Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Not justifying the cost does not necessarily mean being in debt if you buy it.
That only seems to apply to Americans. Other parts of the world don't live on debts like the USA does. That's why creditcards aren't popular all around the world, only in some parts.
 
  • Like
Reactions: tryrtryrtryrt
Can't tell if you're joking, but those aren't comparable tests. I've already listed the comparable results.
No you haven't. You've compared iGPU only to dGPU inclusive trying to mislead people into your narrative. It's apples to oranges. I've compared dGPU inclusive to dGPU inclusive. It's apples to apples.
So my point stands and yours - does not.
 
Last edited by a moderator:
Then your comparison is far more misleading because the dGPU can actually consume an awful lot of power reducing the battery life to something like 1.5 to 2 hours. It highly depends on how much it has to compute. The iGPU is not as powerful but far more energy efficient. The problem with all this is that both of them are in the 15" MBP and macOS will be switching between them quite often. If you want to compare then you need to do so with only the iGPU enabled and only the dGPU enabled. Out of the box the latter can be done, the former requires the use of 3rd party software.

There is way too much context lost in translation here due to only posting screenshots of a very small part of a chart which makes it completely meaningless.

Not to mention that difference in software versions isn't even taken into account at all!

Anyway, to give the screenshot some context here is a quote with some important things highlighted in bold:
In our light Wi-Fi browsing test (which only uses the integrated GPU) with the screen set to a brightness of 200 nits, it actually lasts for around 10 hours, besting the old model by half an hour. In the heavier WebGL test (which only uses the dedicated GPU), both models last for about three hours, but the 2014 GeForce-equipped MacBook lasts a bit longer. The more powerful AMD GPU is also a bit more power-hungry, it would seem.

We did try to run the WebGL test on the integrated GPU using GFXCardStatus to see what the difference was, but the results were anomalously low so we haven’t reported them here. Just know that the entry-level, Iris-only model may behave slightly differently from the ones we’ve reviewed here.

Don't just look at pictures and screenshot them, read the text that explains them too.
 
Last edited:
No you haven't. You've compared iGPU only to dGPU inclusive trying to mislead people into your narrative. It's apples to oranges. I've compared dGPU inclusive to dGPU inclusive. It's apples to apples.
So my point stands and yours - does not.
But, again, congratulations on the 13th successful attempt!
Read the text, not only the chart, which is mislabeled. You have misunderstood. Yet again.
 
Read the text, not only the chart, which is mislabeled. You have misunderstood. Yet again.
Trust a review where text differs from the graphs? :D Thank you. I'll pass.
[doublepost=1492898310][/doublepost]
Then your comparison is far more misleading because the dGPU can actually consume an awful lot of power reducing the battery life to something like 1.5 to 2 hours. It highly depends on how much it has to compute. The iGPU is not as powerful but far more energy efficient. The problem with all this is that both of them are in the 15" MBP and macOS will be switching between them quite often. If you want to compare then you need to do so with only the iGPU enabled and only the dGPU enabled. Out of the box the latter can be done, the former requires the use of 3rd party software.

There is way too much context lost in translation here due to only posting screenshots of a very small part of a chart which makes it completely meaningless.

Not to mention that difference in software versions isn't even taken into account at all!
No, my comparison is less misleading because it's apples to apples. Actually it's not misleading at all.
However it does not show what it's meant to. And there is no apples to apples comparison showing the necessary thing.
So what can be stated is that 15" '16 battery runtime is 1. enough, 2. similar to '15. @JustinRP37, @profmatt and others agree I believe.
 
  • Like
Reactions: profmjh
Trust a review where text differs from the graphs? :D Thank you. I'll pass.
Then read the other one instead. They actually compare the 2016 to the 2015 model and aren't finding much of a difference in the power it draws or the battery life (which is a bit better on the 2016...15 minutes more).

No, my comparison is less misleading because it's apples to apples.
There is no context and you are purposely leaving out the context from the reviews. By doing so you are misleading people. The fact that you do not speak of the second review that was linked speaks volume. That review actually compares it to the previous model that also came with the dGPU and thus is an apples to apples comparison. Yet that review is very specific about the battery life being slightly (15 min) better than the previous model.

@JustinRP37, @profmatt and others agree I believe.
It is not about agreeing, it is about basing things on facts and evidence which is exactly what the linked reviews do. What did you not get from all the scientists demonstrating worldwide today? If you want proper reviews then do proper research by doing actual measurements and providing context along them.
 
Anyway, to give the screenshot some context here is a quote with some important things highlighted in bold:

Don't just look at pictures and screenshot them, read the text that explains them too.
To understand the context: there are following tests presented
1. '16 with dGPU allowed but not forced so when there is something resource intensive it's being calculated on dGPU in short time and less energy spent overall,
2. '15 with dGPU forcibly disabled using GFXCardStatus so when there is something resource intensive instead of calculating it on dGPU in short time and spending less energy overall it's calculated on iGPU in more time with more energy spent overall even though this '15 can do the same '16 does in point 1,
3. '16 with dGPU forcibly enabled using Apple system settings so it always spends more energy.

So 1 is much beneficial compared to 2 and no surprise it shows better results. 2 is beneficial to 3 and no surprise it shows better results.
Overall? This review cannot be used to support @Sanpete's claim since it does not compare apples to apples. It only shows that the battery runtime is enough which I, @JustinRP37 and @profmatt showed earlier.
[doublepost=1492899346][/doublepost]
Then read the other one instead. They actually compare the 2016 to the 2015 model and aren't finding much of a difference in the power it draws or the battery life (which is a bit better on the 2016...15 minutes more).
Yes, this is what I and @JustinRP37 have been stating for the last pages.
There is no context and you are purposely leaving out the context from the reviews. By doing so you are misleading people. The fact that you do not speak of the second review that was linked speaks volume. That review actually compares it to the previous model that also came with the dGPU and thus is an apples to apples comparison. Yet that review is very specific about the battery life being slightly (15 min) better than the previous model.
No, @Sanpete is misleading people by providing bogus sources and not explaining the misleading in them. I just laugh his claims off without misleading anyone. I show screenshots for laughs (have a look - I don't claim anything on the battery runtime of 15" '16 when I provide screenshots - my claims are in different posts) and instead for several pages in a row state that battery runtime is the same as of '15.
It is not about agreeing, it is about basing things on facts and evidence which is exactly what the linked reviews do. What did you not get from all the scientists demonstrating worldwide today? If you want proper reviews then do proper research by doing actual measurements and providing context along them.
I did. I even made @Sanpete provide the sources. Had you not failed me (a joke :p) I've made him make proper analysis as well, we were quite nearing it. Science, *****! :p
 
Last edited:
Have you actually read the review at all? If you did then read my previous reply and tell us what you've noticed ;)
Tell us we're dying to know! (No we're not.)
Is it connected to the fact that you misattributed something to the review and I've agreed with your conclusions saying nothing about the fact I've actually read the same thing in the review. Are you, by chance, misleading people? Or just joking? :p
[doublepost=1492899821][/doublepost]
You have made my point yet again. There is no room for reasonable doubt about this, the text spells it out very plainly.
Yes, the fact there is an error in the review is spelled out quite plainly. :p
 
Last edited:
Nope. Those who've read the review spot it immediately, it isn't that difficult to spot. The funny thing is, you are repeating it.
 
To understand the context: there are following tests presented
1. '16 with dGPU allowed but not forced so when there is something resource intensive it's being calculated on dGPU in short time and less energy spent overall,
2. '15 with dGPU forcibly disabled using GFXCardStatus so when there is something resource intensive instead of calculating it on dGPU in short time and spending less energy overall it's calculated on iGPU in more time with more energy spent overall even though this '15 can do the same '16 does in point 1,
3. '16 with dGPU forcibly enabled using Apple system settings so it always spends more energy.

So 1 is much beneficial compared to 2 and no surprise it shows better results. 2 is beneficial to 3 and no surprise it shows better results.
Overall? This review cannot be used to support @Sanpete's claim since it does not compare apples to apples. It only shows that the battery runtime is enough which I, @JustinRP37 and @profmatt showed earlier.
[doublepost=1492899346][/doublepost]Yes, this is what I and @JustinRP37 have been stating for the last pages. No, @Sanpete is misleading people by providing bogus sources and not explaining the misleading in them. I just laugh his claims off without misleading anyone. I show screenshots for laughs (have a look - I don't claim anything on the battery runtime of 15" '16 when I provide screenshots - my claims are in different posts) and instead for several pages in a row state that battery runtime is the same as of '15. I did. I even made @Sanpete to provide the sources. Had you not failed me (a joke :p) I've made him make proper analysis as well, we were quite nearing it.

You just made up what you present as test 2. The Ars Technica review of the 2015 doesn't say anything about forcing anything. It only notes that the test only engages the iGPU, which is normal for internet browsing with Safari. (I have 25 tabs open, none using the dGPU.) It's the same test as what you call test 1.
 
Last edited by a moderator:
You just made up what you present as test 2. The Ars Technica review of the 2015 doesn't say anything about forcing anything. It only notes that the test only engages the iGPU, which is normal for internet browsing with Safari. (I have 25 tabs open, none using the dGPU.) It's the same test as what you call test 1.
No, I haven't. '16 review states this fact "I ran the Wi-Fi test twice—once with switchable graphics enabled and once with the dedicated GPU on all the time.". '15 review states these facts "In our light Wi-Fi browsing test (which only uses the integrated GPU)" and "We did try to run the WebGL test on the integrated GPU using GFXCardStatus". From these facts we can conclude that
a) the webpages might have used autoswitching since otherwise is not stated and webpages can use it indeed;
b) they differentiate between "switchable graphics enabled" and "which only uses the integrated GPU" by using different phrases; proving otherwise is necessary;
c) they used GFXCardStatus to force iGPU to run another test with similar phrase "on the integrated GPU" so we can imply this is what they've meant by saying "only uses the integrated GPU"; proving otherwise is necessary, again.
 
Last edited by a moderator:
No, I haven't. '16 review states this fact "I ran the Wi-Fi test twice—once with switchable graphics enabled and once with the dedicated GPU on all the time.". '15 review states these facts "In our light Wi-Fi browsing test (which only uses the integrated GPU)" and "We did try to run the WebGL test on the integrated GPU using GFXCardStatus". From these facts we can conclude that
a) the webpages might have used autoswitching since otherwise is not stated and webpages can use it indeed;
b) they differentiate between "switchable graphics enabled" and "which only uses the integrated GPU" by using different phrases; proving otherwise is necessary;
c) they used GFXCardStatus to force iGPU to run another test with similar phrase "on the integrated GPU" so we can imply this is what they've meant by saying "which only uses the integrated GPU"; proving otherwise is necessary, again.
Science, *****!We need some analogue of Godwin's law for Trump. Like "he who mentions Trump first loses". It's getting out of hand lately.
You're really making hash of something that's quite clear and simple. To add to the confusion, you claim using only the iGPU decreases performance, the opposite of what Ars Technica found. Whatever. Those who can read will understand the tests just fine. I accept that you have a complicated view that mixes unrelated stuff together and arrives at an interpretation more fitting your desires.
 
No, I haven't. '16 review states this fact "I ran the Wi-Fi test twice—once with switchable graphics enabled and once with the dedicated GPU on all the time.". '15 review states these facts "In our light Wi-Fi browsing test (which only uses the integrated GPU)" and "We did try to run the WebGL test on the integrated GPU using GFXCardStatus". From these facts we can conclude that

8<
You purposely leave out information here. I'll quote the entire paragraph:

In our light Wi-Fi browsing test (which only uses the integrated GPU) with the screen set to a brightness of 200 nits, it actually lasts for around 10 hours, besting the old model by half an hour. In the heavier WebGL test (which only uses the dedicated GPU), both models last for about three hours, but the 2014 GeForce-equipped MacBook lasts a bit longer. The more powerful AMD GPU is also a bit more power-hungry, it would seem.

The other paragraph that follows the previous one I also quote in full because you've left out information there too:
We did try to run the WebGL test on the integrated GPU using GFXCardStatus to see what the difference was, but the results were anomalously low so we haven’t reported them here. Just know that the entry-level, Iris-only model may behave slightly differently from the ones we’ve reviewed here.

Nowhere in that review you can find the sentence "I ran the Wi-Fi test twice—once with switchable graphics enabled and once with the dedicated GPU on all the time." (it finds 0 results when you type in I ra; the Arstechnica review about the 2016 model even zeros out at I r). Thus you cannot draw the conclusion you are drawing because you are basing it on information that doesn't exist.

And for those thinking "hey, isn't that sentence in the other review?": no it isn't, that review also zeros out at I r.
 
You're really making hash of something that's quite clear and simple. To add to the confusion, you claim using only the iGPU decreases performance, the opposite of what Ars Technica found. Whatever. Those who can read will understand the tests just fine. I accept that you have a complicated view that mixes unrelated stuff together and arrives at an interpretation more fitting your desires.
They say quite the contrary.
And yes, those who can read already understand that the battery runtime of '16 is similar to '15. You're beating a dead horse.
[doublepost=1492902215][/doublepost]
You purposely leave out information here. I'll quote the entire paragraph:



The other paragraph that follows the previous one I also quote in full because you've left out information there too:


Nowhere in that review you can find the sentence "I ran the Wi-Fi test twice—once with switchable graphics enabled and once with the dedicated GPU on all the time." (it finds 0 results when you type in I ra; the Arstechnica review about the 2016 model even zeros out at I r). Thus you cannot draw the conclusion you are drawing because you are basing it on information that doesn't exist.

And for those thinking "hey, isn't that sentence in the other review?": no it isn't, that review also zeros out at I r.
Let's start with "Nowhere in that review you can find the sentence "I ran the Wi-Fi test twice—once with switchable graphics enabled and once with the dedicated GPU on all the time."
http://bfy.tw/BNPX
Снимок экрана 2017-04-23 в 0.02.43.png


After this the rest is not worth discussing before you apologise.

You could not find it because it's on the 4th page of the review and the pages are being loaded dynamically while you scroll so of course browser Cmd+F won't find it until you scrolled down. Instead of asking me politely where this quote is you've accused me of lying while being wrong (and not for the first time). This says a lot about you.

P. S. My points still stand.
 
Last edited:
The iGPU is a perfectly capable GPU which can be used for gaming as well. It just isn't as powerful as the dGPUs from AMD or Nvidia which is why Apple and many others are using them in their machines. In case of Apple they are using both with the OS switching between them. Throughout the years they have tweaked that switching mechanism in macOS because at first it would switch to the dGPU the moment an app was using a specific API. It hasn't been doing that for some years now due to the finetuning. One of the reasons why software versions matter.

Battery runtime isn't just a matter of GPU, it is impacted by all the system components. The review simply highlights iGPU vs dGPU because people want to know the difference between the two when it comes to power consumption/battery life. It is not a good way of comparing the old with the new MBP.

Anyway, the second review is the wrong one to link to because it is a comparison of the high end 15" MBP 2016 compared to the low end version of the same model (and this is a big big hint for the question I asked earlier). Somewhere in that review there is a link to the review of the low end version. If you lookup the battery runtime in that review you'll find the following observation from the reviewer:
Even though the battery capacity was reduced by almost 25 % compared to the predecessor, Apple still managed to improve the overall runtimes with the 76-Wh battery. Our WiFi test determines an advantage of 130 minutes, so little more than two hours.
The reviewer is observing an increase of 2 hrs compared to the previous model (the 2015 one). One can see the exact numbers in the table below that observation. Those numbers are also in the table in the high end model review.

One thing the both of you need to take into account is the fact that these are from a pre-set workload that doesn't represent real life usage of a machine. In real life the load of the system changes constantly and with modern hardware being in idle all of the time and throttling up when necessary it makes it very complex to test this properly (the estimate the OS gives you is now completely useless because of that). I for one have never seen the same values are came anywhere near the values of any review of the notebook I owned and that makes perfectly sense to me. That doesn't mean their measurements are wrong, their data doesn't exist, etc., it only means I'm seeing different things and that can be explained with the fact that my usage is different. Respect what the reviewers are seeing and try to understand what's behind it. That way you can also question the reviewers methodology (because measuring is one thing, doing it properly is another; people do make mistakes in how they measure something).

You could not find it because it's on the 4th page of the review and the pages are being loaded dynamically while you scroll so of course browser Cmd+F won't find it until you scrolled down.
I couldn't find it because you are leaving out information. You left out the part that you are talking of a completely different review than any of the ones linked here (one to an Arstechnica review of the 2015 model and one from notebookchk for the 2016 model...so what do you think people will look for when you speak of "the '16 review"?).

Next time link to the information and stop leaving out things. And as explained above neither your nor Sanpete's point stand.
 
Last edited:
1. it wasn't mentioned by Sanpete
2. neither did you, you simply mentioned "the '16 review" so we still have to guess which review you mean.

Like I said, you need to be more specific about what you mean, stop leaving out information and stop being so darn lazy for not even naming it properly.
 
Last edited by a moderator:
as explained above neither your nor Sanpete's point stand.
Can't tell with much confidence what point you're disagreeing with, but if it's that tests show the 2016 15" has better battery life than the 2015 for light to moderate use, what you say doesn't give any good reason to think otherwise. The browsing and video tests, while more controlled than ordinary use, are applicable to ordinary use. The test results for the 2016 fit with what I observed in my own test and what other post here.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.