I'm running the i7 2015 model. I get a bit annoyed at the ambient 1200rpm noise (after a year of basically silent MBP, and a year of totally silent MB). I assume the ambient noise from the 2017 model is the same. I have heard the fan go very loud, but usually just for small moments since these powerful machines can finish the task quickly.
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.
The new iMac is a lot noisier!
- Thread starter Darajavahus
- Start date
- Sort by reaction score
I put this in another thread but probably more relevant here...
I have an i7/580 coming but I also got a chance to test CPU stress on a 3.4GHz i5 base model 27".
With just CPU stress could not get temps over 68degC or Fan off of 1200RPM. Idle temp ~42degC. As a Pro Audio (recording) guy I need the best speed and the least noise. Been using a 2013 MacPro for 3 years - as close to silent as it gets at any load.
Downloaded latest istat menu (30d free if you don;t own it) for all the measurements
1 YES = 25%CPU Didn't record Watts after 10min Peak~52degC shares cores well! 1200rpm
2 YES = 50%CPU 21W CPU after 10min Peak~58degC
3 YES = 75%CPU 27.6W CPU after 10min Peak~64degC 1200rpm
4 YES = 100%CPU 33.4W CPU …… 66degC! 1200RPM.
I would be very interested in this same text with the i5 3.8/580 and the i7 4.2/580. In general the temps stabilize in a few minutes so it doesn't take that long...
http://osxdaily.com/2012/10/02/stress-test-mac-cpu/
I have an i7/580 coming but I also got a chance to test CPU stress on a 3.4GHz i5 base model 27".
With just CPU stress could not get temps over 68degC or Fan off of 1200RPM. Idle temp ~42degC. As a Pro Audio (recording) guy I need the best speed and the least noise. Been using a 2013 MacPro for 3 years - as close to silent as it gets at any load.
Downloaded latest istat menu (30d free if you don;t own it) for all the measurements
1 YES = 25%CPU Didn't record Watts after 10min Peak~52degC shares cores well! 1200rpm
2 YES = 50%CPU 21W CPU after 10min Peak~58degC
3 YES = 75%CPU 27.6W CPU after 10min Peak~64degC 1200rpm
4 YES = 100%CPU 33.4W CPU …… 66degC! 1200RPM.
I would be very interested in this same text with the i5 3.8/580 and the i7 4.2/580. In general the temps stabilize in a few minutes so it doesn't take that long...
http://osxdaily.com/2012/10/02/stress-test-mac-cpu/
I would be very interested in this same text with the i5 3.8/580 and the i7 4.2/580. In general the temps stabilize in a few minutes so it doesn't take that long...
The question boils down to several factors.
First, the fan in an imac is clearly louder above a certain rpm. If you have keen ears, and are used to fanless computers (such as a macbook, or an ipad), even the base 1200 rpm will be annoyingly non silent-- I can't hear it above the noise floor in my home. For me, 1800 rpm is audible, and it quite gradually ramps up to something objectionable-- to be "masked" by the sound effects of a game, or itunes.
But, apple may have tuned the fan design in the last few iterations. So even my perception of the speed bands may be different. One thing is for sure, the effect doesn't come across in a you tube video--mike placement and mike gain, as well as playback volume will change the result, drastically.
There's a power profile associated with the CPU. Essentially, the closer the CPU comes to reaching it's thermal limit, the faster the fans spin, until the cpu either cools down, or underclocks. You can see this in action with Intel Power Gadget (which may not yet be available for the latest imacs.)
So, you need to either run a benchmark which stresses the CPU for as long and as hard as the apps that you run, or you need to run your loads on it.
A multiple of yeses don't really stress the CPU as hard as Prime95. And Cinebench is often over before serious heat accumulates in the CPU cores. (Video playback is not even close. That sort of question makes me wonder "What the hell is wrong with your computer")
The real question is whether your workload is sufficiently intense enough , and your chosen CPU insufficiently efficient enough to boost the fan semi permanently into that audible range.
PS. Oh joy. Prime95 has now kicked my CPU into underclocking. It's now 3.4 Ghz (down from 3.5), and about 96 degrees, and the fan is working at 2700 rpm. Think I'll stop now.
Last edited:
Gotcha alright let us know how that works and once mine lands I'll report back with my findings.
1) In my experience on all Macs I have owned so far the algorithm for fan speed is quite simple. It will stay at the idle speed till the CPU temp gets somewhere around 95degC. It will then increase to keep the CPU at that temp.
2) I am always up for learning so you may very well know what stresses a CPU more. To my mind - the YES approach is just an easy way to do loading. On the 2017 27 i5 imac base mode - one instance of YES loads one core. When I used 4 YES, the CPU was 100% busy. So much that a file copy I had going in parallel dropped to almost zero MB/s. Is there another way to make the CPU 100% busy and be more stressful? Interesting so I invite you to explain. I looked for a MAC version of Prime95 to try - could not find one. If you can point to a download site and video of how to do it I am happy to try it.
3) Of course the workload is the key. Low Latency audio recording work IME is much like the YES approach. In general I do 50% or less CPU load - but it is a constant load. On a 2013 MacBook Pro i7 - this would incur full speed fans. On an i7 imac 2013 it would take 70% load. On the 2013 i5 iMac (which was similar to the i5 2017) - 100% CPU load from "YES" would also only get ~80degC CPU and never ramp fans. Audio work on that machine was the same. The current base model never went above 68degC. That is a significant improvement over the 2013 IMO.
At 1200rpm it is silent. I was worry about the noise after using a 2013 silent Mac Pro. The fan kicks in when stressed but not annoying (not at frequencies that bother me) enough for me to return it.
I think the questions at hand are compounded:
1) fans at the same RPM, even the "idle" 1200, users report are noticeably lounder than previous iMacs;
and
2) i7 7700k and/or RX850 heats up much quicker than the i5 options, so fans kick in sooner, and needs to work harder
So in practice, higher fan noise and longer time of it may be the case with 2017, especially with i7. The question is how much more noise is there for particular use cases. The only way to know is to have bought the model you want and set it up to run exactly how you want the iMac to, with your typical room condition etc. 14 day return policy is there if you need it (unfortunately where I am we don't have it.)
1) fans at the same RPM, even the "idle" 1200, users report are noticeably lounder than previous iMacs;
and
2) i7 7700k and/or RX850 heats up much quicker than the i5 options, so fans kick in sooner, and needs to work harder
So in practice, higher fan noise and longer time of it may be the case with 2017, especially with i7. The question is how much more noise is there for particular use cases. The only way to know is to have bought the model you want and set it up to run exactly how you want the iMac to, with your typical room condition etc. 14 day return policy is there if you need it (unfortunately where I am we don't have it.)
I would assume the noise level to be similar since the general consensus seems to be that CPU (i7) heats up faster and thus causing the fan to kick in.
Yea but still, an ambient MBP is much more silent than my iMac, and that also has a fan right?
It is a smaller fan with much less air intake. Mine runs at ~2000rpm when barebone OS X is running, CPU die at ~50 degrees Celsius, and the fan noise is completely drowned by other ambient floor noise in my working environment.
Last edited:
I think it would only matter if you're stressing the GPU. In that case, the 575 is likely to be cooler and quieter.
Yes, and those fans idle at ~2000 rpm. They're also a lot smaller, so when you do hear the fans the tone is annoyingly high-pitched. Once you start stressing the MBP it's much more likely than an iMac to spin the fans loudly. Especially on warm days and/or if you have an external monitor plugged in.
At least the 2015 MBP 13" is a semi-passive machine - in other words the fans don't spin at all (or veeeeery little) under normal workloads. You can press your ear right next to the machine in absolute quietness without being able to hear anything.
The iMac at 1200rpm is a slight downer for me since it's quite audible in comparison. When turning the machine on, you can immediately hear the 'swoosh' starting when the log in screen appears. Mine also makes very quiet ticking sounds (comparable to modern hard drives loading something, quiet ticks) that are slightly uneven too (I have the pure SSD, so it isn't that).
I was also almost sure I didn't notice the fan that much a couple of days ago when I got the machine, but now I can't seem to un-notice it. Fiendish loop.
I dunno. Maybe I should contact Apple and ask for a replacement while I'm within the 14 days - is that how it works? Grr.
Edit: On the upside, playing WoW and Divinity: Original Sin results in virtually no increase in fan RPM, so definitely no complaints there (i5/575), not-overly-intensive gaming has been a pleasantly quiet and cool experience.
The iMac at 1200rpm is a slight downer for me since it's quite audible in comparison. When turning the machine on, you can immediately hear the 'swoosh' starting when the log in screen appears. Mine also makes very quiet ticking sounds (comparable to modern hard drives loading something, quiet ticks) that are slightly uneven too (I have the pure SSD, so it isn't that).
I was also almost sure I didn't notice the fan that much a couple of days ago when I got the machine, but now I can't seem to un-notice it. Fiendish loop.
I dunno. Maybe I should contact Apple and ask for a replacement while I'm within the 14 days - is that how it works? Grr.
Edit: On the upside, playing WoW and Divinity: Original Sin results in virtually no increase in fan RPM, so definitely no complaints there (i5/575), not-overly-intensive gaming has been a pleasantly quiet and cool experience.
Last edited:
At least the 2015 MBP 13" is a semi-passive machine - in other words the fans don't spin at all (or veeeeery little) under normal workloads. You can press your ear right next to the machine in absolute quietness without being able to hear anything.
The iMac at 1200rpm is a slight downer for me since it's quite audible in comparison. When turning the machine on, you can immediately hear the 'swoosh' starting when the log in screen appears. Mine also makes very quiet ticking sounds (comparable to modern hard drives loading something, quiet ticks) that are slightly uneven too (I have the pure SSD, so it isn't that).
I was also almost sure I didn't notice the fan that much a couple of days ago when I got the machine, but now I can't seem to un-notice it. Fiendish loop.
I dunno. Maybe I should contact Apple and ask for a replacement while I'm within the 14 days - is that how it works? Grr.
Edit: On the upside, playing WoW and Divinity: Original Sin results in virtually no increase in fan RPM, so definitely no complaints there (i5/575), not-overly-intensive gaming has been a pleasantly quiet and cool experience.
Thanks for the input. So on your i5 575 is WoW playable at 5k and also what is your temps? I ordered the i7 580 and hoping to push it but want to know if I ordered a vacuum or an actual quiet machine.
My i5 / 580 iMac has arrived! So, w.r.t thermals:
– maximum load with Prime95 initially draws ~77W from the CPU. But this drops to 70W (thermal throttling!) and the fan goes to 2700 rpm. Stable at around 95 degrees and 70W power consumption.
– medium load with Prime95 (two cores) draws ~45W. Fan slowly fluctuates between 1200 and 1600 rpm. CPU temp 90 degrees.
The following are with Unigine running in the background:
– two loaded cores with Prime95 + Unigine Valley (50W CPU power usage total) = CPU core temp 95 degrees, GPU core temp 77 degrees, 2700 rpm fan. I cannot get the CPU to consume more than 60W. Notably, the FPS during Unigine do not drop, so it looks like the GPU keeps its cool nicely. It will not throttle.
– as soon as I quit Prime95, with Unigine still running, the fan goes to 1900 rpm and stay there. GPU die temp stays at around 77 to 80 degrees.
– if I start Prime95 again, but with only one thread (35W CPU draw), the fan spins down to 2300 rpm.
Ambient temperature is a toasty 25.5 degrees. Idle temps are about 46 degrees for CPU and GPU. All chip temperatures given are the die temperature. Not something nearby the CPU or GPU, but the temperature of the chip itself.
I conclude that the Intel CPUs run pretty warm. Even the fastest i5 is thermally-constrained. Although you need to push the CPU with a heavy sustained load in order to ramp up the fans, it's these loads that the i7 is good at. However, the i7 will probably throttle down to the same speeds as the i5. I would really advise against getting the i7 just for the sake of getting the highest specs – only get it if you really need it. And even then, it probably won't be much faster due to thermal constraints.
Most games shouldn't cause the fans to spin up to max, unless they're very CPU intensive. Especially if you use vsync to cap the framerates, so GPU utilization is lower than 100%, you should keep the iMac running cool and quiet.
If I guesstimate that the GPU can sustain using 95 W of power, this shows that there is a considerable difference between the iMac's ability to cool the CPU vs the GPU. Perhaps the difference lies in the fact that AMD solder the heat spreader to the chip, whereas Intel use a thermal paste. It's a bit strange.
A few other thoughts:
– maximum load with Prime95 initially draws ~77W from the CPU. But this drops to 70W (thermal throttling!) and the fan goes to 2700 rpm. Stable at around 95 degrees and 70W power consumption.
– medium load with Prime95 (two cores) draws ~45W. Fan slowly fluctuates between 1200 and 1600 rpm. CPU temp 90 degrees.
The following are with Unigine running in the background:
– two loaded cores with Prime95 + Unigine Valley (50W CPU power usage total) = CPU core temp 95 degrees, GPU core temp 77 degrees, 2700 rpm fan. I cannot get the CPU to consume more than 60W. Notably, the FPS during Unigine do not drop, so it looks like the GPU keeps its cool nicely. It will not throttle.
– as soon as I quit Prime95, with Unigine still running, the fan goes to 1900 rpm and stay there. GPU die temp stays at around 77 to 80 degrees.
– if I start Prime95 again, but with only one thread (35W CPU draw), the fan spins down to 2300 rpm.
Ambient temperature is a toasty 25.5 degrees. Idle temps are about 46 degrees for CPU and GPU. All chip temperatures given are the die temperature. Not something nearby the CPU or GPU, but the temperature of the chip itself.
I conclude that the Intel CPUs run pretty warm. Even the fastest i5 is thermally-constrained. Although you need to push the CPU with a heavy sustained load in order to ramp up the fans, it's these loads that the i7 is good at. However, the i7 will probably throttle down to the same speeds as the i5. I would really advise against getting the i7 just for the sake of getting the highest specs – only get it if you really need it. And even then, it probably won't be much faster due to thermal constraints.
Most games shouldn't cause the fans to spin up to max, unless they're very CPU intensive. Especially if you use vsync to cap the framerates, so GPU utilization is lower than 100%, you should keep the iMac running cool and quiet.
If I guesstimate that the GPU can sustain using 95 W of power, this shows that there is a considerable difference between the iMac's ability to cool the CPU vs the GPU. Perhaps the difference lies in the fact that AMD solder the heat spreader to the chip, whereas Intel use a thermal paste. It's a bit strange.
A few other thoughts:
- the minimum brightness setting is really dim. Even dimmer than my 2012 rMBP. I love it!
- I do not notice the fan at idle. If I put my ear to the case, I can hear the fan motor (and maybe even the ticking noise people are talking about), but IMO it's much ado about nothing. In a quiet room, sitting in front, I can only just hear the very faintest rushing of air at idle. When sitting in front of the iMac I never hear the noise of the fan motor itself, no matter what the fan speed – only the rushing of air.
- The fans at maximum speed are much less annoying – and subjectively quieter – than my rMBP.
- the screen is amazing!
I had a top of the line 2015 and now top of the line 2017. The 2017 fan is noticeably quieter when the cpu is stressed out than the 2015 was. I was expecting much worse, so I am very pleased.