Question Real world power consumption of Intel 13 and Ryzen 7000

[vacip]

Hello,
I'm looking to upgrade to a new system, and the first and hardest decision is CPU.
The Intel 13th gen non-k paper launch was very very disappointing, so I'm looking at either i7 13700k or the Ryzen 7900 (yes, the non-x version with a decent MOBO so that I can overclock if I need in 5-7 years).

I'm baffled that all around the internet all I see are "All core full synthetic load power consumption" benchmarks. This is cute, but completely useless. In reality, my CPU will be running idle (web browsing, online meetings, emails, light MS Office, background downloads etc) for 8-12 hours a day, then be used for some gaming, some Airvideo encoding (sometimes in parallel), occasionally some video/photo editing.

Do you know of a good real world power consumption test? I know "my mileage may vary", but I'm specifically interested in the comparison. I'm interested in power consumption while idle, during light gaming (1080p 60fps esport-like titles), and during AAA gaming.
Why doesn't anyone do this (i found literally 1 video and 0 articles)?

View: https://www.youtube.com/watch?v=JHWxAdKK4Xg

The only video I found. Or is it just my google-fu that let me down?

With a green mindset, the current energy prices, and with my plan on running this thing for 10+ years, yes, power consumption matters.
Thank you!

 

[Aeacus]

R9 7900 consumes about half the power what i7-13700K consumes, if this is what you want to know.

For specific numbers;
R9 7900 - 65W normal, 88W turbo
i7-13700K - 125W normal, 253W turbo

Source: official specs + reviews.
Intel review: https://www.tomshardware.com/reviews/intel-core-i7-13700k-cpu-review/2
AMD review: https://www.tomshardware.com/news/amd-ryzen-5-7600-cpu-review

I'm interested in power consumption while idle, during light gaming (1080p 60fps esport-like titles), and during AAA gaming.

CPU alone or entire PC as a whole?

With a green mindset, the current energy prices, and with my plan on running this thing for 10+ years, yes, power consumption matters.

I'm curious, what PSU you're planning to buy?

 

[TechyInAZ]

Sadly not a lot of reviewers test idle power consumption of chips, but i whole heartedly agree with your statement. Idle power consumption can be just as important as full load power consumption, for general usage. Since most of the time our chips do idle.

In general, I've learned over the years that Intel and AMD's idle power consumption generally remains the same, aswell as low load power consumption. Intel is known as a power hog these days, but they are also very good at keeping their chips power efficient under lower usage scenarios.

I don't know exactly or know where to find idle power consumption numbers for both chips. But the difference should be so small between the two it won't make a difference.

 

[vacip]

Hi Aeacus, thank you for the reply!

R9 7900 consumes about half the power what i7-13700K consumes, if this is what you want to know.

For specific numbers;
R9 7900 - 65W normal, 88W turbo
i7-13700K - 125W normal, 253W turbo

Source: official specs + reviews.
Intel review: https://www.tomshardware.com/reviews/intel-core-i7-13700k-cpu-review/2
AMD review: https://www.tomshardware.com/news/amd-ryzen-5-7600-cpu-review

Well, this is my problem:
Processor Base Power
The time-averaged power dissipation that the processor is validated to not exceed during manufacturing while executing an Intel-specified high complexity workload at Base Frequency and at the junction temperature as specified in the Datasheet for the SKU segment and configuration. - Intel

So the 125W normal is not an idle power consumption. Not even a light browsing workload. It is a synthetic workload that I'll likely not often reach. And it is not comparable to AMD's such number.
So these numbers are meaningless in the real world.

A bit like fuel consumption for cars: We have a lot of tests and benchmarks about fuel consumption on a freaking racetrack, (useless) and we have the official number from the manufacturer about a secret manufacturer workload that enables comparison of cars within the same manufacturer.
None of this helps...

CPU alone or entire PC as a whole?

For now, CPU alone; total system power usage is more or less the sum of all parts. But I'd be happy to see full system comparisons too; anything could be useful.

I'm curious, what PSU you're planning to buy?

Didn't start that investigation yet. Something that can comfortably power everything, and has a high efficiency.

 

[TerryLaze]

22-23W for the intel vs 41W for the 7900x doing a single core encoding,sadly they don't have the non-x one.
Gaming is 81W for the 7900x vs 89W for the i7 but the i7 is also around 8% faster so if you play with vsync on for example so that both run the same amount of FPS things will be different.
techpowerup.com/review/intel-core-i7-13700k/22.html

 

[Aeacus]

So the 125W normal is not an idle power consumption. Not even a light browsing workload. It is a synthetic workload that I'll likely not often reach. And it is not comparable to AMD's such number.
So these numbers are meaningless in the real world.

Not meaningless, but instead helpful for other aspects. Namely CPU cooler selection and also PSU selection.

As far as idle power consumption goes, there are way too many variables that skew the results and no meaningful testing can be done.

Moreover, what state of PC can be considered "idle"?
Is it when you boot into Safe Mode, with only bare minimum of services running?
Is it when you boot to OS, but doesn't do anything with PC, letting it sit by, consuming meaninglessly power? (Might as well put your PC to sleep.)
Is it when you browse around File Explorer/doing office work (e.g Excel)?
Is it when you surf the web? Perhaps listen some background music while you're at it (*.mp3 files or internet radio)?

IF given that the "idle" is when PC is booted to OS, but you don't do anything with PC, it has loads of background programs still running. OS services + 3rd party services. And not all OS'es or even their versions, consume the exact same amount of CPU either. They vary. So does the 3rd party programs that are running. <- All that combined can vary between 5W-20W. And unless you have that very same, identical PC, as the test setup is, with ALL the same 3rd party apps - this power consumption data is useless. It is only relevant for that one specific system, running those specific services.

For example:
I have i5-6600K with stock clocks, and my average power draw for my CPU currently, is ~18.5W (source: HWinfo64). This includes all my OS services, all my 3rd party background services, Firefox and me typing this reply.
So, even if you also have i5-6600K but without the very same 3rd party services as i have, my CPU power consumption is meaningless for you.

But I'd be happy to see full system comparisons too; anything could be useful.

Same issue is with whole build power consumption as well. Depending on what hardware build has, power draw differs. It even differs when fan speed is increased/lowered (when using 3-pin fans and/or DC mode).

Best anyone could do, is add up all the hardware, calculate the maximum power draw and based on that, buy proper sized PSU.

E.g my whole PC power consumption + monitor, at current moment, is from 62-85W (source: CyberPower CP1300EPFCLCD UPS). My build that i took as an example here, is Skylake, full specs in my sig.

Though, based your usage, PC power draw will differ. Same is with car fuel consumption. If you are a lead foot, your car fuel consumption will be higher than that of manufacturers. And even when you drive sensibly, manufacturer, for the most part, has tested their cars in ideal conditions. While out of the street is anything but ideal conditions.

Didn't start that investigation yet. Something that can comfortably power everything, and has a high efficiency.

Unless you do not look towards 80+ Titanium efficiency PSU (highest there is), you'll be wasting quite a bit of electricity as excess heat.

80+ Titanium efficiency PSU will cost you quite a lot. Mine did. I have two Seasonic PRIME TX-650 PSUs in use (one powering Skylake build, another powering Haswell build), while my 3rd PC (AMD) is powered by 80+ Platinum PSU (Seasonic PX-550).
SSR-650TD inside Skylake build, costed me €206.80, 7.5 years ago.
SSR-650TR inside Haswell build, costed me €205.50, 2 years ago.

So, unless you aren't willing to fork out premium price for 80+ Titanium PSU, you can throw your "green mindset" out of the window.

And it doesn't end there.

E.g case fans.
Are you going to buy case fans based on: eyecandy? performace? or power draw?

Most people buy case fans based on eyecandy. I bought mine based on performance.
For example, RGB/ARGB LED fans, will consume more power than fans without LEDs. Since LEDs also draw power.
Here, are you willing to buy non-LED hardware? Since all that illumination does draw additional power.

And it still doesn't end there;

so that I can overclock if I need

CPU OC is huge waste of electricity, with little, if any gains. Since CPUs are so efficient today, best you could get would be ~200Mhz increase over Boost clocks. That's neglible difference, but power consumption will be very high. So.... where's the green mindset regarding that?

-----

All-in-all, at some point, you have to draw a line between power savings and hardware performance. You can not have both: low power consumption and high performance.

E.g i5-13400 is solid, 65W CPU. Doesn't consume nowhere near as much power as i7-13700K does. And since it isn't K-series either, it can't be OCd, removing the extreme power hogger (CPU OC) out of equation.

The Core i5-13400's power consumption is surprisingly good, especially given that our test chip uses the Alder Lake C0 die. The Ryzen 5 7600 is more power efficient in our renders-per-day metric and consumes less power in every workload, but the deltas aren't large enough to be of concern for most users. The Core i5-13400 is very power efficient in these workloads, largely due to the addition of four e-cores. That confers a generational improvement against the Core i5-12400 in both performance and power consumed.

Source: https://www.tomshardware.com/news/i...f-thermals-power-consumption-and-boost-clocks

So, why not go with far less power consuming i5-13400, instead of your power hogger i7-13700K/R9 7900?

 

[vacip]

I like your thinking.

As far as idle power consumption goes, there are way too many variables that skew the results and no meaningful testing can be done.

Moreover, what state of PC can be considered "idle"?

I agree that there are many factors, but I think technically it would be possible to have close enough comparisons by installing the same image on a Ryzen and an Intel machine, and take power readings while the same programs are running. Just like with full synthetic load testing, but this time instead of a zip test, just run a Chrome refreshing the same tab every 5 seconds. Anyhow, I don't have the resources journalists have, I'm just baffled/annoyed nobody did such testing.

So, unless you aren't willing to fork out premium price for 80+ Titanium PSU, you can throw your "green mindset" out of the window.

Thank you! I'll look into some tests, and if it really is that much more efficient, I'll get an 80+ Titanium PSU.

Are you going to buy case fans based on: eyecandy? performace? or power draw?

I'm not in my teens or 20s anymore... Rainbow led irritates me. Computer is under my desk, I'll go for good airflow and good dust filtration. No led. Silence is important; air cooling, I don't trust liquid solutions to last 10 years without maintenance need. If possible, passive cooling while idle is preferred, but this is nice to have - to be researched later.

CPU OC is huge waste of electricity, with little, if any gains. Since CPUs are so efficient today, best you could get would be ~200Mhz increase over Boost clocks. That's neglible difference, but power consumption will be very high. So.... where's the green mindset regarding that?

Good point. I am looking to build a system and then use it for 10+ years. This is where the green mindset is. This "buy buy buy buy new new new" mentality somewhat disgusts me. I am upgrading from an i7 3770 (non-k) system after 10 years (I swapped video cards after 6 years from a 7870 to a used 970, that's all). The last ~2 years were a bit shaky on this system, if I had a 3770k, that could have extended the comfortably usable life of the system by 3-4 years with another VGA upgrade. So having an overclockable system gives me an extra few years on that system. I might be wrong, but I figured less e-waste>>a bit more energy usage [insert "doubts about overengineered led lightbulbs that rarely last 10 years as advertised VS locally manufactured simple incandescent bulbs" here].

Also, OC (heat) reduces system life expectancy, so I'd only use it in rare occasions, towards the end of my system's lifespan.

So, why not go with far less power consuming i5-13400, instead of your power hogger i7-13700K/R9 7900?

As above, I intend on using this system for 10+ years, with 1-2 VGA upgrades down the line. For this, I need to go big.


But, i5-13400, non-k? Do these things actually exist? I didn't find any for sale outside of prebuilts. I'll look into them again. On the other hand, Alder lake is last gen, which means -1 year of expected lifespan.

Thanks for provoking some thoughts though! I'll look into intel non-k again, definitely get a great PSU, and maybe investigate e-waste VS energy consumption; last I checked there was very little research into the topic (I grudgingly assume because it would go against the current economy doctrine of buying brand new... everything... every 2-3 years).

 

[Aeacus]

I'm just baffled/annoyed nobody did such testing.

Well, no point to do such test since Core i7/Ryzen 9 isn't known for their efficiency. And desktop PCs doesn't have issues with power delivery either.
This kind of test is only useful where power delivery is limited, namely laptops and the like.

Thank you! I'll look into some tests, and if it really is that much more efficient, I'll get an 80+ Titanium PSU.

Compared to 80+ Gold, 80+ Titanium is more efficient, 4% to be exact.
Wiki: https://en.wikipedia.org/wiki/80_Plus

4% for some is little, while for others, it's quite a bit.
Like i said above, i'm running 80+ Titanium units. Namely because my PCs are backed up by an UPS (one UPS per PC) and better efficient PSU = longer runtime of my UPS. That, and i don't like waste power as excess heat due to the inefficiency, hence why i bought 80+ Titanium PSUs. Both of my Seasonic PRIME PSUs were the very best 650W PSUs, money could buy, at the time of purchase. And even today, PRIME TX is one of the very best (if not the best) PSU out there (depending on who you ask).

I'll go for good airflow and good dust filtration. No led. Silence is important; air cooling, I don't trust liquid solutions to last 10 years without maintenance need. If possible, passive cooling while idle is preferred, but this is nice to have - to be researched later.

Fans wise, look towards Noctua then. Simple as that. Noctua is the only one, in consumer market, who is consistent across the board (all their fans perfom as advertised and they don't have any weak/poor fans). Arctic Cooling is also good, and their fans are known for quiet operation (i had few of Arctic Cooling fans myself, until i went with high-end fans for optimal performance and longest lifespan).

I am looking to build a system and then use it for 10+ years.

So having an overclockable system gives me an extra few years on that system.

Back in the day, with older CPUs, CPU OC was worthwhile.

E.g i have i5-6600K with 3.5 GHz base and 3.9 Ghz boost. With CPU OC, i could get it 4.5 Ghz all core (increase of 600 Mhz over boost), or with delid, ~4.7 Ghz all core (800 Mhz over boost). And there have been some delidded i5-6600K CPUs, that can hold 5 Ghz all core.

Your planned CPU, i7-13700K, is so efficient, that most chips out there, can only hold all core stable at 5.5 Ghz (100 Mhz over boost). With this, there is so little OC headroom, that CPU OC, IMO, isn't worthwhile.
Here's TH review of i7-13700K,
link: https://www.tomshardware.com/reviews/intel-core-i7-13700k-cpu-review/2
where TH got 5.5 Ghz with 280mm rad AIO. TH review also says that depending on silicone lottery and even more cooling, 360mm rad AIO, CPU OC could hold stable at 5.7 Ghz (300 Mhz over boost).

So, 100-300 Mhz OC over boost clocks is so little, that CPU OC with current, highly efficient chips, isn't worthwhile. There won't be any meaningful performance increase.
If the headroom would be bigger, like it is with my 6th gen CPU, where on minimum, i look towards 600 Mhz increase over boost clocks (or up to 1.1 Ghz over boost, if very lucky with delidded chip), then CPU OC makes sense.

All-in-all, CPU OC is dying niche and outside of record breaking, isn't worth the effort anymore.

But, i5-13400, non-k? Do these things actually exist?

Sure it does. It's the latest CPU from Intel, released week ago or so.

I didn't find any for sale outside of prebuilts.

Plenty of buying options. Amazon, Egg, etc, pick your poison ,
pcpp: https://pcpartpicker.com/product/7xLFf7/intel-core-i5-13400-25-ghz-10-core-processor-bx8071513400

and maybe investigate e-waste VS energy consumption; last I checked there was very little research into the topic (I grudgingly assume because it would go against the current economy doctrine of buying brand new... everything... every 2-3 years).

Current economy, in general, is yes, buy new item. Especially since repairs have been made so complex that sending the item into repairs, often costs more than brand new item.
But with research, one can buy solid hardware that lasts for years or even decades. E.g i've done a lot of research before i've bought my PC hardware and while my system has some years on it already (~6.5 years), my i5-6600K is still going strong and i haven't had a need to OC it or otherwise upgrade the CPU. Only upgrade was GPU, to match the monitor upgrade, but other than that, none of the hardware that i have in all of my PCs, haven't died. Nor are they performing below our needs.

Oh, almost forgot;
Following goes into the niche of enthusiast computing and is to do with dust filtration. So, if you're interested about proper dust management, expand the spoiler.

Spoiler: Click here to view

If you checked out my builds from my signature, you probably noticed from my images, that i have custom fan filters on my Corsair 760T V2 Black and Corsair 750D Airflow Edition cases. Namely, i have DemciFlex filter sets, custom made to my PC cases,
link: https://www.demcifilter.com/

E.g here's set for;
Corsair 760T: https://www.demcifilter.com/demciflex-magnetic-dust-filter-for-corsair-760t-dust-filter-kit
Corsair 750D: https://www.demcifilter.com/demciflex-magnetic-dust-filter-for-corsair-obsidian-750d-dust-filter-kit

Demciflex offers custom filters for all sorts of PC cases. Their full lineup is here, just select the case brand and look if they have custom filter kit for specific PC case you plan to go with,
link: https://www.demcifilter.com/magnetic-computer-dust-filters-for-your-computer

You can also buy standard sized filters from Demciflex, or when needed, order a custom made filter, according your own specific needs/dimensions. All filters are either magnetic or come with glued edge (to mount on plastic). Also, Demciflex now offers a wide arrange of colors as well (back then when i bought mine, options were only black or white).
And cleaning them is also very easy (guide on Demciflex site), but in a nutshell, either gently brush the dust off, or when filter is very dusty, rinse in cold water and afterwards dry with towel by gently tapping the filter.

Here's video review of Demciflex set for 760T, showcasing the filters and how fine the holes really are, compared to the stock filter;

View: https://www.youtube.com/watch?v=m3LFD_M8Ev0

One of the reasons, among many, as of why i went with 760T and 750D AF PC cases, was that i could also get Demciflex filter kit for them, for optimal dust management.

 

[Vlado1428]

R9 7900 consumes about half the power what i7-13700K consumes, if this is what you want to know.

For specific numbers;
R9 7900 - 65W normal, 88W turbo
i7-13700K - 125W normal, 253W turbo

Source: official specs + reviews.
Intel review: https://www.tomshardware.com/reviews/intel-core-i7-13700k-cpu-review/2
AMD review: https://www.tomshardware.com/news/amd-ryzen-5-7600-cpu-review



CPU alone or entire PC as a whole?



I'm curious, what PSU you're planning to buy?

Abacus. thanks the info .. and apologies to all. for writing back so late ( I had a serious family issue .. and make me forget about all this ) but all is settle down now .
7900 seems a nice cpu .. but I had a surprise when I saw that in idle and cpu consumes more than intel .. ( maybe its something AMD will fix in future updates .. ).
CPU: Yeap cpu Alone.
PSU: something around 650-750 w. brands I can get here .. are corsair .evga..(gold or platinum.) mm maybe IM forgetting some..
its possible to put a Wattage limit to CPU ?

 

[Vlado1428]

Sadly not a lot of reviewers test idle power consumption of chips, but i whole heartedly agree with your statement. Idle power consumption can be just as important as full load power consumption, for general usage. Since most of the time our chips do idle.

In general, I've learned over the years that Intel and AMD's idle power consumption generally remains the same, aswell as low load power consumption. Intel is known as a power hog these days, but they are also very good at keeping their chips power efficient under lower usage scenarios.

I don't know exactly or know where to find idle power consumption numbers for both chips. But the difference should be so small between the two it won't make a difference.

TechyInAZ. yeah agree .. it's not an easy thing to find .. moreover. its also (at least for me) what ''Idle '' means .. I always thought its power pc on monitor . windows loaded.. (but sometimes they're ar programs running in the background ..
it would be very interesting to see tests doing common stuff .. watching YouTube vid .. doing something in word, .. photoshop .. bah. or other popular programs /apps..
its like a big gap between testing full load ./stressed system gaming all ultra or rendering .. and. idle ( if it means pc turned on . ..
all this fuzz on my side its because .. honestly. .. the increasing electricity bills .. I mean. its not IM a cheap rat lol .. but .. I want to save and build an efficient system as much as I can.

 

[Vlado1428]

Hi Aeacus, thank you for the reply!



Well, this is my problem:
Processor Base Power
The time-averaged power dissipation that the processor is validated to not exceed during manufacturing while executing an Intel-specified high complexity workload at Base Frequency and at the junction temperature as specified in the Datasheet for the SKU segment and configuration. - Intel

So the 125W normal is not an idle power consumption. Not even a light browsing workload. It is a synthetic workload that I'll likely not often reach. And it is not comparable to AMD's such number.
So these numbers are meaningless in the real world.

A bit like fuel consumption for cars: We have a lot of tests and benchmarks about fuel consumption on a freaking racetrack, (useless) and we have the official number from the manufacturer about a secret manufacturer workload that enables comparison of cars within the same manufacturer.
None of this helps...


For now, CPU alone; total system power usage is more or less the sum of all parts. But I'd be happy to see full system comparisons too; anything could be useful.


Didn't start that investigation yet. Something that can comfortably power everything, and has a high efficiency.

Vacip .. yeah ''none of that helps'' I feel the same hah! .. I would be happy to see tests too cpu alone or entire system ..
ps: I understand that. motherboards .. can add their coin to power consumption .. like top of the line can consume. more than h or B .. mobos .. anyway meh IM not saying anything new

 

[Vlado1428]

I like your thinking.


I agree that there are many factors, but I think technically it would be possible to have close enough comparisons by installing the same image on a Ryzen and an Intel machine, and take power readings while the same programs are running. Just like with full synthetic load testing, but this time instead of a zip test, just run a Chrome refreshing the same tab every 5 seconds. Anyhow, I don't have the resources journalists have, I'm just baffled/annoyed nobody did such testing.



Thank you! I'll look into some tests, and if it really is that much more efficient, I'll get an 80+ Titanium PSU.



I'm not in my teens or 20s anymore... Rainbow led irritates me. Computer is under my desk, I'll go for good airflow and good dust filtration. No led. Silence is important; air cooling, I don't trust liquid solutions to last 10 years without maintenance need. If possible, passive cooling while idle is preferred, but this is nice to have - to be researched later.



Good point. I am looking to build a system and then use it for 10+ years. This is where the green mindset is. This "buy buy buy buy new new new" mentality somewhat disgusts me. I am upgrading from an i7 3770 (non-k) system after 10 years (I swapped video cards after 6 years from a 7870 to a used 970, that's all). The last ~2 years were a bit shaky on this system, if I had a 3770k, that could have extended the comfortably usable life of the system by 3-4 years with another VGA upgrade. So having an overclockable system gives me an extra few years on that system. I might be wrong, but I figured less e-waste>>a bit more energy usage [insert "doubts about overengineered led lightbulbs that rarely last 10 years as advertised VS locally manufactured simple incandescent bulbs" here].

Also, OC (heat) reduces system life expectancy, so I'd only use it in rare occasions, towards the end of my system's lifespan.



As above, I intend on using this system for 10+ years, with 1-2 VGA upgrades down the line. For this, I need to go big.


But, i5-13400, non-k? Do these things actually exist? I didn't find any for sale outside of prebuilts. I'll look into them again. On the other hand, Alder lake is last gen, which means -1 year of expected lifespan.

Thanks for provoking some thoughts though! I'll look into intel non-k again, definitely get a great PSU, and maybe investigate e-waste VS energy consumption; last I checked there was very little research into the topic (I grudgingly assume because it would go against the current economy doctrine of buying brand new... everything... every 2-3 years).

I don't know i5 13400. but IM pretty sure. there's the i5 13500 . ( is there a i513600 ?)

 

[Aeacus]

PSU: something around 650-750 w. brands I can get here .. are corsair .evga..(gold or platinum.) mm maybe IM forgetting some..
its possible to put a Wattage limit to CPU ?

Regarding Corsair, look towards RM/RMi/RMe/RMx/HX/HXi/AX/AXi. And SF as well, if you have mini-ITX build. The rest are best to be avoided.
Other good PSUs are: Seasonic Focus/PRIME (if very low on cash, then Core too) and Super Flower Leadex III series (Leadex II is also good but a bit old).

PC components draw as much power as they need. So, PSU will not output it's max wattage at all times. But you can limit the power consumption of GPU (undervolt) and CPU too if you so desire. Though, do note that bringing the voltage too low will make the component unstable.

is there a i513600 ?

There is.

Intel 13th gen CPU lineup;
Core i5: https://ark.intel.com/content/www/u...ration-intel-core-i5-processors.html#@Desktop
Core i7: https://ark.intel.com/content/www/u...ration-intel-core-i7-processors.html#@Desktop
Core i9: https://ark.intel.com/content/www/u...ration-intel-core-i9-processors.html#@Desktop

Currently, there is no Core i3 in 13th gen lineup. Perhaps in the future we'll also get Core i3.

 

[Valesti]

Just want to say that Intel power consumption at low-usage (web browsing, youtube, coding, word/excel) is MUCH lower than AMD.
I have 3700X. Only web browsing - CPU usage 1% - CPU Wattage 38-40W.
My wife has 10600K. Web browsing - CPU Usage 1% - CPU Wattage 10W.

Only focusing power consumption at heavy workload is flawed. In typical usage scenario for most people, Intel CPUs actually save a lot of power compared to AMDs in the long term..

 

[BFG-9000]

Intel does seem to idle lower for CPU only:

However for total system power 13700k idles higher than Ryzen 9 7900x and 7950x (click the image for latest article showing this):

but close enough to be a wash

 

[hotaru.hino]

What's important to point out with Intel CPUs is that:

• Everything the processor is capable of is fully integrated, where as Ryzen since Zen 2 has at least two chips, and one of them is at a higher process node.
• The I/O die on Ryzen processors supports a lot more functionality than Intel's I/O portion on their processors.
  
  Compare this:
  
  
  
  
  To Alder Lake's system diagram:
  
  

So I think when comparing idle processor power consumption, it's not really an apples-to-apples comparison because of this. Or at least, you need to remember that Ryzen processors have a lot more going on than Intel's.

 

[TerryLaze]

What's important to point out with Intel CPUs is that:

• Everything the processor is capable of is fully integrated, where as Ryzen since Zen 2 has at least two chips, and one of them is at a higher process node.
• The I/O die on Ryzen processors supports a lot more functionality than Intel's I/O portion on their processors.
  
  Compare this:
  
  
  
  
  To Alder Lake's system diagram:
  
  

So I think when comparing idle processor power consumption, it's not really an apples-to-apples comparison because of this. Or at least, you need to remember that Ryzen processors have a lot more going on than Intel's.

That's why they measure the power directly from the pins, that's also what software reads out.
There are pins for Processor IA Cores and Ring power rail alone, that is only powering these things so you can only measure that there and nothing else.

Processor Power Rails - 002 - ID:743844 | 13th Generation Intel® Core™ Processors