MacBook Air 13. I5 or i7?

[ezekielrage_99]

Have you downloaded Intel Power Gadget like w00d suggested and actually OBSERVED how the CPU reacts versus just speculating? I did it a week ago and have spent quite a bit of my time looking at how my 2012 i7 MBA performs when doing "everyday" tasks.

The 2012 i7 has a base speed of 2.0GHz, with a turbo of 3.0 in dual core and 3.2 in single core. With all background apps and processes killed, it sits there at 0.8GHz to 1.2GHz. If you launch Safari, it immediately jumps to 3.0GHz for a second and then evens back out at 1.2GHz or so. Start moving the cursor around and it goes to 2.0 to 2.4. Grab a slider on the side of the window, move the window up and down and it again shoots up to 3.0 - 3.1 GHz. Kill Safari and it immediately drops back down to under 2.0GHz and eventually settles down to 0.8GHz.

Similarly, launch iTunes. Again, screams up to 3GHz upon launching and updating Genius, then again settles out. Scroll your albums and it jumps up to 3.0 again. Play a song and it levels back out to 1.0 to 1.2

This is played out time and time and time again. The only app I didn't see this high level of processor speed is when I opened Pages. But when opening a document, it again shoots up over 2.8GHz.

Download it, see for yourself, and report back. After observing this for a week, I am more than convinced that the 2012 (and probably the 2013) processors spend more time at the high end of their turbo range than you would expect. At the very least, i7 doesn't need to be "driven hard" or "maxed out on high demand apps" to really go to high clock speeds; they are consistently running at speeds higher than their i5 counterparts even when doing tasks like mail, Safari, iTunes, etc.

This is absolutely spot on, it's how they are fundamentally design to work and do function in an everyday scenario.

 

[Ronnoco]

Again you are reinforcing my point with that response. Dunning-Kruger effect methinks.



I posted the same thing before and the answer was similar, he is not accepting of facts instead clinging to sophistical logic that and i5 = i7 at same clock speed for "average" tasks where it has been proven several times this is not the case. Not even taking into account the L3 cache is very different, this does impact performance...

Honestly, I finally had to put that mattferg dude on ignore. I grew weary of his ridiculous non-factual arguments and his bitterly clinging to nonsensical logic.

 

[AXs]

you guys are just going in circles at this point.

It's one person's word, against the other. No one is going to win.

Maybe if someone would put up test results related to time/speed it would be more conclusive.

I too would like to see how much the i7 improves my browsing. As it is, my i5/8 on 30mbps cable and wireless n300 loads youtube homepage in about 1 second from clicking the bookmarked tab on chrome.

There's no delay or stutter ever, it takes 1 second to open, load, and play an HD video as well.

Microsoft Word takes 1 second to boot, and 1 second to load templates. Maybe you'll save a tenth of a second with the i7?

I think you guys are missing his point entirely, or at least are not open-minded enough to consider 'alternative perspectives'.
Not saying you should agree with him, but at least try and understand what he's saying.

Because as someone with college level debate experience, you guys are arguing over different points, or are equivocating quite a fair bit.

 

[ezekielrage_99]

Honestly, I finally had to put that mattferg dude on ignore. I grew weary of his ridiculous non-factual arguments and his bitterly clinging to nonsensical logic.

LOL that is a good call.

 

[mr.bee]

you guys are just going in circles at this point.

It's one person's word, against the other. No one is going to win.

Maybe if someone would put up test results related to time/speed it would be more conclusive.

I too would like to see how much the i7 improves my browsing. As it is, my i5/8 on 30mbps cable and wireless n300 loads youtube homepage in about 1 second from clicking the bookmarked tab on chrome.

There's no delay or stutter ever, it takes 1 second to open, load, and play an HD video as well.

Microsoft Word takes 1 second to boot, and 1 second to load templates. Maybe you'll save a tenth of a second with the i7?

Why are you extrapolating 'opening one program' with everyday use of a computer given two different processors?

It's much more complex than 'substract [insert your daily task] with 20%' to know if i7 is 'faster' than i5.

 

[AXs]

I guess we should first define 'everyday use' with relativity to this discussion, to avoid arguing over cross-purposes right?

Well, we did that.

I'm not going to get into it again. The Air is marketed to a specific segmentation. And just because a few particulars from a different segmentation decide that the Air may be useful to them, it doesn't redefine the Air's placement in the market.

At the end of the day, the real question in 1 month will be - The Air or The Pro.

Let us see how many will pick the Air over the Retina Pro 13". The i5 vs i7 discussion will become obsolete fast.

Btw, Just noticed i3 was edited to the topic title. So hopefully someone will come tell us why i3Y is better than i5u and is the best chip for browsing

 

[m98custom1212]

you guys are just going in circles at this point.

It's one person's word, against the other. No one is going to win.

Maybe if someone would put up test results related to time/speed it would be more conclusive.

I too would like to see how much the i7 improves my browsing. As it is, my i5/8 on 30mbps cable and wireless n300 loads youtube homepage in about 1 second from clicking the bookmarked tab on chrome.

There's no delay or stutter ever, it takes 1 second to open, load, and play an HD video as well.

Microsoft Word takes 1 second to boot, and 1 second to load templates. Maybe you'll save a tenth of a second with the i7?

I think you guys are missing his point entirely, or at least are not open-minded enough to consider 'alternative perspectives'.
Not saying you should agree with him, but at least try and understand what he's saying.

Because as someone with college level debate experience, you guys are arguing over different points, or are equivocating quite a fair bit.

You chimed in with nothing to support or argue just though out about one second here and there.

Plus, on 30mbps cable your hold up is your internet speed that's only 3.75 Megabytes per second. That's nothing

 

[AXs]

You chimed in with nothing to support or argue just though out about one second here and there.

Plus, on 30mbps cable your hold up is your internet speed that's only 3.75 Megabytes per second. That's nothing

I'm sorry that doesn't make any sense. Can you clarify in clarity please. I find it hard to understand some of your posts, so It is hard to respond.

And please, don't make this about my internet speed. red herring 1o1.

I didn't make any assertions, but you highlighted my point exactly. No one has statistical nor illustrated backing to their claims.

I'll tell you now that my 13" Macbook Air runs Chrome for 15 hours. I'll say it's true because you can't disprove it. You'll say prove it or it is false.
Appeal to ignorance 1o1

Okay seriously, i'm out.

 

[mattferg]

Well I'll probably go with an i7 because I don't want apples slowest CPU. I also found it annoying that the i5 was about the same speed as last years air. I also want it to be more powerful because I was disappointed that the pro didn't get updated and I was most likely going to get that one. I just don't want to wait for October.



Seriously?! Did you even read the whole thing?! It said that even on the same tasks the i7 performed better and more efficiently then the i5. The i7 does everything better. Whatever though.

The i7 doesn't do everything better, that's physically possible and if you understood turboboost you'd know this. Whatever though.

----------

Again you are reinforcing my point with that response. Dunning-Kruger effect methinks.



I posted the same thing before and the answer was similar, he is not accepting of facts instead clinging to sophistical logic that and i5 = i7 at same clock speed for "average" tasks where it has been proven several times this is not the case. Not even taking into account the L3 cache is very different, this does impact performance...

You are aware you're perfectly exhibiting this effect, right? If you took the time to Google it you'd understand that it's unskilled individuals (which is clearly yourself judging by this thread) arguing that they understand a technology as much as those who do, do.

The i5 pretty much equals the i7 at the same clockspeed. If you can't understand this, you really are perfectly demonstrating the Dunning-Kruger effect. At clockspeeds slower than 2.6ghz this extra increase in cache isn't going to give the clearly superior performance you claim, and that's all it is. Extra cache, nothing more.

----------

Ummmmm, what am I missing? Yes the i5 turbo boosts to a max of 2.6ghz and matches the i7 assuming the i7 doesn't boost itself up to its max of 3.3ghz.

So if you are saying that an i5 boosted matches a non-boosted i7, then yes, you win.

I find it hard to believe that even you can make such as weak argument.

i5 max boost = 2.6GHz.

i7 max boost = 3.3GHz.

i7 > i5. Any other flawed comparisons you want to make?

Oh, and BTW,

BASE clock speed of i5 = 1.3GHz.

BASE clock speed of i7 = 1.7GHz.

Just in case you wanted to drop that every day task, non maxed out argument again....

Again, ZBoater showing your misunderstanding of both logic and Intel processors. They don't turboboost up to 3.3ghz, they don't need to. I've run the software on my Mac people keep claiming I haven't, my i5 never goes above 2.3ghz for everyday tasks. A turboboosted i5 at 2.3 will match an i7 at 2.3, and as such spending the $150 is wasted for everyday users. Base clockspeed is irrelevant in Intel's current generation of processors.

----------

Have you downloaded Intel Power Gadget like w00d suggested and actually OBSERVED how the CPU reacts versus just speculating? I did it a week ago and have spent quite a bit of my time looking at how my 2012 i7 MBA performs when doing "everyday" tasks.

The 2012 i7 has a base speed of 2.0GHz, with a turbo of 3.0 in dual core and 3.2 in single core. With all background apps and processes killed, it sits there at 0.8GHz to 1.2GHz. If you launch Safari, it immediately jumps to 3.0GHz for a second and then evens back out at 1.2GHz or so. Start moving the cursor around and it goes to 2.0 to 2.4. Grab a slider on the side of the window, move the window up and down and it again shoots up to 3.0 - 3.1 GHz. Kill Safari and it immediately drops back down to under 2.0GHz and eventually settles down to 0.8GHz.

Similarly, launch iTunes. Again, screams up to 3GHz upon launching and updating Genius, then again settles out. Scroll your albums and it jumps up to 3.0 again. Play a song and it levels back out to 1.0 to 1.2

This is played out time and time and time again. The only app I didn't see this high level of processor speed is when I opened Pages. But when opening a document, it again shoots up over 2.8GHz.

Download it, see for yourself, and report back. After observing this for a week, I am more than convinced that the 2012 (and probably the 2013) processors spend more time at the high end of their turbo range than you would expect. At the very least, i7 doesn't need to be "driven hard" or "maxed out on high demand apps" to really go to high clock speeds; they are consistently running at speeds higher than their i5 counterparts even when doing tasks like mail, Safari, iTunes, etc.

Done this, I don't have the same results (maybe that's because we're running different models... as such your data is irrelevant in this thread)

----------

This is absolutely spot on, it's how they are fundamentally design to work and do function in an everyday scenario.

No, it really isn't. If the processors constantly turboboosted all the way, Intel would just keep the clockspeed higher, as turboboosting takes a lot of power and it'd be a waste for an Intel U chip. If you think it's natural for a processor to go from 0-100-0 all the time with no steps in between, Dunning-Kruger effect for you, again.

----------

Honestly, I finally had to put that mattferg dude on ignore. I grew weary of his consistent logic and arguing of the facts which made me feel bad that I wasted $150.

Fixed this for you.

Literally I'm so done with this thread. I've posted stuff to support my arguments, used consistent logic and facts, and stats, to prove that what you guys are arguing is nonsense. You personally attack me, call me childish and stupid, accuse me of being unskilled, and do the same to anyone who dares to agree with me.

You then bring up irrelevant arguments, refuse to address the points I raise, then when you do, you use ridiculous stuff like the OpenOffice benchmark, arguing that is somehow everyday usage.

For anyone who takes the time to read this whole thread, get the i5 if you don't do anything CPU intensive, including gaming. These guys don't have a clue, and need to justify their purchases to themselves on the internet. I own an i5 and have used an i7 regularly, unless you do lots of editing/coding, you'll never notice the difference.

 

[m98custom1212]

I'm sorry that doesn't make any sense. Can you clarify in clarity please. I find it hard to understand some of your posts, so It is hard to respond.

And please, don't make this about my internet speed. red herring 1o1.

I didn't make any assertions, but you highlighted my point exactly. No one has statistical nor illustrated backing to their claims.

I'll tell you now that my 13" Macbook Air runs Chrome for 15 hours. I'll say it's true because you can't disprove it. You'll say prove it or it is false.
Appeal to ignorance 1o1

Okay seriously, i'm out.

You got my point..

You claimed youtube loads in one second.. That has nothing to with the I5 you will be bottlenecked by your internet speed far before the I5..

 

[Mike in Kansas]

Done this, I don't have the same results (maybe that's because we're running different models... as such your data is irrelevant in this thread)

It's relevant in the fact that folks need to be aware of this application. That way they can download Power Gadget and see for themselves how the clock speed changes in their 2013 MBAs as various tasks are performed on the machine.

My 2013 MBA comes back from college this weekend; I'll be sure to put Power Gadget on it and see how the 2013 differs from the 2012.

 

[AXs]

you got my point..

You claimed youtube loads in one second.. That has nothing to with the i5 you will be bottlenecked by your internet speed far before the i5..

exactly

i finally realized that a lot of people could be complaining about flash being slow, or stuttering a little- when it is directly related to internet speed, and not cpu speed.

Any well haswell laptop chip is overkill for browsing flash. Wifi speeds and reception surely has something to do with it too, but usually will be faster per second than internet speed.

the bottleneck is most likely internet connection.

 

[w00d]

For anyone who takes the time to read this whole thread, get the i5 if you don't do anything CPU intensive, including gaming. These guys don't have a clue, and need to justify their purchases to themselves on the internet. I own an i5 and have used an i7 regularly, unless you do lots of editing/coding, you'll never notice the difference.

I don't understand why you're so adamant about pushing your incorrect perception as fact and insulting the rest of us. Why should anyone blindly believe you... because you can use a bold font and call us idiots? You can try to close your eyes and keep talking louder and louder but it doesn't change reality.

The facts here are in plain sight. The i7 is a superior chip in every regard. In addition to the higher clock speeds (both CPU and graphics) it has more cache and additional instruction sets that the i5 does not. It is a mistake to ignore the effectiveness of these features.

How can you explain that under a light workload (which I am assuming loosely translates to your so called "everyday tasks") the i7 has better battery life than the i5? It's not a big difference, but it is there. Essentially, the i7 is a more powerful CPU at any speed compared to the i5. It is able to get more done in less cycles because of the superior performance of the CPU and then spend more time idle conserving the battery (when doing these super minimal tasks you keep talking about).

All the points I am making have been backed up with specs and data.

Bottom line, why should we believe you when you say the rest of us are idiots and don't know what we are talking about? All the data that has been presented suggests the opposite.

 

[AXs]

Not EVERY aspect.

Not power efficiency and battery consumption. This has been the whole base argument in this thread - vs loss of batter life vs power gain.

And I think we've concluded that "Whatever makes you happy" is the only right answer.

 

[ZBoater]

...And I think we've concluded that "Whatever makes you happy" is the only right answer.

Agreed. I think the continuing issue is the insistence that the i5 and i7 perform the same at "lower" levels of usage, and that the i7 heat and battery usage is "excessive" for what it gives you. I think the Anandtech review provides a well thought out comparison. Yes, it gets hotter. By 4-5 degrees. Yes, it uses up more battery. 52 minutes more over 11-12 hours. Yes, its faster. Up to 28%.

As long as we keep these differences in perspective, we can make educated choices and be "happier".

 

[AXs]

Agreed. I think the continuing issue is the insistence that the i5 and i7 perform the same at "lower" levels of usage, and that the i7 heat and battery usage is "excessive" for what it gives you. I think the Anandtech review provides a well thought out comparison. Yes, it gets hotter. By 4-5 degrees. Yes, it uses up more battery. 52 minutes more over 11-12 hours. Yes, its faster. Up to 28%.

As long as we keep these differences in perspective, we can make educated choices and be "happier".

Exactly. And further more, we should take this time to prepare to battle the oncoming wave for 13" Haswell Retina Pro users.

They will be attacking us soon enough talking about how the Retina is more portable than the Air and is Faster and has same battery life (even if Apple announces 9 hours on the Retina model). Yea, 9 hours will be 'same' as 12 hours.

Brace yourselves. Apple Winter is coming.

 

[SchodMC]

Exactly. And further more, we should take this time to prepare to battle the oncoming wave for 13" Haswell Retina Pro users.

They will be attacking us soon enough talking about how the Retina is more portable than the Air and is Faster and has same battery life (even if Apple announces 9 hours on the Retina model). Yea, 9 hours will be 'same' as 12 hours.

Brace yourselves. Apple Winter is coming.

The enemy of my enemy is my friend... What happened over the time? I mean, where are the times where the "battle" was against Windows / PC user?

cu
SchodMC

 

[PDXPean]

Agreed. I think the continuing issue is the insistence that the i5 and i7 perform the same at "lower" levels of usage, and that the i7 heat and battery usage is "excessive" for what it gives you. I think the Anandtech review provides a well thought out comparison. Yes, it gets hotter. By 4-5 degrees. Yes, it uses up more battery. 52 minutes more over 11-12 hours. Yes, its faster. Up to 28%.

As long as we keep these differences in perspective, we can make educated choices and be "happier".

I am curious. Do you use yours frequently on your lap? If so, doing intense work? Does it feel hot on your lap at all?

 

[ZBoater]

I am curious. Do you use yours frequently on your lap? If so, doing intense work? Does it feel hot on your lap at all?

I do not. Although it is a "lap"top I never use it on my lap. I use it 80% on a desk, and 20% on my chest while in bed, but the (very sharp edge) of the bottom half against my chest, not the flat bottom. So "heat" is never an issue for me.

 

[PDFierro]

Exactly. And further more, we should take this time to prepare to battle the oncoming wave for 13" Haswell Retina Pro users.

They will be attacking us soon enough talking about how the Retina is more portable than the Air and is Faster and has same battery life (even if Apple announces 9 hours on the Retina model). Yea, 9 hours will be 'same' as 12 hours.

Brace yourselves. Apple Winter is coming.

If I was getting a MBA, I would definitely max it out and get the i7. But on the rMBP, it's not as much of a clear-cut choice.

 

[sofianito]

For anyone who takes the time to read this whole thread, get the i5 if you don't do anything CPU intensive, including gaming. These guys don't have a clue, and need to justify their purchases to themselves on the internet. I own an i5 and have used an i7 regularly, unless you do lots of editing/coding, you'll never notice the difference.

What do you mean by editing/coding? programming & compilation? Single threaded tasks?

----------

I do not. Although it is a "lap"top I never use it on my lap. I use it 80% on a desk, and 20% on my chest while in bed, but the (very sharp edge) of the bottom half against my chest, not the flat bottom. So "heat" is never an issue for me.

Do you feel the magnetic field on your chest? Does it turbo-boost your heart?...

 

[Scott6666]

Noise

In all this discussion between them, there is almost nothing mentioned of the fan noise that comes with each.

$150 is a relatively minor sum which I can see paying for the (sometimes) extra power of the i7.

HOWEVER, if the cost of the extra power is fan noise it's better to go with the i5. Can we have 2 pages of back and forth on this please.

I'm ready to buy but still after all this thread, I can't figure out which I should get. I want power but not if it comes with noise.

I know the i7 must be noisy sometimes, say with games or handbrake, but in everyday use (including coding) will I notice extra noise over the i5? I read Anand's review but found his notation on noise to be rather terse and, for me, not so convincing.

 

[ZBoater]

If you so a search on "fan noise", you will see that topic has been beaten to death even more so than this one. And that's saying a lot.

So the question is not whether it is noisier, as they have the same fan, but whether one runs the fan more than the other. The problem is "noise" is somewhat subjective. Time with the fan running over 2000rpm is a little more objective. Even when my fan it running I barely hear it. No, I am not hard of hearing. But the fan blowing at 6000 rpm doesn't really bother me nor do I notice it.

Other people are more sensitive to this "noise". Probably the same people who can't live without retina. If you are noise sensitive, then you want the one that is likely to run the fan less.

I have not seen objective comparisons on fan runtime. The i7 is likely to reach a higher temperature quicker than the i5 because it is working harder/faster. So one can assume the fan will run longer. How much longer or how much harder will be dependent on use, and whether that is enough "noise" to make a difference for you is entirely dependent on your "noise" sensitivity.

When you do your search and read some of the noise threads, you will notice some make it seem like there is a leaf blower inside their systems, or there is a death rattle that will pierce your ear drums. These, of course, are likely exaggerations. Others will dismiss the noise (like me) as nothing to be concerned about as it is just a very low level, barely perceptible woosh sound. Both processors will cause the fans to turn on. I've been sitting at my i7 all day browsing the web and smc Fan Control reports the rpms at 1991-2013, and it has barely moved. When I run Civ5, I can't see it but I can hear it. I turn on the game music volume and the problem is fixed.

 

[l.a.rossmann]

You'll notice the difference in your wallet MUCH more than the difference in performance.

 

[Dr. McKay]

Just ordered my MBA 13/i7/8/256 from an Apple Premium Reseller (closest thing we have to a real Apple Store over here in Belgium.
Oooh, the agony of waiting