Mac mini i3 vs i5 temps

[Cashmonee]

He did an update and says here that the i5 model runs really hot to the touch.

I'm returning mine when it arrives and am going for the i3.

Why exactly are you returning the i5 for the i3? Genuinely curious. It seems if you need the speed the i5 will still be faster and if you don't need the speed and concerned with quiet operation, then the i3 seems best regardless. I am not understanding why the temperatures are a big enough concern to return the i5 for the i3.

 

[Spectrum]

Thank you. Yours is probably the first such result on the internet.

So the conclusion to this question is that neither the i5 nor the i7 can maintain their rated multicore speeds even with a 100C allowance. The i3 can and does so at under 90.

True - I guess with a much more serious cooling system it may be possible to allow the i5 and i7 hit their maximum multicore turbos. But regardless, they are still going to be faster than the i3 because they have 50% extra cores.

They will also be faster for less threaded tasks because their base speed and boost speeds are both higher.

 

[Jorbanead]

True - I guess with a much more serious cooling system it may be possible to allow the i5 and i7 hit their maximum multicore turbos. But regardless, they are still going to be faster than the i3 because they have 50% extra cores.

They will also be faster for less threaded tasks because their base speed and boost speeds are both higher.

Exactly - even if the i3 doesn’t technically throttle, the i5 and i7 will still be faster for multithreaded tasks due to more cores and more threads.

Whether you need an i3, vs i5 or i7 greatly depends on your workflow. A lot of these stress tests being done are arguably the most demanding tasks you will do. I do a lot of audio production, and rarely does each core max out fully, but it definitely will utilize as many threads as possible. For me, the i7 makes total sense. I’ll rarely be pushing each core to their max, because Logic will be able to spread out the workload more evenly. If I had an i3, each core would be closer to max frequency and could possibly overload.

 

[Mol1n]

It seems if you need the speed the i5 will still be faster and if you don't need the speed and concerned with quiet operation, then the i3 seems best regardless.

There is a limit for which temps within it can be tolerated. A temperature in the high 90's exceeds that by leagues.

Throttling also causes short interrupts when the power state is being adjusted, causing hiccups that running at a consistently slower speed wouldn't see.

MarkosXT also says the fan doesn't spin up in the i3 model.

 

[F-Train]

He did an update and says here that the i5 model runs really hot to the touch.

I'm returning mine when it arrives and am going for the i3.

I think that it's worth noting that this gentleman also said that it was 19ºC in Sydney, which he characterised as "hot". Nineteen degrees Celsius is 66ºF. He has a pretty expansive view of what "hot" is.

I did a number of tests on an i5/8GB today, including tests that put stress on the computer, and the case didn't even get warm, let alone hot.

I was pleased that it could convert a 640MB 4K video from H264 to H265 in 05:16.

 

[SoCalReviews]

There is a limit for which temps within it can be tolerated. A temperature in the high 90's exceeds that by leagues.

Throttling also causes short interrupts when the power state is being adjusted, causing hiccups that running at a consistently slower speed wouldn't see.

MarkosXT also says the fan doesn't spin up in the i3 model.

If you are that paranoid about running high thermals and willing to neuter the performance of a 2018 Mini by having less cores and a lower clock speed why don't you just buy a new 2014 dual core model. I'm sure it would achieve the cooler temperatures you seem to be looking for and you might even get it at a discounted price.

 

[FreakinEurekan]

He did an update and says here that the i5 model runs really hot to the touch.

I'm returning mine when it arrives and am going for the i3.

So... you're under the impression that a less powerful CPU, will run cooler under the same load??

 

[Cashmonee]

There is a limit for which temps within it can be tolerated. A temperature in the high 90's exceeds that by leagues.

Throttling also causes short interrupts when the power state is being adjusted, causing hiccups that running at a consistently slower speed wouldn't see.

MarkosXT also says the fan doesn't spin up in the i3 model.

I think you are reading far too much into the hype that is being created about temperatures. I really would not worry about getting the i5 or i7 from a performance or reliability standpoint.

 

[Dr. Dr. Evil]

I think that it's worth noting that this gentleman also said that it was 19ºC in Sydney, which he characterised as "hot". Nineteen degrees Celsius is 66ºF. He has a pretty expansive view of what "hot" is.

I did a number of tests on an i5/8GB today, including tests that put stress on the computer, and the case didn't even get warm, let alone hot.

I was pleased that it could convert a 640MB 4K video from H264 to H265 in 05:16.

Did you run some more reports during the test like intel power gadget?
Thank you

 

[Mol1n]

I think that it's worth noting that this gentleman also said that it was 19ºC in Sydney, which he characterised as "hot". Nineteen degrees Celsius is 66ºF. He has a pretty expansive view of what "hot" is.

I did a number of tests on an i5/8GB today, including tests that put stress on the computer, and the case didn't even get warm, let alone hot.

I was pleased that it could convert a 640MB 4K video from H264 to H265 in 05:16.

It was his 10 minute stress test in the first video that was more interesting. In any case I have already tested the i5 Mini with Cinebench in that other thread of yours. So these earlier videos are no longer relevant.

If you are that paranoid about running high thermals and willing to neuter the performance of a 2018 Mini by having less cores and a lower clock speed why don't you just buy a new 2014 dual core model.

This thread was created to ask for benchmark info, not fishing for clever snipes.

So... you're under the impression that a less powerful CPU, will run cooler under the same load??

The 'impression' comes from reading off benchmark results submitted on this forum directly.

 

[SoCalReviews]

So... you're under the impression that a less powerful CPU, will run cooler under the same load??

If anything this thread has further sold me on wanting to own the i7... even if it's just for bragging rights while posting in this forum.

 

[F-Train]

Did you run some more reports during the test like intel power gadget?
Thank you

I'm going to run Intel Power Gadget in the morning. Speaking of which, on the Windows version one can apparently access an expanded Graphical User Interface by right clicking. The Mac equivalent doesn't work, at least with the new version of IPG that I downloaded. Is there another way it's accessed on a Mac?

 

[brentsg]

Obviously that is based on the assumption that the i3 runs substantially cooler. I'm still trying to find someone who can post their i3 thermals. No one seems to have it.

A CPU that's doing far less work per unit time should be cooler. So what? As long as it isn't throttling so significantly that the i5 or i7 can't outperform the i3 by an acceptable margin.. who cares? And even if the thermal/power envelope of the Mini is such that they all would converge to the same general performance under 100% sustained load indefinitely, I still wouldn't be too concerned b/c the faster CPUs would be better performers in all other situations.

Buy what you need. If you need an i3, buy an i3. I'd argue if 100% sustained load all the time is your objective, you're buying the wrong tool with a Mini anyways.

There has been a lot of talk of late about Intel CPUs and the various parameters that are involved with determining what "turbo" clocks will be used, and for what duration. In short, the discussion centers around manufacturers exceeding the guidelines while Intel looks in the other direction. The end result is that we have CPUs far exceeding the rated TDP values (which are likely at stock base clocks), often indefinitely. This is fine in the enthusiast space, as people are using motherboards with overbuilt VRMs and water, or air coolers (some of which are larger than a Mac Mini) that can handle it.

Expecting something as small as a Mini to maintain maximum turbo indefinitely under 100% load with a 6 core CPU is asking too much. If that was the objective then they wouldn't be called turbo, or boost.. they'd just be the stock clocks.

I couldn't care less if you return your i5, but I did get some entertainment from your concern about this and the associated temps.

 

[Dr. Stealth]

Did you run some more reports during the test like intel power gadget?
Thank you

Dr. to Dr.

Thanks for this... Never saw this app before but really seems like a very nice tool!

 

[SoCalReviews]

It was his 10 minute stress test in the first video that was more interesting. In any case I have already tested the i5 Mini with Cinebench in that other thread of yours. So these earlier videos are no longer relevant.


This thread was created to ask for benchmark info, not fishing for clever snipes.


The 'impression' comes from reading off benchmark results submitted on this forum directly.

Just put a $20 laptop heatsink/fan underneath your Mini.... lower thermals achieved... problem solved.

 

[Mol1n]

Just put a $20 laptop heatsink/fan underneath your Mini.... lower thermals achieved... problem solved.

A cooling pad under the plastic cover? You tried this before?

Edit: I don't think it works.

 

[F-Train]

Just put a $20 laptop heatsink/fan underneath your Mini.... lower thermals achieved... problem solved.

Mine is going to live on the sill of the window that I keep partially open in the winter

 

[SoCalReviews]

A cooling pad under the plastic cover? You tried this before?

Edit: I don't think it works.

I've read posts from Mini owners in this forum who have done this with positive results. I haven't done it with any of my Mac Minis because I haven't had to but since much of the heat is dissipated through the aluminum case any additional air flow blowing on to it should work. You could even use small silent desk fan blowing from the sides, back or top of the Mini casing.
[doublepost=1541982017][/doublepost]

Mine is going to live on the sill of the window that I keep open in the winter

We will be able to blame you for the high carbon footprint of your 2018 Mini.

 

[F-Train]

[doublepost=1541982017][/doublepost]
We will be able to blame you for the high carbon footprint of your 2018 Mini.

Hey, I live on the 6th floor of an old New York building. I'm just letting some of the heat from the lower floors escape, and getting some healthy fresh air in the process

 

[SoCalReviews]

Hey, I live on the 6th floor of an old New York building. I'm just letting some of the heat from the lower floors escape, and getting some healthy fresh air in the process

You're not one of those "global warming deniers" are you?

 

[zandorf]

https://www.anandtech.com/show/13544/why-intel-processors-draw-more-power-than-expected-tdp-turbo

From this article, re new vendor strategy to run cpus to thermal limits:

Why do the vendors do this?

Firstly, it means that users can have turbo on all the time, and get the all-core turbo every second of every day. This means that the benchmark scores go through the roof, and it looks good on reviews, or when people are comparing numbers.

Secondly, that the products are engineered for it.​

The latter quoted sentence is what we are hoping for, i.e. that Apple and Intel have this sorted such that longevity is not degraded.

But given this is a new and possibly riskier cpu thermal management game, it is reasonable to wish for some assurance from Apple that this has been well tested.

 

[Dr. Stealth]

This may be just a little bit off topic but very related....

If anyone is interested in seeing the maximum heat, cpu speed (MHz), and Fan noise that you will ever see with your new Mini you can run this simple test. (Free)

Download Intel Power Gadget here: Intel Power Gadget

Download Temperature Gauge Pro here: Temperature Gauge Pro

Start both apps then run these commands in Terminal: (This is a MAX CPU Stress Test)

To start a 4-core test Cut and Paste in Terminal: yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null &

To start a 6-core test Cut and Paste in Terminal: yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null &

To end the test Cut and Paste in Terminal: killall yes

This test will push your CPU to the highest level it will ever achieve, and the max temps you will ever see, and the max noise your fan will ever make. The CPU will throttle at 100c and decrease in speed to stay below 100c, as designed. So no worry about damaging anything. It's actually very interesting to see just how well the Mini protects itself, and can give you some piece of mind.

The Fan will be pegged at 4,400 RPM which is maxed out. For those concerned about noise.

 

[PJivan]

https://www.anandtech.com/show/13544/why-intel-processors-draw-more-power-than-expected-tdp-turbo

From this article, re new vendor strategy to run cpus to thermal limits:

Why do the vendors do this?

Firstly, it means that users can have turbo on all the time, and get the all-core turbo every second of every day. This means that the benchmark scores go through the roof, and it looks good on reviews, or when people are comparing numbers.

Secondly, that the products are engineered for it.​

The latter quoted sentence is what we are hoping for, i.e. that Apple and Intel have this sorted such that longevity is not degraded.

But given this is a new and possibly riskier cpu thermal management game, it is reasonable to wish for some assurance from Apple that this has been well tested.

That has not much to di with temperature but more like “cheating” benchmarks, cpu will throttle once it hits thermal treshold regardless of motherboard power settings.
Not sure what assurance Apple needs to give you outside normal warranty and running components under manufacturing requirements. This would be different if Apple would allow (not even sure is possible) to run the cpu above the JT of 100c

 

[zandorf]

That has not much to di with temperature but more like “cheating” benchmarks, cpu will throttle once it hits thermal treshold regardless of motherboard power settings.
Not sure what assurance Apple needs to give you outside normal warranty and running components under manufacturing requirements. This would be different if Apple would allow (not even sure is possible) to run the cpu above the JT of 100c

I posted that to explain to the "anything over 90C is BAD" folks that their old fashioned notion about CPU thermal control is now being replaced by a push-it-to-the-limit approach.

I actually trust Apple to get this right. But even though redline in your car is an "OK" limit, driving around in first gear redlined may very well shorten its life. I don't know if this analogy applies here, and I suspect most posters here don't really know either. Thus my plea for some actual expert reassurance.

 

[F-Train]

I've read this whole thread and I want to thank all the contributors. It has been very helpful in deciding whether thermal issues should be a factor in deciding whether I should keep the i5 or exchange it for an i7.

My personal conclusion is that thermal issues aren't a factor, and in any event I have a three year AppleCare+ warranty on a computer that I am likely to sell in three to five years.

I just can't get excited about this.

Nice to be able to eliminate this issue as a consideration.