Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

My real world experience with TSMC vs Samsung

roeiz said:
thanks. i really appreciate the efforts to calm down various anxieties here.
i will be on the 4K shooter and HD footage handling user spectrum,
so my reasons to feel it is unfair are more founded.
i won't check my chip (i hope) when i get the phone.
the only thing that will really ease my mind is seeing a test when shooting 4K and stuff.
if that will show 5-10% difference, i'll feel much more relaxed.
if this test will show 20-30%, i will feel very pissed.
and i don't live in US, so the many "easy" replacement options are non existent for me.
Click to expand...

if a difference of 20-30% will be demonstrated in real world situations I'll be with you on this matter, mate

Ries said:
It's been replicated by several different people and tech sites. There is a difference and loading the cores at 30% will give you 2 hours less battery life. No one has been able to get a samsung phone to run same length as TSMC in that test. In doesn't matter what does it in the phone, it will happen on samsung vs TSMC phone.
Click to expand...

please, link me any other controlled test about it ...

This is the net: one unknown dude post BS on YouTube, 100 blogs report it as "a fact" ...

Dented said:
I don't thing at this point anyone else needs to run the geekbench test, it's very well established that the Samsung phones do considerably worse on that test, it's not a question of the sample size or statistical error, it's an eminently repeatable result.

What is still very unclear, as I and others have said before, is what actual bearing that result has on anything other than geekbench.

None of the other tests or situations people are trying to emulate indicates anything like that kind of difference between the chips. With a few exceptions (and history tells us there are always people with terrible iPhone battery life out there, whatever the iteration) the majority of Samsung users actually using their phones seem happy. The "real world" data, including the results of other benchmarks, is all a lot closer to Apple's 2-3% than Geekbench's 20-30%. So the question is, is geekbench revealing a real problem that will effect anyone, or is it simply revealing a flaw in the test that Geekbench has devised, as a measure of real world battery life?
Click to expand...

That's the point.
Before say something, I'm waiting for a more structured test about the matter.
Ars tecnica demonstrated that only Geekbench is showing a difference, so far.
It could be the single benchmark.
 
Max(IT) said:
And you analyzed how many post exactly ? 10 ? 100 ? 1000 ?
With 20 MILLIONS iPhone sold your "statistically significant" analysis is just trash.


Who said there is a better phone ?

reported for personal attack

unfortunately not .....

you are right, and did a very good analysis, but whiners here aren't going to read a post like this.
Most of them don't even understand what's you are speaking about ...
Click to expand...

Personal attack? You can't be serious?
Some people.
 
Max(IT) said:
again "controlled test on YouTube" .... lol ....

Did you at least read the Ars Technica article ? read at least the title "Apple was right" ....
Click to expand...
The 20-30% difference has been proven conclusively in several controlled tests so far. Just because one or more groups use YouTube to show the controlled tests and results doesn't preclude "proper controls". Please make intellectually honest cases, Max, and not "red herring" arguments.

And to be clear - the 20-30% difference has only been proven on 6S.. NOT ON 6S PLUS!! Those of you returning your phones for battery problems on 6S Plus are living in fantasy land - and just wanting to jump on the Chipgate bandwagon.
 
  • Like
Reactions: hammy434 and roeiz
NewdestinyX said:
The 20-30% difference has been proven conclusively in several controlled tests so far. Just because one or more groups use YouTube to show the controlled tests and results doesn't preclude "proper controls". Please make intellectually honest cases, Max, and not "red herring" arguments.

And to be clear - the 20-30% difference has only been proven on 6S.. NOT ON 6S PLUS!! Those of you returning your phones for battery problems on 6S Plus are living in fantasy land - and just wanting to jump on the Chipgate bandwagon.
Click to expand...

Oh, didn't realize that! good to hear.
 
NewdestinyX said:
The 20-30% difference has been proven conclusively in several controlled tests so far. Just because one or more groups use YouTube to show the controlled tests and results doesn't preclude "proper controls". Please make intellectually honest cases, Max, and not "red herring" arguments.

And to be clear - the 20-30% difference has only been proven on 6S.. NOT ON 6S PLUS!! Those of you returning your phones for battery problems on 6S Plus are living in fantasy land - and just wanting to jump on the Chipgate bandwagon.
Click to expand...
No... YouTube is the LAST place where I'm going to look for controlled test.
And, again, 20% was "proved" (not really) on a single synthetic test.

Speaking about "intellectually honest cases" .... :rolleyes:
 
Max(IT) said:
No... YouTube is the LAST place where I'm going to look for controlled test.
And, again, 20% was "proved" (not really) on a single synthetic test.
Click to expand...
This is a complete fallacy. Willfully overlooking many tests by huge respected 3rd party testing groups.
 
Max(IT) said:
On a side note I'm always astonished on how people here could believe Apple is lying.
They are a big multinational company, they are under the magnifier every day, they have shareholders to report, they have customers to keep loyal.... A company like that could avoid comments, could make press conferences to appear in a good light, but they can't just lie. It would be a suicide
Click to expand...
In unsure if you're following the Volkswagen global scandal, but I thought the same thing. How on earth did they think lying was a viable option?

I think asking questions is always a good thing. A company may not be specifically lying. It may be something that legitimately wasn't tested in specific way (this is any aspect of a product, not specifically limited to a SOC, like we are speaking of here). It's a big reason why recalls exist; companies find out about flaws in a product once it gets into the hands of the public.

Again, I don't see evidence in this specific case that this is an issue, but asking questions is always something I condone. At the same time, if we are to make a claim against a company, we should have reasonable evidence against them. The reason I even posted in this thread is because I don't consider the evidence the OP presented to be reasonable at all. Just one man's opinion, of course.
 
lordofthereef said:
In unsure if you're following the Volkswagen global scandal, but I thought the same thing. How on earth did they think lying was a viable option?

I think asking questions is always a good thing. A company may not be specifically lying. It may be something that legitimately wasn't tested in specific way (this is any aspect of a product, not specifically limited to a SOC, like we are speaking of here). It's a big reason why recalls exist; companies find out about flaws in a product once it gets into the hands of the public.

Again, I don't see evidence in this specific case that this is an issue, but asking questions is always something I condone. At the same time, if we are to make a claim against a company, we should have reasonable evidence against them. The reason I even posted in this thread is because I don't consider the evidence the OP presented to be reasonable at all. Just one man's opinion, of course.
Click to expand...
When a big company lie, they are going to pay ...
VW is a good example.
Apple's executives know that, and I tend to believe they may twist things a little bit, but they won't lie...

So far I didn't se any evidence. Just ONE benchmark, and only that, on a ridiculously little number of units (TWO !).
 
  • Like
Reactions: realeric
  • Like
Reactions: Max(IT)
hammy434 said:
look at TSMC and Samsung on geekbench battery website.

TSMC: https://browser.primatelabs.com/battery3/search?dir=desc&q=n71map&sort=score
Samsung: https://browser.primatelabs.com/battery3/search?dir=desc&q=n71ap&sort=score

thats more than just a single test, and it shows there is a consistent difference, at least on geekbench. and apart from one tsmc anomaly, tsmc always wins.
Click to expand...
Yes except that they are all results from the same, single benchmarking app which gives this kind of margin. No other test performed by anyone else shows anything like this sort of deviation between the chips. Tom's Hardware ran their own standardised testing today (which doesn't include geekbench) and gave the efficiency advantage to Samsung - that's how far left field the geekbench scores are.

I totally get that the geekbench scores are repeatable, and have been repeated many times. That still doesn't mean they're right..
 
Dented said:
Yes except that they are all results from the same, single benchmarking app which gives this kind of margin. No other test performed by anyone else shows anything like this sort of deviation between the chips. Tom's Hardware ran their own standardised testing today (which doesn't include geekbench) and gave the efficiency advantage to Samsung - that's how far left field the geekbench scores are.

I totally get that the geekbench scores are repeatable, and have been repeated many times. That still doesn't mean they're right..
Click to expand...

true, but theres youtube video
which also suggests the tsmc lasts longer.

although that tomshardware article is interesting. if more evidence like that comes through, which i hope it does as id rather not worry about this, then we can hopefully relax. hopefully software updates will help too.
 
Max(IT) said:
When a big company lie, they are going to pay ...
VW is a good example.
Apple's executives know that, and I tend to believe they may twist things a little bit, but they won't lie...

So far I didn't se any evidence. Just ONE benchmark, and only that, on a ridiculously little number of units (TWO !).
Click to expand...
You kind of took 90% of what I said and threw it out the window with that response lol. Oh well.
 
Dented said:
I don't thing at this point anyone else needs to run the geekbench test, it's very well established that the Samsung phones do considerably worse on that test, it's not a question of the sample size or statistical error, it's an eminently repeatable result.

What is still very unclear, as I and others have said before, is what actual bearing that result has on anything other than geekbench.

None of the other tests or situations people are trying to emulate indicates anything like that kind of difference between the chips. With a few exceptions (and history tells us there are always people with terrible iPhone battery life out there, whatever the iteration) the majority of Samsung users actually using their phones seem happy. The "real world" data, including the results of other benchmarks, is all a lot closer to Apple's 2-3% than Geekbench's 20-30%. So the question is, is geekbench revealing a real problem that will effect anyone, or is it simply revealing a flaw in the test that Geekbench has devised, as a measure of real world battery life?
Click to expand...

Do you or anyone knows that how many Samsung chips are used? Apple used Samsung after they ran out of TSMC right? If that's the case Apple might not have used lot of Samsung chips because Samsung was just covering up for the extra demand. Since now the people are concerned is Apple still using Samsung chips to meet the demand?
 
Max(IT) said:
No... YouTube is the LAST place where I'm going to look for controlled test.
And, again, 20% was "proved" (not really) on a single synthetic test.

Speaking about "intellectually honest cases" .... :rolleyes:
Click to expand...

Max(IT), you just pretending? there are many many, not only two.
Are you an IT people? why You can't google, I know, from your avatar. But macrumors is where you are around. The geekbench result is much more than 20%. OP's delta is 25% (4710/3778). Another one from RonFromOregon 6s tested a delta of 51%(3448/2276). So burst your bubble..
 
Zune55 said:
Do you or anyone knows that how many Samsung chips are used? Apple used Samsung after they ran out of TSMC right? If that's the case Apple might not have used lot of Samsung chips because Samsung was just covering up for the extra demand. Since now the people are concerned is Apple still using Samsung chips to meet the demand?
Click to expand...

Nobody outside of Apple really knows any of that kind of stuff, but it's believed that Samsung are making (or at least selling) their chips for a lower price, so it's in apple's interest to use more of them rather than less.
 
hammy434 said:
look at TSMC and Samsung on geekbench battery website.

TSMC: https://browser.primatelabs.com/battery3/search?dir=desc&q=n71map&sort=score
Samsung: https://browser.primatelabs.com/battery3/search?dir=desc&q=n71ap&sort=score

thats more than just a single test, and it shows there is a consistent difference, at least on geekbench. and apart from one tsmc anomaly, tsmc always wins.
Click to expand...

Prince134 said:
Max(IT), you just pretending? there are many many, not only two.
Are you an IT people? why You can't google, I know, from your avatar. But macrumors is where you are around. The geekbench result is much more than 20%. OP's delta is 25% (4710/3778). Another one from RonFromOregon 6s tested a delta of 51%(3448/2276). So burst your bubble..
Click to expand...

Tom's Hardware just demonstrated the contrary. Their Samsung unit is better than the TSMC ....

Stop using YouTube as a way to demonstrate something....
 
realeric said:
Tom's Hardware test was wrong. They should have monitored CPU throttling while doing the battery test.
Click to expand...
What was that word a lot of TSMC owners were using against those with Samsung chips a couple of days ago? Starts with DE and ends in NIAL?
 
Max(IT) said:
Tom's Hardware just demonstrated the contrary. Their Samsung unit is better than the TSMC ....

Stop using YouTube as a way to demonstrate something....
Click to expand...

Toms hardware is the only one where samsung is better. Also they didn't test geekbench battery on that.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back