• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Power consumption and efficiency in gaming: AMD vs Intel

Associate
Joined
14 Aug 2013
Posts
206
As far as idle wattage goes, the last intel chip I had, the 7700k, used to auto down clock the cores to pretty much nothing when idle. my 7600 doesn't seem to ever go below 3ghz, which might explain the differences in idle usage to some extent.
 
Associate
Joined
5 Mar 2017
Posts
2,255
Location
Cambridge
The difference about GPUs is that you can’t have 4090 performance at 100W. Not for the next few generations at least. But you can have top performance for gaming at 100W, how justify the 250-300W?
99% of the time my GPU is at 200-220W, CPU 70 ish W at most.
But the 360 AIO as requirement to avoid thermalthrottle if left at its own desire is poor. Intel could be 5% slower at same power. Or same performance for 5% more power. But as it is, not worth for gaming at least.
 
  • Like
Reactions: G J

G J

G J

Associate
Joined
3 Oct 2008
Posts
1,404
@mattyfez I've got good results undervolting my 13600K as it uses about 125w which was just abit lower that the 8700K I had.

@Drumroll Intresting take as I'd not thought of it in that way but I agree.
 
Soldato
Joined
28 Oct 2002
Posts
9,479
Location
Returning some videotapes
Sorry to resurrect this thread but it is pretty useful.
Looking at an UNRAID setup to use a setup with decent NVME support I will need to go for a non embedded CPU + with a decent integrated iGPU.

It looks like the i3-13100 or i3-14100 fit that bill + have the low idle power consumption. Is there a comprehensive comparison of idle CPU power consumption anywhere about (I have looked).

(I actually have the option of a cheap 5700G which might fit the bit going by the OPs video), it just doesnt play as nice/as powerful for Plex transcoding.
 
Last edited:
Man of Honour
Joined
22 Jun 2006
Posts
11,671
Sorry to resurrect this thread but it is pretty useful.
Looking at an UNRAID setup to use a setup with decent NVME support I will need to go for a non embedded CPU + with a decent integrated iGPU.

It looks like the i3-13100 or i3-14100 fit that bill + have the low idle power consumption. Is there a comprehensive comparison of idle CPU power consumption anywhere about (I have looked).
hwcooling.net did an article here on the difference between the two 12400 versions:

computerbase.de often includes these numbers in their reviews:

Historically ASRock has been pretty good, Kontron are probably better, high-end motherboards are usually bad.

Ryzen CPUs have higher idle power consumption with XMP/DOCP enabled and/or memory above 2666. I believe 12th-14th gen also have this issue, but I haven't seen much testing on it.

Threads like this can be helpful, but you have to dig a bit for the info you need. NAS forums are generally good sources.
 
Man of Honour
Joined
22 Jun 2006
Posts
11,671
Wow fairly big changes between Intel Core i3-14100F, i3-13100F and i3-12100F, with the 13th gen coming off far better.
Just a word of warning: most reviewers don't really care about power consumption and they're even more indifferent to idle power, which means they rarely specify important information like motherboard and memory changes (including BIOS updates) and with an -F CPU since it doesn't have integrated graphics: the graphics card and driver version.

To be more explicit: I am very suspicious of idle numbers that show CPUs with the same die (12100, 13100 and 14100 have the same die and the same amount of cache) that show significant differences in idle power, since with Intel CPUs the variance is usually a few watts at most.
 
Soldato
Joined
28 Oct 2002
Posts
9,479
Location
Returning some videotapes
Just a word of warning: most reviewers don't really care about power consumption and they're even more indifferent to idle power, which means they rarely specify important information like motherboard and memory changes (including BIOS updates) and with an -F CPU since it doesn't have integrated graphics: the graphics card and driver version.

To be more explicit: I am very suspicious of idle numbers that show CPUs with the same die (12100, 13100 and 14100 have the same die and the same amount of cache) that show significant differences in idle power, since with Intel CPUs the variance is usually a few watts at most.
It does look somewhat suspicious.

I am somewhat surprised there isn't much of a comparison at idle given the rise in NAS/personal servers these days.
 
Man of Honour
Joined
22 Jun 2006
Posts
11,671
I am somewhat surprised there isn't much of a comparison at idle given the rise in NAS/personal servers these days.
Yeah, the best source I've found is when a thread like the one I linked does a spreadsheet comparing the numbers that their users report. I've never found the reviews particularly helpful because they're very hard to compare when they don't keep the same test system. I think there is almost certainly a significant difference between H0 and C0 CPUs, but other than that, it is hard to say anything reliably. I've even seen some claims that CPUs with E-Cores idle lower than CPUs without them, but I've never seen any evidence of that.
 
Soldato
Joined
28 May 2007
Posts
18,262
It does look somewhat suspicious.

I am somewhat surprised there isn't much of a comparison at idle given the rise in NAS/personal servers these days.

Idol CPU power use is largely irrelevant in this market. Its over all system power at the wall that matters and that’s largely down to everything else including software.
 
Soldato
Joined
28 Oct 2002
Posts
9,479
Location
Returning some videotapes
Yeah, the best source I've found is when a thread like the one I linked does a spreadsheet comparing the numbers that their users report. I've never found the reviews particularly helpful because they're very hard to compare when they don't keep the same test system. I think there is almost certainly a significant difference between H0 and C0 CPUs, but other than that, it is hard to say anything reliably. I've even seen some claims that CPUs with E-Cores idle lower than CPUs without them, but I've never seen any evidence of that.

I saw a spreadsheet on that forum but looked a but outdated unfortunately, but looking through the thread I do see mention of the 12100 and its idle consumption. :)


Need ChatGPT to trawl the web and create a google sheet pulling it all together (multiple forums). ;)
 
Soldato
Joined
28 Oct 2002
Posts
9,479
Location
Returning some videotapes
Idol CPU power use is largely irrelevant in this market. Its over all system power at the wall that matters and that’s largely down to everything else including software.
What if its running 24/7 tho. I think its pretty important tbh, however my case might be particular since I am building a fanless server so lower TDP and idle consumption are important.
 
Caporegime
Joined
17 Mar 2012
Posts
47,668
Location
ARC-L1, Stanton System
If CPU A is pulling 300 watts under load with CPU B pulling 150 watts what what does it matter if CPU A pulls 5 Watts less at idle?

I mean.... do you use your computer or do you like to stare at it all day every day?
 
Soldato
Joined
28 May 2007
Posts
18,262
What if its running 24/7 tho. I think its pretty important tbh, however my case might be particular since I am building a fanless server so lower TDP and idle consumption are important.

The CPU is going to change the profile, but other parts are much more important.

I’ve been building min max systems recently (just because I can and out of interest) and my lord are they complicated builds and efficiency (especially the last 10%) is stupidly expensive.
 
Soldato
Joined
28 Oct 2002
Posts
9,479
Location
Returning some videotapes
If CPU A is pulling 300 watts under load with CPU B pulling 150 watts what what does it matter if CPU A pulls 5 Watts less at idle?

I mean.... do you use your computer or do you like to stare at it all day every day?

Its for an UNRAID server/homelab so will be running docker containers - Plex, OpenVPN, Sonarr, Radarr, Nextcloud etc. This isn't my main rig.

Also I will be using a Streacom fanless case so efficiency key - silence is golden!

The CPU is going to change the profile, but other parts are much more important.

I’ve been building min max systems recently (just because I can and out of interest) and my lord are they complicated builds and efficiency (especially the last 10%) is stupidly expensive.

I guess I haven't got to the rest yet, as long as I meet the threshold for the PSU then I'm happy. Given the CPU has an iGPU it is going to help a lot.
 
Last edited:
Man of Honour
Joined
22 Jun 2006
Posts
11,671
What if its running 24/7 tho. I think its pretty important tbh, however my case might be particular since I am building a fanless server so lower TDP and idle consumption are important.
The motherboard and PSU are usually more important than the CPU to the overall power use @ idle (assuming the same gen/socket of CPU) and if the power consumption is very low, they're much more important.
 
Last edited:
Soldato
Joined
28 May 2007
Posts
18,262
The motherboard and PSU are usually more important than the CPU to the overall power use @ idle (assuming the same gen/socket of CPU) and if the power consumption is very low, they're much more important.

The PSU is critical and lower the system power draw the more complex the more difficult the calculations become. I’ve been through 9 PSU’s.

I’d also test at the power at least at the wall socket pre PDU surge device.
 
Last edited:
Back
Top Bottom