Intel® Core™ i5-8600 65w vs. 95 w

All about them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Sun May 27, 2018 6:19 am

i'm building a new desktop.
trying to decide between this 2 cpu's:
https://ark.intel.com/products/129937/I ... o-4_30-GHz

https://ark.intel.com/products/126685/I ... o-4_30-GHz

beside of the difference in Processor Base Frequency there is also one in TDP - 65 w vs. 95 w
i don't care to spend more money for better performance.
the only thung that bother me is if the 95 w will make the pc work more hot and therefore more fan working= more noice.
or maybe i'm wrong about it and the opposite will happen.
appreciate if someone can clear this subject to me.

Abula
Posts: 3662
Joined: Tue Nov 02, 2004 12:22 pm
Location: Guatemala

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by Abula » Sun May 27, 2018 7:34 am

Either CPU will perform well, the 8600K has a higher base clock at 3.6 vs 3.1 ghz, up to you if its worth it. As long as you get a decent tower cooler you will be able to cool it quietly. I prefer the K series, even if you dont overclock out of better resell value and in case down the road you need a little more performance you can oc it.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Sun May 27, 2018 8:04 am

i was thinking to get the 8600k for the reasons you wrote but i've been told by one of the stores that it will definitely run warmer then the 8600
as i said i'm not working usually with very heavy applications.

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by Olle P » Mon May 28, 2018 3:28 am

The motherboard will allow the two to draw different amounts of power.
With a higher limit the K-version will make use of more power and reach a higher temperature if/when exposed to a sizeable workload that the non-K-version can't complete in a few seconds.

Extended low load and/or a high load with limited duration (few seconds) will make the difference between the two diminish.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Mon May 28, 2018 3:53 am

i apologize but this sentence is not clear to me.
can you please simplify it to me ?
"With a higher limit the K-version will make use of more power and reach a higher temperature if/when exposed to a sizeable workload that the non-K-version can't complete in a few seconds."

Vicotnik
*Lifetime Patron*
Posts: 1831
Joined: Thu Feb 13, 2003 6:53 am
Location: Sweden

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by Vicotnik » Mon May 28, 2018 4:13 am

The faster CPU will get hotter at full load, but if the load is momentary and not sustained the heat will even out. The faster CPU will to the task quicker, a hotter but smaller burst. The slower CPU will do the same task over a slightly longer time, a less hot but longer burst. Total power will be about the same, and temp will also be about the same.

Load both CPUs to the max and the faster CPU will get hotter but do more work. The slower will be slightly cooler but will do less work.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Mon May 28, 2018 4:19 am

thank you/
i convinced to go on the K version

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by Olle P » Wed May 30, 2018 3:33 am

nimo11 wrote:... can you please simplify it to me ?
"With a higher limit the K-version will make use of more power and reach a higher temperature if/when exposed to a sizeable workload that the non-K-version can't complete in a few seconds."
Vicotnik wrote:The faster CPU will get hotter at full load, but if the load is momentary and not sustained the heat will even out. ... Total power will be about the same, and temp will also be about the same.
That's not exactly what I meant...

The info below assume that sufficient cooling is provided, so that temperatures become irrelevant.
a) The "maximum" power consumption of a CPU is limited by its TDP and that number is read and acted upon by the motherboard's chip.
b) The CPU is allowed to draw more than the TDP by using the maximum boost frequency, but only for a few seconds at a time. If the task at hand is not completed in those few seconds the speed will be throttled so that the power consumption is at TDP. With higher TDP the CPU will thus draw more when a more extended workload is present.

What counts as "sufficient cooling" will of course also depend on the TDP and workload.
If you only expose the CPU to "quick" tasks at sizeable intervals there doesn't need to be much cooling, but if the CPU load is higher the TDP is the minimum design target for the cooling solution.

BrianF
Posts: 151
Joined: Sat Apr 16, 2005 2:59 am

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by BrianF » Fri Jun 01, 2018 6:43 am

Olle P wrote:b) The CPU is allowed to draw more than the TDP by using the maximum boost frequency, but only for a few seconds at a time. If the task at hand is not completed in those few seconds the speed will be throttled so that the power consumption is at TDP. With higher TDP the CPU will thus draw more when a more extended workload is present.
I'm experimenting right now with an 8500 and it WILL, if asked to, go well over its 65W TDP indefinitely. With an unrealistic P95 running, TDP is over 100w without any throttling. :)

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Fri Jun 01, 2018 7:50 am

did you checked the temp. difference between 65w and 100?

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by lodestar » Fri Jun 01, 2018 9:53 am

TDP is Thermal Design Power, so 65w, 95w etc are thermal watts, that is measures of heat not electrical watts. For Intel processors the intended use of TDP is to give a relative indication of how much cooling is required in normal usage. TDP has nothing to do with how much power a CPU uses. The limit on how much power a CPU can use is determined ultimately by how hot it gets. At a certain level the CPU will throttle. This explains the extremely high overclock figures you see when liquid nitrogen cooling is used.

Differing TDPs within the same processor family are not a measure of relative performance. For example, the Coffee Lake i5-8400, i5-8500, i5-8600 and i7-8700 all have a TDP of 65w. A 95w TDP is essentially just a warning flag that this processor will potentially need more cooling. But if you use almost any reasonable third party cooler, particularly a 160mm tower cooler with a 120mm PWM fan then a 95w TDP will not be an issue.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Fri Jun 01, 2018 10:57 am

thank you.
i admit i thought it's electrical power wars.
so if i understand well a higher tdp will run at higher temp. all the time with no correlation to the pc load, ?
if this is the situation will it demand more cooling that may cause more noise?
that lead me to a question if to to give up some cpu performance for better cooling/noise results.

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by lodestar » Fri Jun 01, 2018 11:36 am

From a quietness point of view you can always trade higher CPU temperatures for less CPU cooler fan (and case fan) noise. The main issue is that you need to ensure that the CPU remains below the throttling point. This should not be difficult given the current level of BIOS and software fan control available. So the plan should be to set up the CPU and case cooling to allow the processor to achieve its full performance while maintaining a stable temperature under load conditions. On a practical level I have found that exhaust fan configuration has a big influence on how hot the CPU will get under load. Using a PWM exhaust fan and linking it to the CPU fan, that is running it at the same PWM % duty cycle can be a big help in containing CPU temps. This linking can be done with a PWM Y cable or some Asus motherboards have a CPU_OPT PWM fan header which will do the same thing. A case fan PWM header could also be used with the CPU temperature selected as the control source.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Fri Jun 01, 2018 11:58 am

sorry but my English is not perfect.
it'll help me a lot if you please approve or dis approve what i assuming:
"higher tdp will run at higher temp. all the time with no correlation to the pc load, ?
if this is the situation will it demand more cooling that may cause more noise?"

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by lodestar » Fri Jun 01, 2018 12:34 pm

Yes, more CPU cooling will mean more higher fan speeds which will mean more noise. But there are ways to keep the noise down. For example replacing the stock Intel cooler with a third-party cooler like the Noctua NH-U12S. And if your case has case fans plugging them into the motherboard and letting the motherboard control their speeds.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Fri Jun 01, 2018 1:51 pm

of course i'm planning to use third party cpu cooler.
apologize but i feel that i steel didn't get an answer for the main purpose of my question .
maybe i didn't put clear enough so i'm sorry about it.
is 95 w cpu will run hotter under all circumstances then a 65 w cpu, or it dependence on the loan on it?

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by lodestar » Fri Jun 01, 2018 2:10 pm

It is dependant on the load.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Fri Jun 01, 2018 2:17 pm

thank you.
so if my load will be not heavy i may get temperatures on the 95w not higher then on the 65 w?

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by lodestar » Fri Jun 01, 2018 2:46 pm

Yes, that's right.

nimo11
Posts: 266
Joined: Wed Dec 09, 2009 3:58 am
Location: tel aviv

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by nimo11 » Fri Jun 01, 2018 3:19 pm

great, thank you.

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by Olle P » Tue Jun 05, 2018 12:21 am

BrianF wrote:I'm experimenting right now with an 8500 and it WILL, if asked to, go well over its 65W TDP indefinitely. ... TDP is over 100w without any throttling. :)
How do you measure the CPU's power consumption?
What motherboard do you have?
Some high end motherboards have a setting that by default allow the CPU to stay a bit above TDP as long as there's sufficient cooling.
lodestar wrote:TDP is Thermal Design Power, so 65w, 95w etc are thermal watts, that is measures of heat not electrical watts. ...
There's practically no difference!
It's just that the electric power fed to the CPU is all transformed into thermal power by the work done in the CPU.
The TDP does exclude power lost outside the CPU package, in for example the power supply and voltage regulators.

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by lodestar » Tue Jun 05, 2018 1:19 pm

Olle P wrote:There's practically no difference! It's just that the electric power fed to the CPU is all transformed into thermal power by the work done in the CPU.
TDP for Intel desktop CPUs is simply intended to be a relative guide as to how much cooling a particular CPU needs. It is quoted as a maximum figure and is intended primarily for OEMs. It has nothing to do with power consumption. CPUs do not reference the TDP number is any way when operating. The maximum sustainable speed and power draw of a CPU is primarily determined by how much cooling is available. It is not true that there is practically no difference between TDP and power consumption. CPUs with the same TDP number can and do have significantly different power consumption under load. For example the i3-8100 has the same TDP (maximum 65w) as the i5-8400. A reviewer tested both of these CPUs on the same system. Under load the system power consumption for the 8100 was 105w and for the 8400 143w.

BrianF
Posts: 151
Joined: Sat Apr 16, 2005 2:59 am

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by BrianF » Wed Jun 06, 2018 3:31 am

Olle P wrote:
BrianF wrote:I'm experimenting right now with an 8500 and it WILL, if asked to, go well over its 65W TDP indefinitely. ... TDP is over 100w without any throttling. :)
How do you measure the CPU's power consumption?
What motherboard do you have?
Not power consumption per sé, but TDP as reported by Intel XTU (although the two are intrinsically linked as you aptly point out). I'm using an Asus H370-F which, interestingly, defaults to "unlimited" for both Turbo Boost Power Max and Turbo Boost Short Power Max (most of the non-Z boards I've looked at come with those hard set at 95/107, a few really crappy ones at 65/95). Even though this BIOS has settings for per-core-count multiplier, I of course cannot set it with this CPU, yet even at the stock all-cores-loaded cap of 3.9Ghz it can chew up a ton of juice and put out a commensurate amount of heat (IF asked to do so). This is in line with Tom's findings in their original review of the 8400: with an admittedly unrealistic load, these chips will happily burn over 100W of TDP...IF the motherboard doesn't reign them in (and you have a half decent HSF). But even under more realistic heavy load (rendering a Blender animation for example), I'm still seeing sustained TDP figures well above that generalized "65" spec (I can get you exact figures tonight).

I will say that this is the first Intel CPU, in a long line I've had, for which the stock cooler (IMO) is inadequate: even in a nominally decent (cool) case/environment, it has to ramp up to its max (~3200rpm) while just barely keeping the chip from throttling under test-type load. Thats not everyday use of course, but I'd like to see a little more margin there, regardless of use-case.

To get back to the OP, the bottom line is that IF you do not overclock (or invoke MCE), then the power consumption (and/or heat generated) by the 8600 and 8600K will be practically identical: the all-core clock limits of the two chips (the only speed which really matters in the real world) is the same 4.1Ghz.

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by Olle P » Thu Jun 07, 2018 4:07 am

lodestar wrote:TDP for Intel desktop CPUs is simply intended to be a relative guide as to how much cooling a particular CPU needs. ...
That's true in general.
If you look at a range of CPUs they typically have the same TDP but not the same power consumption.
But... the TDP is essentially the same as the maximum power consumption averaged over some time period for the top end of that range!

How this works, and the result of it: https://youtu.be/CSI6N6RKd5A?t=47s

lodestar
Posts: 1683
Joined: Fri Aug 05, 2005 3:29 am
Location: UK

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by lodestar » Thu Jun 07, 2018 7:47 am

Intel Desktop CPU TDP numbers cause confusion because they are quoted in watts. This leads to the assumption that it's linked to power consumption, that is electrical watts when in fact it is about the heat in thermal watts that the CPU produces when it consumes power. As Intel put it themselves* "...Thermal Design Power (TDP) should be used for processor thermal solution design targets. TDP is not the maximum power that the processor can dissipate. The processor TDP specification is very important because your thermal solution (fans, heat sink, etc.) must be able to be able to dissipate the rated TDP value and keep the processor within its thermal specifications". Now I think this advice is aimed primarily at OEMs. And as a result Intel tend not to have an individual TDP figure for every processor but as you say cover several different CPUs of varying power with the same TDP. So for example 65w is applied to a wide range of CPUs. But all Intel are saying is that an 'up to' 65w thermal solution is needed to allow this range of CPUs to operate reliably. 'Reliably' is defined in thermal terms, below the temperature at which the CPU throttles. I get the point that there must be some link between the power a CPU consumes and the heat it produces. But if the same TDP covers a variety of CPUs then it must involve a range of system power draws and maximum CPU temperatures under load then TDP is not it.

For that reason, apart from OEMs TDP does not seem to me to be of much value. Partly because most people that assemble their own PCs use performance motherboards from manufacturers such as Asus and MSI rather than OEM parts. Some of these, generally the 'Z boards' have features that make TDP numbers even more meaningless than they are already. An example is Multi-Core Enhancement(MCE) that runs all cores at the maximum Turbo mode, and to do it typically increases the CPU voltage a touch. MCE is not a problem if a reasonable third-party cooler is used. However, if you are not aware of it and have decided to use the stock Intel cooler you may find the extra power consumption and heat will be an issue.

In the context of this thread, the 65w TDP i5-8600 vs the 95w TDP i5-8600K, MCE will turbo all the cores of both CPUs at 4.3Ghz. There will no difference between them in power consumption or heat generation or the cooling needed if only Turbo modes are used. The 8600K can of course be overclocked and I think this is why it has the 95w TDP, to simply indicate that it needs a beefier cooling setup to allow for overclocking. Neither the 65w or 95w TDP numbers have any relation to the power consumed by these processors using either MCE or overclocking in the case of the K.

(* Intel White Paper Measuring Processor Power)

BrianF
Posts: 151
Joined: Sat Apr 16, 2005 2:59 am

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by BrianF » Thu Jun 07, 2018 9:12 am

I'll just add for clarity that a watt is a watt is a watt. If a CPU, or a light bulb for that matter, is giving off 100Watts of heat...then it is also consuming 100watts of electricity.

The confusion arises when people look at TDP and think "that is exactly the maximum this CPU will consume" when in fact it could be more, it could be less. As has been said, it is a guideline only, or as Intel puts it an "average under predefined test load". As such it is, IMO, an overly broad spec. The proof is in the 8400: it WILL consume (or give off in the form of heat) over 100 watts, and NOT with any MCE or other clocking shenanigans! This is why I take exception to cheap coffee lake boards which have a sustained power delivery cap of 65 watts (and yes some do): they might be fine for "every day" use, but they cannot supply even an 8400 with enough juice to sustain an esoteric load to the best the CPU is capable of.

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Re: Intel® Core™ i5-8600 65w vs. 95 w

Post by Olle P » Fri Jun 08, 2018 5:27 am

BrianF wrote:... This is why I take exception to cheap coffee lake boards which have a sustained power delivery cap of 65 watts (and yes some do): they might be fine for "every day" use, but they cannot supply even an 8400 with enough juice to sustain an esoteric load to the best the CPU is capable of.
Or one can see it from the other aspect:
The motherboards that don't keep it at 65W will cause the cooling to run louder than expected when optimised for that power level.

Post Reply