I just tested this on both 2015 and 2017 iMac i7s, both with and without hyperthreading enabled. I'd be very interested if you could verify whether an iMac 27 with i7-7600 CPU remains "near silent" with fan at 1200 rpm for 10 minutes when running that same Handbrake transcode job.
On both my 2015 and 2017 iMac 27 i7, when transcoding that H265 4k video to H265 output using the Roku 4k preset, the fans rapidly increased to about 2700 rpm by +60 sec, and temps on both fluctuated from 95-100C. I could not tell any audible difference in the noise from either 2015 or 2017 machine when running this test -- they were both very noticeable.
This was unchanged when I disabled hyperthreading using the CPUSetter utility. This in effect makes the i7 like an i5. This was confirmed by watching all eight logical CPU cores in iStat Menus. When hyperthreading was on, all eight cores were at 100%. When hyperthreading was disabled, only four logical cores were at 100%, the other four were zero.
To expedite testing I used a shorter H265 video from that site with the similar characteristics: "Philips Supershop Demo 1". Using this video and the above encoding I got the following performance using Handbrake 1.0.7 to H265 with the Roku 4k/30 preset:
2015 iMac 27 i7, hyperthreading off: 12:21
2015 iMac 27 i7, hyperthreading on: 10:53
2017 iMac 27 i7, hyperthreading off: 11:11
2017 iMac 27 i7, hyperthreading on: 09:53
Wow, that was some quick testing for multiple machines.
To be completely honest I couldn’t be 100% sure that was the exact file I used but I think it was. So for a brief quasi-confirmational test before I left for work, I tried it again with that specific file but only let it run for a few minutes, not the whole 10 mins (because I was running short on time). After those several minutes the fan was still at 1200 rpm.
Note though that while the i7-7700K with HT off is effectively an i5, it is a 91 W TDP i5 running at 4.4 GHz multi-Core Turbo. In contrast, my 7600 is a 65 W TDP i5 running at 3.9 GHz multi-core Turbo. Interestingly, I believe I saw somewhere the i7’s first throttling speed to cool down (if the fan isn’t sufficient) is 3.9 GHz which happens to be the exact same as the regular Turbo speed of the 7600.
BTW, I found my post on this:
https://forums.macrumors.com/threads/the-new-imac-is-a-lot-noisier.2051501/page-23#post-24783381
A bunch of us did some YES testing in the terminal which basically maxes out the CPU with a useless command. I ran 4 instances of yes to peg all 4 cores at 100%. After 10 minutes of this the machine plateaued at 76C, with no throttling and with the fan at 1200 rpm. Interestingly, the 7600K performed the same. No movement of the fan above 1200 rpm and similar temps. The 7600K is a 91 W TDP chip but I guess if you’re lucky it can perform similarly to the 7600. However, its max multi-core Turbo is only 4.0 GHz, which is barely more than the 3.9 GHz of the 7600 so that’s not a huge surprise.
In that post I also show my h.265 encode which after 10 minutes brought my 7600 to 83C, yet still with 1200 rpm fan.
Lots of pertinent screen grabs in there by the way, as well as some graphs from a Tom's Hardware review comparing the power utilization curves of the 7600 vs the 7600K. That same review also has the 7700K (and 7700). The graphs for the 7700K show that it is a power utilization monster, in comparison to the 7600.
http://www.tomshardware.com/reviews...i7-7700k-i7-7700-i5-7600k-i5-7600,4870-8.html
Here is the 7700K (with HT on), with a stress test (but not a power virus test):
Here is the 7600 with the same load:
With the power virus test the differences were even more striking.
7700K:
7600:
