This is going to be a bit of a spinoff of my previous requests for help, but I think the topic is different enough to grant a separate post.
I have a Macbook Air mid 2012 11" and a Macbook Air early 2015 13".
I love the 11" (I bought the 13" as backup) but it has fan issues: even after SSD and battery change as soon as I use something a bit more demanding it starts spinning like crazy. I’m still trying to wrap my head around the reasons and to understand if it’s just “too old”. No fan problems with the 2015 model.
My question in this case concerns a comparison between processors. This is what they look like:
Macbook Air 13 Early 2015
| Processor Name:|Intel Core i5|
| Processor Speed:|1,6 GHz|
| Number of Processors:|1|
| Total Number of Cores:|2|
| L2 Cache (per Core):|256 KB|
| L3 Cache:|3 MB|
| Memory:|8 GB|
Macbook Air 11" Mid 2012
| Processor Name:|Intel Core i7|
| Processor Speed:|2 GHz|
| Number of Processors:|1|
| Total Number of Cores:|2|
| L2 Cache (per Core):|256 KB|
| L3 Cache:|4 MB|
| Memory:|8 GB|
In a layperson’s eyes, such as mine, the 2012 machine would seem to come up on top, is that even possible, given that it’s almost three years older? Am I missing anything?
I realize the 2012 might just have been overused and there’s nothing one can do, but I would still like to understand.
Based on the Geekbench scores, the 2015 machine is a bit faster, but not hugely so.
(make sure you’re looking at the 2ghz entries for the 2012 11")
That said, I suspect that the difference in fan noise is probably due to something other than differences between the chips. Maybe the cooling system in the 13" is better, maybe the 11" has a lot of dust bunnies in it, maybe the thermal paste attaching the heat sink to the processor has degraded over time, etc.
You can use iStat Menus to get a better idea of where the heat source is, if any. If you increased the size of the SSD, that might could be the culprit.
Thank you so much, I didn’t know this website. Do you know how tests are made and why are there differences (with the same parameters)? For some of the tests the 2012 actually scores better.
As for dust and thermal paste: it has been cleaned and the thermal paste changed. So I will have to go for a better cooling system for the 2015. My configuration was the best available one at the time (made to order), so maybe the regular fan wasn’t up to the challenge. Either way it’s a bit irritating. I grow very attached to my Macs!
The fan problem pre-dates the SSD upgrade, but the SSD was bigger than the basic configuration right from the start (I think the minimum was 128 and I got a 512, built in by Apple, I now have 1TB). I’ve had the impression before that they gave these upgrading options without actually thinking that they would need more cooling, so it may very possibly be a congenital shortcoming. Having said that activity monitor often does flag one or two app as eating up a lot of CPU when the machine gets loud.
I’m not quite sure exactly what you’re asking but I’ll provide a couple of answers that may or may not be correct and may or may not answer you
CPUs generally become more efficient over with each generation, so it’s quite possible that a newer i5 running at a lower frequency will have better performance than an older i7.
The thermal design of the computers is likely much different. The larger 13" computer is probably better at managing heat than the smaller 11", which might account for the 11" Air’s tendency to spin up the fans more aggressively. That heat could also cause the system to slow the CPU’s operation, which may also explain why the 13" Air is faster (or faster than its clock and place within the CPU line might indicate).
Just trying to understand why the 2015 doesn’t panic when the 2012 does (fans spinning) , even though on paper the 2012 equals or exceeds the results of the 2015.
Thanks for your contribution, it makes sense. That’s the kind of stuff I don’t know. I wish I could do something to make the 2012 more resilient, it remains my favourite!
To reinforce that: We - on the mainframe - dropped the clock speed from 5.5GHz to 5.0GHz (zEC12 to successor z13).
Most of the time we got faster single engine speed (and faster multiple). Occasionally not - and those cases were well understood.
(And features such as SIMD and SMT-2 - where exploited - are examples of where each generation of processor brings something new. Possible Intel added accelerators like that.)
Thanks Martin. I only understand this partially, but I get the gist.
Bottom line: CPU numbers (in my case I5 vs i7 and 1,6GHz vs 2Ghz) only matter so much by themselves and when something is so much newer it ought be better too (my own translation for those with limited technical knowledge).
This brings me to my next musing: would it be possible to create a Frankenstein machine? Say solder a newer processor (and more RAM) onto an older motherboard (so that it still fits the small 11" case)? I wouldn’t attempt this myself, but if I were to get a few spare pieces and find someone willing and able to proceed, could it work? Does anybody have concrete examples of this having been done before? This part is just for the sake of discussion, no actual emergency, but I would still be curious. Thanks again!
I’m writing with hobbyist level understanding, but I fancy offering my thoughts.
You are getting into system-on-a-chip, SOC territory here.
The cpu as a discrete chip could easily have 168 connecting pins. What that means is that it is expecting to be plugged in to a board and have lots of other chips etc on that board for it to connect to.
It also expects those other chips to be a particular specification.
So, typically, if you take out a chip from one generation and solder it onto a board from a different generation it won’t work because the connections just don’t match up.
I wouldn’t be surprised if there were examples of when this worked, but they’d be quite old. Note that changing within the same generation, or closely matched generation, was and maybe still is common, with chip sockets being built with levers to ease chip removal.
Within SOC, those pins on the cpu have become hard wiring to other parts on the SOC.
One last thought (maybe): could less VRAM make the fans spin? The problem usually manifests itself with many Safari windows open and/or with video chats, audio recording, etc.- Does that make sense? In my current machines the 13" (2015) has 1.5 GB VRAM against 512 MB of the 11" (2012).
It all uses energy, so yes… but CPUs use the most so VRAM is unlikely to be the issue.
If this fan noise is something new, it is most likely (imo) to be due to either the CPU generating too much heat because a software bug is causing it to work too hard, or the heat is not being dissipated fast enough due to poor heat sink adhesion or dust in the system.
Video calls can tax the cpu massively, so that could be the source.
It is also possible the i7 chip handles video more efficiently, but I’ve no idea about if it does have extra capabilities like that.
No, not new, unfortunately. No dust, it’s been cleaned, thermal paste changed.
No software bug either, methinks: new SSD and clean OS install. Only a few programs installed, one by one, no user transfers. There’s hardly anything that can create a conflict.
Over 450!
These chips, especially in mobile devices, are usually soldered directly to the board. That involves applying solder paste to the board, positioning parts, and putting the whole thing in a reflow oven. That melts the solder between the parts and the board, and makes the thousands of total connections.
It’s not impossible to exchange parts, but nearly so without special equipment and/or hacking skills.
Mostly Safari, but also Zoom, Adobe Connect, Audio Hijack, Movavi Screen Recorder and other random things I do not even recognise. For a while I thought they might be pieces of software from a distant past (I have been cloning and importing data from older machines for years), but now the computer is clean, so they can’t be.
The same stuff on the 2015 machine causes no problem.
