, remember what Apple was boasting about. smaller battery, same battery life. yeah right everyone was saying. notice the result for the 6S Plus is 30%+ better then then the 6 Plus. considering every review and reading the battery life threads on here (TSMC or Samsung) the majority compare their battery life as about the same as last generation, this is strange. surely someone would have noticed a difference that big.
if we check Rene Ritchie's review he has the TSMC chip and claims the same battery life as last gen, Leo Laporte and Marques Brownlee have the Samsung chip and also claim their battery life is the same as last gen, reading through the battery life threads here there are a couple of people who have reported up to a hour less, an hour more with both chips but most have described theirs as the same with both chips
there is no indication of anyone getting 2 hours screen on time more the the 6 Plus that the geek bench battery test shows. none at all, combining that with previous evidence it provides a very strong argument the geek bench battery test doesn't in any way reflect real world battery life and may even be flawed.
of course I'm sure someone is going to mention gaming again at some point so let me include the 1st post in the thread below
https://forums.macrumors.com/threads/iphone-6s-plus-battery-performance.1926699/
there is someone with the 6S Plus and the apparently better chip, one and a half hours of intensive gaming has resulted in 17 hours of battery life on a 6S Plus!. for the sake of your battery's long term health I suggest just plugging the phone in if you want to play hours of games
this turned out longer then I thought, sorry
Yeah what's the matter with you, bringing logic, reason and critical analysis to a geektest benchmark thread. It's posts like your's that could ruin this whole none sense topic.
Simple, let me help you out. Just re-read the entirety of all posts on all threads pertaining to topic. By then your brain will be so fried, you won't care.help!!! i don't know what to do!
my 6S plus is stuck on 100% battery for 4 hours...
i have the %^$&# chip..
Yeah what's the matter with you, bringing logic, reason and critical analysis to a geektest benchmark thread. It's posts like your's that could ruin this whole none sense topic.
oh I don't know, maybe this is why I don't get invited to parties much
honestly if I jumped to conclusions like some people on here based on all the evidence I presented earlier my science, economics, statistics and business studies teachers would have all simultaneously thrown me out of their classes.
sorry Miss the vast majority of the evidence suggests there is no difference but this one result is different and even though there is evidence to suggest it might be flawed I'm going with that one because uh numbers!
not to mention I suspect Apple likely have some of the best lawyers in the world scrutinising everything they do and no way in heck would let them release a statement saying there is no difference if there was due to the legal implications that would lead to and the fact that obviously it would be tested, but I guess we needed a controversy.
it would also be nice if this was all in one thread instead of however many there is now..
Newest rumors state that there significant differences (layout, utiziliation etc.) between two A9 variants, not just manufacturing. We'll see how this will play out.
you do realised that further shows there isn't a difference outside manufacturing variance, I'm guessing you didn't read it?
the first test is the geekbench test, as expected it shows the TSMC with a 20%+ lead which by now we are pretty used to
the second test is a video playback test, that shows a 8% difference
the third test is a CPU intensive web browsing test designed to focus on the CPU power consumption which resulted in only a 7% difference
now remember a 2-7% difference even between even identical chips is common, ananadtech's testing was in favour of the TSMC by about 4%, tomshardware's testing has around a 10% advantage for the Samsung
the videos you linked excluding the geekbench battery test show around a 8% advantage again for the TSMC
heise.de shows about a 4% advantage for the Samsung chip
that other video I linked a while ago showed a 1% difference between the two
between all these tests all of the results fall within normal manufacturing variance except the geekbench test which remains a huge anomaly, also remember these are all benchmarks run to exaggerate CPU power consumption. in normal real world usage when you are using wifi and the GPU and the camera Apple's claim of 2-3% difference sounds to be very accurate, good find though.
Last edited:
you do realised that further shows there isn't a difference outside manufacturing variance, I'm guessing you didn't read it?
the first test is the geekbench test, as expected it shows the TSMC with a 20%+ lead which by now we are pretty used to
the second test is a video playback test, that shows a 8% difference
the third test is a CPU intensive web browsing test designed to focus on the CPU power consumption which resulted in only a 7% difference
now remember a 2-7% difference even between identical chips is common, ananadtech's testing was in favour of the TSMC by about 4%, tomshardware's testing has around a 10% advantage for the Samsung
the videos you linked excluding the geekbench battery test show around a 8% advantage again for the TSMC
heise.de shows about a 4% advantage for the Samsung chip
that other video I linked a while ago showed a 1% difference between the two
between all these tests all of the results fall within normal manufacturing variance except the geekbench test which remains a huge anomaly, also remember these are all benchmarks run to exaggerate CPU power consumption. in normal real world usage when you are using wifi and the GPU and the camera Apple's claim of 2-3% difference sounds to be very accurate, good find though.
You are right. There has been lots of smoke, but so far Geekbench is only fire for sure.
I'm fairly certain 99.9% of the public won't care at all
, it's a very small chance whoever you sell it to unless it's on here will have heard of chipgateYou are right. There has been lots of smoke, but so far Geekbench is only fire for sure.
it's a shame that primate labs do not seem to want offer any kind of insight into this, they seem happy to respond to anything except questions about the result itself :/
Unless you're selling it on MacRumors to a small percentage of people then I would expect that your prospective buyers won't even enquire as to the make of chip inside.
That's a relief lol my battery is pretty awesome so I don't care that it's Samsung
Don't give places like Gazelle any ideas. I can see it now;I'm fairly certain 99.9% of the public won't care at all
it's a shame that primate labs do not seem to want offer any kind of insight into this, they seem happy to respond to anything except questions about the result itself :/
Carrier,
Size,
Color,
Samsung or TSMC chip ($100 less for Samsung chips)
LOL!
Don't give places like Gazelle any ideas. I can see it now;
Carrier,
Size,
Color,
Samsung or TSMC chip
LOL!
I can totally see that happening!! Lol
So as a former pc builder enthusiast this cpu variance has been on my mind being that this is my first iPhone ever. I actually waited in line at 4am on launch day to buy one(coming from a note 3). They only had a 6s+ 16gb available when I purchased, had it for a week, returned it and purchased 6s 128gb. Had that for a week returned it a got a 6s+ 64gb(Samsung A9). Battery life seemed acceptable...but my quest to compare my usage did not cease there. Finally landed a SG S-F128gb 6s+ (what I originally wanted) and its TSMC.
I ran real world test all last night on a 64gb Samsung and a 128gb TSMC and my suspicion was vindicated. In every real world test I ran (nfl mobile streaming, safari browsing, video playback, GAMING) I found there to be a consistant difference.
Gaming for 22 minutes starting at 100%= 95% vs 92%
NFL Mobile streaming, safari, video playback, various apps for 90 minutes starting at 100% = 86% vs 80%
I wanted to do this because I noticed a huge variance from the brightness levels when compared from around the web. People are only testing at full brightness, other than geek bench which tests at min brightness. For my test I let the auto brightness set the levels about 35% and then turned off auto brightness.
Apple said 2-3%.. what they didn't say was that was per 10% at this rate I'm seeing a 20-30% difference in real world testing w/35% brightness.. More to come
I ran real world test all last night on a 64gb Samsung and a 128gb TSMC and my suspicion was vindicated. In every real world test I ran (nfl mobile streaming, safari browsing, video playback, GAMING) I found there to be a consistant difference.
Gaming for 22 minutes starting at 100%= 95% vs 92%
NFL Mobile streaming, safari, video playback, various apps for 90 minutes starting at 100% = 86% vs 80%
I wanted to do this because I noticed a huge variance from the brightness levels when compared from around the web. People are only testing at full brightness, other than geek bench which tests at min brightness. For my test I let the auto brightness set the levels about 35% and then turned off auto brightness.
Apple said 2-3%.. what they didn't say was that was per 10% at this rate I'm seeing a 20-30% difference in real world testing w/35% brightness.. More to come
so you standardised brightness settings but then ran just a bunch of random apps and reached a conclusion from that?, your testing doesn't seem particularly controlled. for results to be valid you need a controlled test in the first place. did you play the game or leave it in the same scenario?, that would affect the results. which websites you went on would have contributed to the result. which apps you used and if they were the same, streaming quality and so on
Last edited:
Every single click or screen rotate was in sync with one another, just browsing the web in general like a real world user would. Whatever the default video streams would be for the apps that ran them.Like I said everything was done identically. The gaming was done on each independently, I played the game for 22 mins. The game was Breakneck.so you standardised brightness settings but then ran just a bunch of random apps and reached a conclusion from that?, your testing doesn't seem particularly controlled. for results to be valid you need a controlled test in the first place. did you play the game or leave it in the same scenario?, that would affect the results. which websites you went on would have contributed to the result. streaming quality and so on