The Intel Problem: CPU Efficiency & Power Consumption

[@GamersNexus]

Thanks, Greg for this idea! We're funding this one ourselves, so grab a brand new GN15 all-over print component mouse mat! store.gamersnexus.net/products/15-yr-mouse-mat or a GN15 metal emblem glass to support the testing! store.gamersnexus.net/products/gn-3d-emblem-glasses Watch the CPU Sample Size video! kzbin.info/www/bejne/hobIi4RppdCsY9k Learn about Intel's Snake Oil: kzbin.info/www/bejne/roa3ZZdrfruMZqM You can also find Wendell from Level1 Techs here: kzbin.info And find our mega charts for CPU power here -- new charts not added yet: gamersnexus.net/megacharts/cpu-power

[@TheRedneckCodecChannel]

Grabbed me one of those 15 yr annimousary pads ol son!😂

[@RNG-999]

Thanks, Greg!

[@eduu1717]

@GamersNexus can I have one of those chips :( uwu

[@weltsiebenhundert]

When did Intel lose in effizient vs AMD? 1st or 3rd Ryzen? (Dont think Bulldoser)

[@Post1157]

​@weltsiebenhundert378 I think during the 3000 series, thought the 2000s might be close

[@dremy746]

I've done extensive testing and I've found that if you power limit both the 14900K and 7800X3D at 0W they perform identically with identical performance per watt.

[@cameronbosch1213]

Of course they would! 😂

[@FreshChriss]

Dont say it to Greg, he would still bet that even at 0w the intel cpus where more efficient and faster

[@GamersNexus]

Wow! This is a huge break-through! We'll immediately test the power consumption of the systems when they're off. hahaha

[@RealAvus]

You are wrong.. Intel CPU will actually give out energy from ether while AMD CPU is just turn off at 0W

[@smncrowe]

@@GamersNexus Actually, that's not as mad as it seems, who actually switches their systems off at the wall .......

[@Noobsaucer]

The low-key tragic detail of this video is that the 14900k is LESS efficient than the 13900k in literally every single test. Ouch.

[@RFC3514]

That was to be expected. They're basically milking the same CPU for a few extra cycles, at the cost of even higher power.

[@claudiobizama5603]

That just proves that BIG NUMBER BETTER

[@miqotelover]

Underrated comment.

[@PyromancerRift]

It is normal, it is a refresh. Same trash CPU overclocked to look better on benchmark. They literally do on desktop what they say AMD is doing on laptop. But intel is doing it on high end parts.

[@tommihommi1]

because it's literally the same chip, just overclocked a little bit more than it already is

[@UncleSilverGaming]

Damn Gregs out here makin' us look bad! From one Greg to another, DAMN IT, GREG!

[@Shiny_Dragonite]

I guess Dave (if you're familiar with the radio commercials) has handed the title off to Greg.

[@Abrasive-Heat]

It’s okay. Let us be the best of Greg’s and hopefully we’ll inspire the others to do better.

[@Eveningbreeze78]

Gregs does ask the stupid questions nobody asks, but when asked want too know the anwser to... :) Results confirm bias, though bias had no quantification.. this is nice

[@OnTheRocks71]

Yeah that other Greg really Greg'd it up for everyone. sigh

[@slaysi]

My name Greg

[@andydbedford]

Greg: you are not testing CPU’s scientifically enough! GN: challenge accepted. Subscribers: Thanks Steve 👍

[@ADesertVoice]

@@amrmoustafa328 er they are comparing based on claims by either company on what their CURRENT chips performs. It doesn't matter what "technology" those chips are build on. It is FAIRLY COMPARED based on the claims the producers of those chips are saying about those chips NOW. Why is it that someone like you can't seem to understand that difference? Are you being shady here and claiming something stupid by saying that those companies like AMD and INTEL are LYING about their chips in the first place? Your assumption is what is truly not fair as the baseline is the comparing based on claims MADE BY THE MANUFACTURERS of their own chips in their current state, no matter what technology is in them.

[@Zesmas]

not really fair considering Gamer nexus has a billion sponsors and has the money to just throw at research while the normal person cant.

[@ihebammar6568]

back to you steve

[@shadow7037932]

The 5800X3D will need to be in a CPU Hall of Fame. What a champ of a CPU. I feel like this CPU alone will keep AM4 around for quite a few more years.

[@newtwitchclipper]

Yea i went to am5 and got the 7800x3d but my brothers build i stayed am4 and for the price to performance the 5800x3d and a b550 velocita being 175$ cheaper than my combo... Lets say i have a 7900xt in my am5 build.. for the same price u can stay am4 with the 5800x3d and get a 7900xtx and have better performance for the same price as the am5 system

[@thepcenthusiastchannel2300]

I bought one for a Dual Vega 64 Crossfire build as I was trying to get around the poor GPU scaling due to CPU bottleneck. Just for a fun project. It worked. Both Vega 64 GPUs get 100% utilized and you get a doubling of performance. Very strong CPU.

[@IRefuseToUseThisStupidFeature]

If they continued to make it.

[@jawohlbxb3534]

Currently on ryzen 2700 when im gonna upgrade i dont know but when i do it will be 5800X3D it will be cpu that a lot of people with old ryzens will upgrade to just because performance of it for just upgrading your cpu is insane

[@Drew1218s]

For gaming yes but for anything non gaming not really.

[@eliasroflchopper3006]

It's hilarious to me that the 7800X3D without FPS limit still pulls less power than the 14900K with an FPS limit Edit: the amount of fanboys defending Intel with all they have, for no good reason at all, is just ridiculous. Stop saying they should have undervolted it. Stop saying they should have turned off the E-cores. Stop saying they should have limited the power draw (which they did for the 14700K photoshop test, actually). Stop trying to make this literal waste of energy better than it is. You get nothing from it by doing this. You only expose yourself to siding with a company that rips you off, and that makes you look like morons. And besides, this video was supposed to be a fair comparison between Intel and AMD. If you tweak Intels chip to the point of being advantaged, you surely can do the same for AMD.

[@ReclusiveXL]

Looking at power consumption of the 7800X3D, I wonder why AMD didn't release it on laptops platforms yet. It would have no match at 50-75W TDP in games. Edit: yes, I know about the R9 7945HX3D, but it still has to share some of its power between two CCDs, instead of focusing all the power on the 3D V-cache one(and laptops usually have 55-75W limits). There's 13980HX on laptops, and it is just identical(in terms of its cores count, frequency and even cache) to power limited 13900K. But they wouldn't need to limit the 7800X3D that much, even 55W would be enough.

[@WWAVES1]

@@ReclusiveXLIt might be a temperature constraint; that and the higher power floor to run them

[@dagameboy]

@@WWAVES1 nah, in laptops the cpu doesnt have a thicc boi IHS and 50-75 watts is relatively normal for laptop CPU's

[@_under2world_]

Thx to the 3D cache performance are higher AND they have to reduce the maximum voltage to avoid damaging this extra cache. Those 2 factor make it incredibly efficient/performance, that and the fact that zen core are more efficient that Intel core by default

[@Neketre]

@@ReclusiveXLit needs desktop levels of cooling to remain stable at 50-75W. There’s a reason it doesn’t boost as high as its non-X3D counterpart, which has to do with voltage and temperature limitations.

[@limemason]

I love how Steve will see a tweet with like 5 likes and take it as a personal challenge necessitating tens of man hours of work. What a mad lad

[@GamersNexus]

This one was about a month of work! And it required 3 of us! And give me some more credit -- it had 2 likes! hahaha

[@EversonBernardes]

"Someone is wrong on the internet!"

[@waynesworldofsci-tech]

@@GamersNexus It’s the sort of work that keeps us coming back. Great stuff.

[@fratista]

He wants the paycheck

[@limemason]

@@GamersNexus Astounding work gentlemen

[@shivcharan5700]

Even though Greg took a fat L, he gave us all a big W for introducing this new set of power efficiency tests! Thanks Greg and Thanks Steve!

[@moritzbecker131]

Thanks Steve

[@Lurch-Bot]

I don't see it that way. It is a win-win. Scientists learn from their mistakes, empirically.

[@imjustapotatoleavemealone]

Now i see Greg as my personal hero!

[@Braiam]

Back to you Steve.

[@FearOfDeath1332]

Thank you for saying it

[@SpiritOfTheHeretic]

I don't think I've ever seen someone get blown out of the water as hard as Greg just was

[@twign6378]

I love it when Steve pulls out these "I didn't need to go this hard but..." videos

[@redacted5744]

Greg owes somebody his paycheck, lol.

[@rmvdhaak]

Can you believe the guy is doubling down on Twitter ?

[@TheHectic0ne]

@@rmvdhaak It's impossible to argue with stupid, even with facts.

[@mangatom192]

​@@rmvdhaakWhat's the @?

[@theavera9ejoe]

The fact you get biased comments like this and still take the time to go through and methodically test them speaks loads to you and your team's character. THIS is the content I'm subscribed for! Love to see it.

[@betag24cn]

i like that you used the word biased, instead of call greg a fanboy, what he is, lol

[@anorax001]

I don't think it matters how many benchmarks or tests Steve runs on these CPU's. Intel fanboys will not accept any result even if it proves them wrong.

[@shaneeslick]

@@anorax001 it doesn't matter what company the FanBois are all as STOOPID as eachother, they are the Flat Earthers of the Tech Community.

[@CptJistuce]

​@@anorax001Look, if AMD was really better, how come they can't afford a five-letter name, HUH? That's right, you can't answer that, because there is no answer!

[@chrisbullock6477]

That's the difference between maturity and also having the intelligence to create out of hate and to make lemonade out of lemons something that only comes with maturity something that many of the Millennials and the under 30 crowd seriously lack and their parents are to blame

[@SuspiciousGirafe]

"We figured that some people might not like this scenario... because it's completely valid, fair, realistic and ootb as you use the processor" LOL, you didn't have to roast them this hard, Steve !

[@MattFowlerBTR]

Steve didn't do the roasting - he just put Greg#2 next to the 14900k heatsink (ok, let's be real, radiator) and let all the stupid amounts of heat do the work.

[@craigdeeter8042]

Yes he did🤪

[@Burmn]

“Thanks Steve!”

[@philslaya]

Greg must be a new userbenchmark for hire 😂

[@IRefuseToUseThisStupidFeature]

LOL, for that matter... They should call this test _The Greg._

[@walterwhite415]

I thought the same. Sadly there are still a whole lot of people who fall for the bullshit Bullshitbenchmark is telling them.

[@j5ylim396]

with how ridiculous these types are getting im starting to think theyre just stockholders with all their life savings on the line

[@Therapistwatermirrur]

@@j5ylim396 I'd honestly buy Unity stocks over ITNC right now. Even with the Unity drama.

[@CptJistuce]

​@@j5ylim396I wish I could believe that. I've seen enough fanboys over the years that this just seems like normal denial of reality to me.

[@dontreadtoofarintoit1868]

I can only imagine its difficult to come up with creative and unique shots for a processor, it doesn't go unnoticed from myself. Some creativity here using the lens with it being the subject of investigation.

[@GamersNexus]

The video team will appreciate that you noticed! I'll share the comment!

[@mnemonik61]

Agreed! The video team has really been upping their game over the past couple of months. Kudos to the GN video and editing crew!

[@lexwaldez]

Can we all just take a moment to thank Greg for threatening us with a good time? This was an EXCELLENT presentation and more than a little eye opening. I knew 14rth gen i7 and i9 CPU's were power hungry but Intel got absolutely abused. I should not have started this at bed time. 30 minutes gone in no time.

[@razoo911]

even 8, 9, 10, 11, 12, 13 gen too were/are power hungry...and i dont know why intel call those cpu 14gen, in fact they are 13gen with a bit overclock

[@vadicus]

Describing CPUs in terms of their fps per watt is brilliant.

[@JFinns]

Intel has left the chat.

[@n3rdy11]

I wish the same would also be done with GPUs, I have the suspicion that AMD will also be ahead there when the workload is pure rasterization without any DLSS/RT.

[@ivofixzone6410]

Well, it's the same as describing EVs of their mile per Kw , because there is the truth of inefficient and expensive usage.

[@sitaroartworks]

@@n3rdy11 the real performance saver is FSR 2. It is really good.

[@Link0402]

So as someone with a 7800X3D living in Germany, where electricity is about 32 cents / kWh... WOW I'M SO BUSY SAVING MONEY HERE.

[@slipknot1943]

You you realize that even if you would use 400w 24/7 for a month @ 40c price, it would only cost you like 100€ E: You do it scales down? Eg. 150w gaming 2-4h per day @ 40c. Month cost: ~3-6€

[@RotaryJunkie]

​@@slipknot1943 I don't think you made the point you meant to make there.

[@dascandy]

Positive side: Saving power. Negative side: It's that expensive because it's all coal, instead of GN's nuclear...

[@IdiotRace]

@@slipknot1943 I think they would rather have the extra 100 euro

[@CarbonPanther]

On the other hand i'm probably wasting money, i have a 5800X3D and it takes about ~60W on Idle which is a lot especially when i watch videos which is when the system spikes to 110W watching youtube, my old i7-6700 system does about ~33W... while watching videos, i know it didn't have a GPU in it, but that's still a heck of a difference especially at idle where the 6700 would draw like 15W at most, that's a huge difference and it's really stupid how little AMD's CPUs save power at idle because of their IO die or whatever, or that my GPU needs apparently at least 40W just to watch a damn video at 1080p.

[@jan__itor]

Idle power consumption for CPU and GPU reviews would be extremely cool to see. I don't think anyone is doing this kind of tests, but it would be nice to know that

[@GamersNexus]

They're tough to get right. Working on that separately!

[@mcod99]

my 7950x3d pulls around 30ish watts at idle and my 4090 pulls 15w.

[@Zapdos0145]

AMD would be hard because the 7000 series still has high idle power under certain circumstances like different monitor resolutions and refresh rates amongst other things

[@tommihommi1]

one problem with idle power testing is the *crazy* swings based on motherboard model as well as software configuration.

[@Zapdos0145]

@@tommihommi1 second this. it’s very hard to have a “control” when there is no “control” everyone’s software is different, everyone’s hard ware is different, even people’s monitors are different, it’s hard to control all that when no one does that 100% of the time

[@davidmstuff]

Anyone else enjoying the ridiculously good quality of stuff that the GN team has been putting out lately? Mfers are heading into 2024 on FIRE!

[@hukl3945]

More efficient CPU also means lower requirements for the CPU Fan, Case Coolers as well as the Powersupply or even the case itself. So there is more merit than just power cost.

[@nickierv13]

Not just lower requirements, but more options as well. CPU trying to pull 300W is going to need a lot different cooler than one chilling around 150W, but run the 300W cooler on a 150W CPU and you can run the fans a lot slower and thus quieter. More efficient isn't going to be less total power, take a 100W CPU that can do X work/hour but a 200W CPU that can do 4X work/hour. Yes total power saving but higher peak power.

[@nickierv13]

@@fluffy_unicorn859 Not sure on exact numbers, but when dealing with heat a 4:1 ratio of heat to power needed to deal with the heat is probably on the good side.

[@Deathscythe91]

also motherboard , dont have to worry about vrm's as much

[@saricubra2867]

My i7-12700K is the only one that can be air cooled. Any i7 above it needs an AIO.

[@griffin1366]

@@saricubra2867 What nonsense. I have a Noctua NH-U14S and it cools my 13700k just fine at 5.4Ghz. Granted I set the power limit to 200W and it sits at 90C in Cinebench. In gaming and everything else it sits at a stable 600 RPM.

[@ExpiredStyrene]

Power efficiency is a critical metric for me these days; thanks for this in-depth testing!

[@thomasschraubt7497]

Why is that?

[@Zandonus]

Yeah, one of my heaters broke, and this 65w gen1 ryzen is not doing it even with a malfunctioning fan. Came here to get some blood to my head and consider buying intel for the extra heat generation.

[@dragomegaman3711]

​@@thomasschraubt7497electricity bill for me 😢

[@annekedebruyn7797]

@@thomasschraubt7497 In some countries, energy bills have been doubled and in some cases tripled since covid.

[@BugThorn]

@@annekedebruyn7797we pay 0,5$ per KWh, when your average net income is around 1750 and half that is price of your rent ...

[@jedimastapeb]

The coverage you all have been doing, on virtually everything, has been outstanding. I know it's not much, but here's a token of appreciation. I debated whether or not I should give you this, considering you got someone's whole paycheck as a result of this video. I'm obviously in the wrong line of work.

[@Brockzillagaming]

I don't think they're actually getting anything from Greg lol. He's all talk. But if you want to support them, it's better to buy from their store, they get a similar cut to what KZbin gives them and you get a thing!

[@GamersNexus]

$20 is huge! Thank you so much. We really appreciate that. That's a large donation for us and they all go a long way!

[@jedimastapeb]

⁠@@Brockzillagaming Yeah, I know they didn’t actually get a paycheck, it was a joke. :)

[@elforeign]

Steve, wow, the GN team delivers some of the best most objective content on KZbin full stop. And while the comment leading to this video was wild, the testing and analysis of power consumption and efficiency in real world application is incredible and eye opening - thank you for this and hope to see this work in future videos for future cpus and gpus

[@silverarm6130]

Nice work on this one. Power consumption and power efficiency are two of the most important yet overlooked aspects of this technology.

[@pino_de_vogel]

sadly optimal power target/limit was never shown as they capped fps for gaming not power which is hot the correct way to go about it.

[@user-lp5wb2rb3v]

intel bulldozer moment lol

[@MultiOlimac]

some intel fanboys only state that the 14900k smokes amd just because of raw performance while drawing three times or more wattage. thats what they never tell users, and also that it gets hotter while doing so.

[@esotericjahanism5251]

@@MultiOlimac Well TBF some people don't really care about it if they have the money for a proper cooling solution. Most people just aren't going to figure power efficiency into a Gaming build. A server or workstation that's going to be running 24/7 and there's going to be dozens of them? yeah people are going to care then. But most gamers are just going to buy the CPU that gets them the best frames that's within their budget or other build constraints.

[@user-to7ds6sc3p]

@@pino_de_vogel no. Why would you limit power for gaming? You want a smooth experience first and power efficiency second. Limiting power low enough to potentially destroy or have fluctuating framerate kinda defeats the whole point. No sane person buys a top end CPU to look themself out of 60+FPS gaming. Limiting power for non gaming makes way more sense, because you don't need a specific speed/framerate to get a enjoyable experience.

[@ijustsawthat]

Greg: places bet Intel: Be careful of what you wish for

[@rabidwallaby84]

Greg is Homer. Intel is Santa's Little Helper

[@arnlr61233]

The 7800x3D is absolutely dominating 🤯 So happy to have chose it for my build

[@MrAnimescrazy]

Yup me too what are your specs?

[@TroyEdinSammyWinchester82]

Same here! It's a beast

[@gehrrman]

what is your spec

[@MrAnimescrazy]

@@gehrrman i have an all white high end build with the gigabyte aero oc 4090/ 7800x3d/ 64 gigs of ddr5 ram. The pc is a beast and I will clean it soon. I clean my pc every 2 weeks to 1 month at the latest.

[@DraxTrac]

I went with an all AMD build, 7800X3D and a 7900XT and I'm so happy with the performance.

[@rellikai945]

Man Greg really on a tear of headassery lately. Someone needs to stop this absolute menace to society!

[@HollywoodEDC]

what ever encoder you guys at GN are using to make videos is BEAST. Your youtube videos look orders of magnitude better than 99% of other youtube channels. It's insane how good these videos look on my G9 monitor and 4090 lol.

[@beachslap7359]

Also add short support of their sockets to the problem.

[@GamersNexus]

That's a different problem!

[@Troonielicious]

Constant crashing, mobo problems, can’t oc ram on 4 sticks etc. etc. on my 7800x3d with NO issues zipping on coffee ☕️

[@sean8102]

That was the main reason I went with a 7800X3D over intel when upgrading from my 8700K this month. After I read that their next gen of CPU's will need a new socket. Not buying a new mobo that's already a upgrade dead end. AMD claims they will support AM5 for "years to come" and with the longevity of AM4 I believe them.

[@thebcwonder4850]

@@sean8102 AMD only gave 5000 support on AM4 because Intel would’ve been more popular if they didn’t

[@priyanshusharma1812]

@@thebcwonder4850 and your point?

[@Yuki2204]

There is a certain level of professional spite here, And I absolutely appreciate it.

[@artofnoise5013]

That's the perfect description for it!

[@MrZestubo]

I'm excited to hear idle power draw is coming, I feel like as live service models get worse there will be more interest in running local servers. I know I am. I don't pay a ton for electricity, but I'm still interested in running as efficiently as possible where I can.

[@JayBe443]

You guys do outstanding testing for consumer products, with data and knowledge of how it applies. All while making it easy to understand. You guys rock. Considering my basic, general understanding of computers, watching a shorts series or videos showcasing your testing methods and what they aim to achieve is needed. Alot of things go over my head or lead me into rabbit holes. Maybe thats why google recommend the channel lmao

[@Lurch-Bot]

I think he's on to something here. There is far too much BS flying around in the industry and we need an independent, nonprofit organization to test PC components for safety, efficiency and performance. UL is a joke with their 3D Farce benchmark. If anyone could get intel off their butts and designing a modern architecture, I bet Steve can. The problem is they've been huffing their own farts for too long over there at intel and they've gotten complacent. Now the entire world can see what BS artists they are. This troll post that prompted all this new testing was undoubtedly the direct result of intel propaganda. Why put the effort into R&D when troll bots are cheaper and more effective?

[@watercannonscollaboration2281]

I’m pretty sure when Amit had his 7 leftover X3D CCDs he never could’ve imagined how much of a game changer (pun unintended) a server part would become in performance and efficiency

[@friendlyhobo6483]

The money you waste on electricity can be deducted when you realize the intel cpu’s can be used as a duel purpose heater for your home.

[@DimkaTsv]

On other hand your power bill will double on air conditioner during summer. P.S. There is centralized heating during winters where i live, so even there Intel CPU will make things worse by overheating room up to point when temperature forces me to open window to cool it down

[@R4d1o4ct1v3_]

This just gave me a flashback to ancient times, when we used to make that joke about AMD CPUs xD The pre-Ryzen era really was dark times all around.

[@greebj]

We just need to run it at negative GHz so it sucks heat out of your room in summer Hurry up and figure it out already, Science

[@walterwhite415]

@@greebj Just go to the bios and set a negative offset of 10ghz on your CPU and boom, your CPU is running backwards now! ggez

[@mikfhan]

@@R4d1o4ct1v3_ dark days indeed, but now we are back to the Athlon days :D competition rocks!

[@BMKTW]

Steve, don't forget that higher power consumption means spending more money for a bigger PSU and Heatsink.

[@kenshirogenjuro873]

Not to mention motherboards that flow the amount of required power

[@rhesaramadhan8474]

not to mention more case fans to make it cool..

[@Lotiks.]

It's so nice to see how Tech Reviewers always come up with new ways to approach things and you really cut it again with this one. Always thought that Power is an underated Aspect in CPU / GPU reviews.

[@DarioCastellarin]

A minute of silence for Greg's paycheck.

[@malinpl2]

This is the best CPU test ever. This should be presented to all people who want to build their new PC

[@snozzmcberry2366]

Never in the history of fanboyism has anyone been so thoroughly, exhaustively, yet 10/10 class act also educationally shut. the. fuck. down. The fact that the fanboy accidentally, by sheer luck, induced GN to explore an entirely new way of measuring and evaluating data about CPUs is just.. I love this. I am more filled with joy over the good that was made than I am sadistic joy over corporate worship idiocy being given the heaviest of boots. There's a lot of both, but new data = awesome still wins out. Love you, Steve & team. ...but, like, imagine provoking an incredibly in-depth effective advertisement to the complete contrary of what you argued. I reeeally doubt any white collar folks over at Intel would be pleased with this turn of events. It's beautiful. The whole thing, just beautiful.

[@futuza]

Intel: "Shutup Greg, shutup! We don't need them taking a closer look-damnit Greg!"

[@gosuprime]

People should make their PC decisions based on performance, not brand loyalty. The current best setups involve an AMD CPU, and Nvidia GPU. I'm not team red or team green, i'm team "whoever has the best/most efficient parts." I've liked Intel CPUs in the past, and wasn't sure if it was worth a platform swap to an AMD chip, but the 7800x3D and 5800x3D have been so impressive and super fast that making an AM5 system was a very easy decision. "But if you like Nvidia you can't support AMD!" Nonsense. I'm not loyal to shareholders, i'm loyal to whoever is making the best components...cause it's my system! I hope the AMD 3D CPUs make Intel improve and maybe down the line they will have a solid competitor, not just Raptor Lake refresh+++.

[@olo398]

preach.

[@JS-be8ll]

​@gosuprime nah, AMD CPU and GPU is best bang for the buck. 6700xt, 7600, or 7800xt, I'd take over any comparable Nvidia gpu right now at current price points. The supers might change it up though.

[@10th_Doctor]

@@gosuprime Being that my major hobby is CGI rendering using software I have used for the past 11 years so I know it extremely well, I have to be team green. Iray only runs on Nvidia GPUs else it drops to glacially slow using the CPU. However CPUs are a different thing entirely. I am doing a full new build next year and know zero about AMD as I have been using intel for so long. I have no clue what CPU to get that would match a 13700 or 14700 nor do I know anything about AMD compatible motherboards.

[@TECHiSuppose]

Great data! Totally not surprised a smaller manufacturing process pulled ahead in efficiency, lol. The kWh charts were especially great. I'd be curious about idle or near idle power consumption between systems. Like whether either brand has better technology to keep computers efficient when they are not doing intense work. Such as browsing and maybe watching videos. Maybe the differences are insignificant but that could be charted out over their expected lifetime make it significant (say 5 years). EDIT, I saw idle was mentioned toward the end, nice!

[@washaa]

Sometimes I can sit down with my desktop for half a day only browsing or watching videos. Sometimes I use the PC daily but don't game for a week. The power consumption of having a few base programs up is just as important as the heavy load consumption since I only game for a few hours each week. Some kind of baseline would be nice to see, however it is defined.

[@Targonis1]

It's not all about the fab process, but is more about the focus. Intel is taking the attitude, "just throw more power at the chip to let it clock higher", and the result is that efficiency goes right out the window. Alder Lake to Raptor Lake to Raptor Lake Refresh, yea, add some E-cores going from 12th gen to 13th gen, add cache, but then boost the power draw for higher clock speeds. Intel has been using this approach since the 7th generation when Intel had locked its upgrade path to the fab process. Oh, Skylake...yea, 10nm wasn't ready yet, so Skylake with more power and higher clock speeds was what we saw from 6th to 7th gen. Then 8th gen, still Skylake, but with 6 cores at the top end. 9th gen, 8 cores at the top end, but still Skylake. 10th gen, yep, for the desktop, Intel still couldn't improve the design, so again, Skylake drawing more power for higher clock speeds. 11th gen, still no 10nm on the desktop, all a joke. That was the big, "same design, throw more power at it for higher clock speeds" approach. Intel is still doing it that way, being so obsessed with higher clock speeds, they don't bother with thinking about efficiency at all.

[@TECHiSuppose]

@@Targonis1Not all about, sure. I'll agree there. Though this video clearly shows differences when the CPUs are given the same power limits. Design can certainly be a factor as well. All that said I'd consider manufacturing process the key attribute.

[@MrAltairantares]

Really do love GN's content these days. Really feel like the channel is pushing hard when a lot of others are struggling.

[@MartinStrazynski]

"and is entirely funded by people like you, our viewers. And, people like Greg..." Instantly paused the clip to make sure I had hit thumbs up after that one. Hilarious (and educational) content, as always. Appreciate you guys. Happy New Year!

[@shaneeslick]

That special ad was hilarious 🤣🤣🤣🤣🤣 I almost choked on my vegemite toast

[@gregoryburzy468]

As a Greg, I appreciate the shout outs. I have plenty of ideas that I like to test on my own because I have no life, however I'm not bitter like these other Gregs. Thanks Steve, you're living my dream job and I appreciate your work and being the gold standard in journalistic tech enthusiasm. Merry Christmas!

[@Mamiya645]

Can't wait to see Intel's army of lawyers try to help them outta this one. Maybe there's another benchmark site they can buy.

[@gg-gn3re]

Greg will have to be giving them a lot of paychecks for the fee

[@JJFX-]

Well it's not like there's no reason someone would choose an Intel system but for the average gamer it's certainly compelling now that DDR5 and budget AM5 boards are much more reasonable. Honestly I'm just stoked there's healthy competition and enjoy both platforms for different reasons. AM5 still has its frustrations but has come a long way since launch. The next few years are going to be a lot of fun and I'm much more excited about all of this than the current state of the GPU market.

[@toufusoup]

Thank you Gregs for continuously blasting comments like this so that Steve can make these videos out of spite

[@JellyLancelot]

Greg out there looking like the CEO of User Benchmark

[@samoh4u]

Best part about the 7800X3D efficiency is I can setup all my case and cpu fan curve to only use 1000 rpm for the fans when it's gaming and still have awesome temps to noise ratio. And still maintain high boost clocks. Seriously such a good gaming CPU.

[@Pikminiman]

These results are striking. Thanks very much for diving this deeply into power efficiency in more use-cases.

[@wouldntyaliktono]

Love this scientific approach to all the comparisons! In situations where you've got two variables to illustrate at the same time, you'll get way more mileage out of your graphic if you use a scatterplot. For example, instead of a bar chart with MIPS per watt, do a scatterplot with the MIPS and watts on the X and Y axes. That way you're not losing any information about the overall power consumption, but you're still illustrating the relationship between power and compute. And I'm betting that the point cloud will beautifully illustrate the diminishing returns of progressively more expensive devices.

[@GamersNexus]

Will play around with that and see how it looks for the next one! Thanks!

[@wouldntyaliktono]

@@GamersNexus Excellent, looking forward to it! I just manually plotted FPS and Watts for your Cyberpunk numbers using this technique and the difference between Intel and AMD _reallllly_ stands out. Thanks for sharing this with the world!

[@prgnify]

@@GamersNexus in that vain, for the GPU busy graph you could get the area beneath each graph and compare the deltas from one CPU to another. I think that would be the best visualisation for the data.

[@prgnify]

because then you can plot both bottleneck AND perf/W at the same time. like a bottleneck per Watt sorta

[@wouldntyaliktono]

@@prgnify I'm gonna be real happy if we start getting integral calculus or GLM out of this channel. I might have to buy one or two of those mod mats if that starts happening.

[@petergimeno6547]

Excellent review and interesting perspective. One thing to note about cpu power consumption: power correlates to heat, which can be a meaningful factor when considering how ambient room temperatures and noise levels are changed over time, especially in hot climates or during the summer months.

[@sean8102]

Just upgraded from a 8700K to a Ryzen 7 7800X3D this month. Super happy with it, and what a upgrade. Not just the CPU but PCI-E gen 3 to PCI-E gen 5 and 4, DDR5 etc. Also grabbed a Samsung 980 Pro 2TB. So it was quite a difference in speed/snappiness after doing the upgrade. Also my decision between intel or AMD was made MUCH easier when I read that the next gen of intel CPU's will require a new socket. While AM5 will be supported for at least ~2-3 years and with AMD's track record and longevity of AM4 I believe em. Didn't wanna buy a new mobo that's a upgrade dead end.

[@luvingyouu]

make sure to adjust pbo to -20 or -30mv in PBO (or auto PBO settings some boards offer) drop SOC to 1.2v (unless you plan on overclocking your memory, but 6000mhzcl30 A/M hynix die should be stable at 1.2v on nearly all 7800x3d cpus) and get the same performance with lower temps and lower power draw :))

[@GamersNexus]

The 8700K was a great CPU. Jumping to the 7800X3D is pretty huge - you basically went from one top-of-the-line build to another. Enjoy the system!

[@GigAnonymous]

Exactly the same upgrade here! I wanted to go for Intel and I waited for the 14th gen... but genuinely even the cost of buying an AM5 MB *and* new RAM couldn't offset how disappointing the new Intel CPUs were. Same thing happened a bit earlier when I replaced my GPU - really wanted to stay with Nvidia, ended up with a 7900 XTX. AMD got my money because its competitors keep shooting themselves in the foot... (also I bought an ASROCK MB, because it was the *best option*. Whatever happened to the motherboard market?!?)

[@sean8102]

@@GamersNexusThanks! Happy holidays and happy new year to the GN team!

[@sean8102]

@@GigAnonymousI was going to order a ASROCK x670e mobo as well. But it went out of stock the day I went to order it. Ended up getting the MSI MAG X670E TOMAHAWK WIFI.

[@MrLense]

It's crazy 10 years ago AMD was in the exact same spot

[@Qalibrated]

Even about 7 years ago, pre-Ryzen. The fake core era. Intel was so far ahead, kinda disappointing to see where they're at now.

[@haunt97]

@@Qalibrated thing could have been different if they did not sit on their butt, milking their customer money for few extra mhz for years. It's quite funny

[@ShanesGettingHandy]

Came here to write the same comment, and I also wonder.... is Intel doing the exact same thing AMD did back then? Maybe they've got something up their sleeves and in the meantime they're pushing the old stuff harder and harder to bide time.

[@CaptainKenway]

@@ShanesGettingHandy Intel's big ace in the hole is... copying AMD. They're moving to chiplets (or "tiles" as they're calling them) and working on their own stacked cache solution. It's all quite amusing really, given Intel's previous malicious marketing about "glued together" CPUs.

[@herbertwalter8693]

​@@Qalibratedwelcome to the trap that is complacency

[@Gnomleif]

Greg: I bet a paycheck that the 14700k will be more efficient in every workload under these conditions. Steve: Challenge accepted! _Boss music soundtrack starts playing_

[@RobAryeeArc]

Thanks Steve.

[@kurtjustiniani1354]

GN: Be cashing out your paycheck soon Greg 😈

[@yarghhargh9345]

reddit moment

[@Spark010]

Love the frames per Watt metrics. Think you should select 2 games as baselines and the current top selling game so that the data is meaningful at the time of benchmarking for gamers who are building a system for the current top seller.

[@tulsatrash]

This is some damn fine journalism. Glad to see it still exists.

[@davidbondy2250]

I built my first PC today, and thanks to your videos (as well as videos from other channels like Hardware Unboxed) I got a 7800X3D. I'm still sussing out some issues with expo, but even with expo off the performance gains I've seen versus my previous PC (which used an R5 3600) are absolutely ludicrous. Even with the expo issues, I'm super happy with my purchasing decision, and your coverage made the decision incredibly easy.

[@tyson8585]

I guess your stock ram kit speed is too high? like above 6400? I read AMD recommends 6000 as the sweet spot, which also mirrored my personal findings. What's funny is an i9 needs to run 7000-8000 ram just to get close to some 7800x3d gaming benches.

[@davidbondy2250]

@@tyson8585 maybe. Expo did say it was targeting 6,000 in bios. I imagine once I'm more comfortable with going into bios, I'll manually increase the speed until I find the maximum I can get away with. My RAM speed is currently at 4800 and performing incredibly well, like BG3 at max settings and 60 FPS no matter what well. But man that additional 1,200 MT/s is something I will want to try to claw back as much as possible.

[@tyson8585]

@@davidbondy2250 Looks like you're in luck, I came across an x3d ram speed benchmark by Hardware Unboxed and looks like they don't really matter, only a couple of % difference between 4800 and 6000 at 140+ fps. I think you won't even have to do any further tweaks because of the huge amount of cache these cpus have.

[@Navi_xoo]

Ryzen have a different style of memory controller that has problems above 6000 mt/s generally. Some people claim even at 6000mhz. Good news it barely affects fps and beats intel the same at 5600 cl28 or cl30.

[@davidbondy2250]

@@tyson8585 I think I know which video you're referencing. I'll go back and rewatch it. Thanks for reminding me, and also thanks for doing some searches. Things are still going super fucking well on my PC, it is absolutely handling BG3. It feels good to finally play that game and have it be stable.

[@MarkNiemann]

This has my purchase of the 7800X3D feel even better when I chose it for the gaming performance and efficiency. I was tempted to go to Intel but I’m glad that I didn’t.

[@frankytanky5076]

Same 7800x3d is truly revolutionary. The first time I gamed on it I legitimately thought it was broken due to the lack of power consumption and heat. Nope it’s just that good.

[@KiSs0fd3aTh]

This is Greg, or should I say Greg#2. Your original post which I replied to was "the 14700k is way less efficient". I said that's not the case unless you are talking about gaming. In order to disprove my point, you went ahead and spent the first 17 minutes of the video testing games. Uhm...okay? Then you finally tested exactly what I claimed, a non gaming workload at same power. Not only was the 14700k in fact faster AND more efficient than the 3d, it was the most efficient CPU on the whole freaking chart - threadripper not included. Let's try to wrap our heads around that, one of the most inefficient Intel cpus in existence (supposedly), when limited to the same power as eg. the 7800x 3d, tops the whole freaking chart. Then you also tested photoshop at same wattage, and the 3d edged ahead. So from a 28+ minute video, you run just 2 tests that had anything to do with my claim, and both of them disproved your original point regardless (unless your original point was strictly talking about gaming, in which case you are correct, but I never disputed that in the first place). Now if you want to pedantic about it and say "Yeah Greg, but you said EVERY test" then sure, of course it's not in every test, and I knew that before I made the post anyways (eg. ycruncher using AVX) . But that's just being pedantic, run 100 workloads at same wattage and the 14700k will win in 95 of them. Im pretty sure you know this so I don't see the point of the video in the first place. If you want to actually make a video about it then run the usual Cinebenches, corona, blenders, 7zips etc at same wattage and shows us.

[@benfoster93]

Watts was actually my 2nd most important metric after max budget when it came to choosing my latest pc parts. Since I'd only be gaming, I knew I wanted to max framerates and power efficient. With the 5800x3d being so efficient, undervolting both cpu and gpu(the 4070) using frame limiters and upscaling were factored in my choices. Really happy in the end I can max gaming everything at 1440p minimum 60fps

[@YDINO69]

I don’t understand how high end CPUs are comparable between AMD and Intel when Intel uses double the power and runs way hotter. Why would you do that to yourself? Just to run 128gb ram?

[@EXiLExJD]

Double the power and runs hotter despite "efficiency cores". Intel CPUs are a joke.

[@theviniso]

@@EXiLExJD Isn't that because they're stuck using larger nodes than AMD?

[@akilh340]

@@thevinisoI'm not sure about “stuck” but yeah

[@JohnWiku]

​@@theviniso The fab fabrication tech is part of the problem. Also Intel is running higher clocks, which by itself brings up power consumption. Remember what happened to both Intel and AMD when they went full pentium 4 (Bulldozer for AMD)? Intel at this point is going full pentium 4 again, while trying to cram as many cores as they can, that is why you got P and E cores, while running chips fabbed a lesser node...

[@tyronewilson6463]

@@akilh340 I'm pretty sure stuck is a good term. They don't have the long term deals with TSMC that AMD and NVidia do because they aren't a fabless company and had financial and executive pressure to do it internally since they spend so much money on their own fabs.

[@kingzilant]

I love the new snarkier Steve lol the energy is just entertaining and informative... keep it up GN.

[@hovant6666]

The IO die idle power requirements answer a question I had asked in an earlier vid's comments about why AMD's idle power consumption is 36-68W compared to 9-28W on the 11900KF; thank you! What gets measured gets done, and since we raked AMD's Vega Frontier Edition over the coals for its power consumption and the Nvidia 30/40 series for that same reason, then it only makes sense to transfer that analysis and criticism to the CPU space to keep pressing these companies to not be so wasteful

[@KimBoKastekniv47]

Steve please do a part 2 for all the Gregs that can't stop talking about idle power.

[@FreshChriss]

Maybe as a test game: Factorio? Its a pretty good optimised game, but can be easy to repeat the workload by just loading the save and standing still. If you get a extreme late game map (by playing, or from other sources like viewers), the cpu workload would be noticeable. Keep in mind that the game is locked to 60fps for the engine. Its also available as portable versions that dont need updates, great for version control. I dont know if the game can really produce good results, because i lack the equipment and time that you have, but that something that you could evaluate by yourselve.

[@temporalmechanic3183]

You discuss how the ultimate electricity savings are relatively negligible one way or the other (which is true), but I find that the appeal of a lower-draw CPU lies elsewhere for me. One, it means you can potentially save a bit of money (or allocate it elsewhere in your build) on your PSU selection. Watts In = Watts Out, so generally that means that it's easier to keep your CPU properly cooled while at lower fan volumes, and that heat energy doesn't just disappear the moment it leaves your case. This is definitely more unique to my own circumstance, but I live in a hot area with my computer setup in a smaller room, and there is a legitimate appeal in a CPU that runs 80-100 watts less than its competitor because it has a tangible impact on how long I can run an intensive workload like gaming before the room becomes uncomfortably hot and muggy--even with the air conditioning running.

[@BodySculptTV]

I run my 14900K overclocked to 6.1Ghz as a daily driver, in my configuration I enabled C-States with a specific value of C0/C1. Doing so allows the 14900K to drop down all the way to 800Mhz on certain cores and not all the time. In Starfield I can average a clock speed of 5897Ghz with a consumption of about 135w, obviously that's still not as good efficiency as a 7800X3D but I don't think its as bad as the numbers you are getting for the 14900k. I have even tried this configuration on Firestorm Viewer for Second Life/Opensim and at 5897Ghz its only using 75w. It varies greatly from application to application as you mention in this video. But what good is it to have a highly efficient processor when you pair it with something like an RTX4090,which can get up to 427w of consumption in certain games. I understand the need for efficiency but I see this relating more towards a corporate server farm than a homebrew PC aiming to get the highest possible FPS, its a bit silly especially when you toss in the consumption of the high end video card. Someone that is blowing so much money for a high end processor and high end video card won't be derailed by power draw. What I am wondering is "will Intel's APO yield not only higher frames but better power usage", this we will have to see.

[@xiraijuakara5988]

Is there any chance we could see something like this replicated for GPU power efficancy in gaming and rendering workloads too? I know this would be abit more complex potentially but would be incredible to see.

[@MLGproKino]

I have to say I would equally interested in seeing these numbers to see if AMD has made any ground on the competition there

[@MauricioRPP1]

RX 6600 non-XT will lead this chart. I compared average gaming performance from a big tech-review site with power consumption a while ago. This one was ahead of all other GPUs. Not that much detailed, but good enough approximation.

[@pino_de_vogel]

kzbin.info/www/bejne/bGHcd52flKx-l80

[@lefthornet]

For the 2 games, I think you would prefer 2 Games that use the CPU in a heavy way and that will not get that many updates or no updates in the future. Said that, my recommendation will be CS:GO (we can still access it through beta access of CS2 on steam) and well for obvious reasons that game will not get more updates. And also Cyberpunk, that is close to not get more updates. In that way, there are 2 games that use in a good amount the CPU, one Esport (so high fps )and one Graphically challenging. Also maybe a better combination that comes to my mind is Rainbow Six Siege on Vulkan and Cyberpunk on DX12 (so we have an esports and a Story game, but also different APIs).

[@tbard]

Maybe the Intel can take the W with idle power, single core loads or light desktop usage (watching youtube, office work, internet browsing and so on)? It would still matter if someone does both that and gaming, like in a work from home scenario? It has efficiency cores and all. IF that is true, and the margin is significant enough, it could still be a good deal for someone that does game a lot (like 3-4 hours a day), BUT also does light work at the PC for 8-9 hours a day.

[@fabianfeilcke7220]

Greg might be another alias of UserBenchmark?

[@Kuriketto]

Is Greg 2 affiliated with Userbenchmark?

[@andersjjensen]

"Shots fired! I repeat: SHOTS FIRED!" :P

[@marcusrwalker]

wow your content is always so very detailed, you must put sooo much work into this. Thanks for doing it!

[@andreasschaufler]

Great analysis. Thank you. Too bad ARM has not entered the gaming space(,yet). Really wonder how those would compare. ….and yes I paid an average of 39.70 Eurocents per KWh in Germany this year. It matters a lot.

[@phale426]

I was just thinking that this locked FPS test would be super useful for GPU power efficiency comparisons as well. Like a lot of the top end Nvidia GPUs are super power hungry when being maxed out but are they more or less hungry when doing the same workload as a mid-tier GPU?

[@Zarcondeegrissom]

yeah, is a bit ironic, people being nit-picky with CPUs going over 60 or 105 watts, yet GPUs blowing past 300 watts like it's a new more-is-better challenge. Now I want to see a FPS/watt chart going back to GTX10 series, lol.

[@MacGuyver85]

Throw in overall system efficiency at a locked framerate as well!

[@serina762]

As if I needed another reason to go with AMD. Thanks GamersNexus, Intel and especially Greg!

[@_EVANERV_]

This matters a lot to me, I live in a tiny apartment unit with terrible ventilation. I often find it very difficult to cool my PC and myself during the summer months. Coming from 13700k to 7800x3d, both on noctua nhd15 cooler, I noticed the temperature is significantly lower on the amd part, and the CPU fan is a lot quieter. It is especially noticeable on my legs, since the PC vents almost directly to my feet, and is validated with HWinfo.

[@Nbwest609]

Im so glad I went withe the 7800x3d when I rebuilt my pc. Great performance and great efficiency.

[@MrAnimescrazy]

Yup me too and what are your specs?

[@TheHighborn]

i'm only 11:21 in and already laughing like I won the lottery. Bro who made the comment, is clearly not familiar with what manufacturing nodes are, and only reads the markating slides :) (well, not that i'm an electrical engineer..)

[@stormrider01]

Conclusion: E cores and P cores are marketing bullshit.

[@ijustsawthat]

Greg: places bet against GN Snowflake: hold my cat toy

[@P4r4d0x1c4l]

I'm really glad this type of videos exist, since I've been asking them since Ryzen 1000 dropped.

[@maxthebean8047]

It would be interesting to see how those CPUs compare at Idle

[@Slimmeyy]

Glad to see different electricity prices included. I don't know our exact rate, but the average in the Netherlands is €0.40 / kWh. I might not game 8 hours per day, but we've got 2 gaming PCs that will probably add up to close to that 8 hours. We've currently got a 5950X and 5600 so I feel the difference won't be that noticeable, but it's good to know.

[@Mammothtruk]

I knew Intel was hurting on power, but I had no idea it was THIS bad. damn. I hope Intel can find a way to bring that in line in a few generations, AMD is already talking about increasing efficiency in the net three generations.

[@SebastianWeiss-c8d]

I love my power-efficient 5600x. Lovely CPU. Intel really needs to come up with something really soon. Not only for themself, but also to keep the competition up. Competition is good for the consumer ( most of the time ).

[@MattTrevett]

Room cooling/comfort and cpu cooler requirements are also a factor to me, especially if you can keep lower fan speeds.. so cheaper part, less costs with other factors, and still excellent performance is kind of a no brainer

[@pirojfmifhghek566]

Thank you for doing this kind of test. I've always been extremely curious about what the results might be from this kind of in-depth comparative efficiency testing. I'd love to see something similar for RAM and CPU pairings. Do some CPUs perform better than others at different RAM timings and megatransfer speeds? It really makes me wonder, especially when we take into account 3D v-cache chips. Do any CPUs really _need_ the fastest RAM, or is it a nonissue? It could be a big money-saver for budget builders if it turned out that RAM speeds within a generation have little-to-no effect on FPS scores. But without that data those builders are going to feel some fear and apprehension that buying slower RAM may have a bigger impact on performance than they're willing to compromise on.

[@abunk8691]

Thanks Steve. Thanks GN. I was thinking about this in the back of my head since about 9th gen with the pumping of clocks without much IPC improvements or a die shrink. I was expecting around a 0.5:1 power consumption and efficiency ratio between Intel and AMD and your tests just shown it was worse and it wasn't surprising. Probably something that contributed to the inefficiency of Intel is what they did since 12th gen with the P and E cores hybrid that kinda feels like going back to Pentium D where they glued the inefficient dies together and kept pumping the clocks at the cost of the customer especially with the degredation and instability of 13th and 14th gen. The i3 12100F was surprisingly efficient though, likely due to not being a hybrid CPU design and just having 4 P cores with hyperthreading. And it was also really nice to see the Ryzen 5 5600X also performing well since I have a 5600 myself and they're quite similar. Also I do hope Greg surrendered his paycheck for some modmats haha.

[@Divinebelf]

I'll bet my paycheck that Greg is an Intel fanboy.....

[@Bobthebuilder2235]

Can you guys debunk AMDIP/frame pacing between 7800x3d vs 14700k for example. I keep hearing about 1% lows/frame pacing being better on intel.

[@manbearpigsereal]

Hardware Unboxed did in a community post

[@Bobthebuilder2235]

@@manbearpigsereal can you pls link me

[@Habixus]

Thanks for the awesome video! That's exactly what I've asked for and I just got this video :D The only little addition I would've liked is a graph for the idle power consumption. Sure, it's awesome that my room stays cool while gaming with the 7800X3D, but I'm 80% doing office stuff and my 10600k only used about 12W compared to the 7800X3D taking ~30W. Doesn't matter that much, but I need to let fresh air in a bit more often now. while working from home.

[@HondaWagovan]

I love the absolute sass that Steve has, makes these super informative videos very entertaining.

[@johnpaulsen1849]

I really enjoyed seeing the in depth power analysis. Baldurs gate and cyberpunk I think would be good options for gaming workload going forward. Any thoughts on GPU power per frame data?

[@joshpointoh]

I wonder how Greg will explain bow wrong these tests were and how right he still is...

[@Liatin1]

Probably just need to take a look at userbench

[@ruizhewang4917]

"AMD social media marketing machine and paid influencer etc..." - Greg

[@TestarossaF110]

That kwh electricity cost chart was amazing would be amazing to be able to play with it on your website by plugging in your own kwh price.

[@liaminwales]

I am amazed how well the 3700X rates, guess that's why the consoles use it.

[@TotallySlapdash]

yeah, zen 2 can hit some crazy efficiency numbers at low power when properly configured. There was an interview recently with one of the developers of the steam deck, comparing its power envelope to something like the ROG Ally in terms of battery life and cooling... Basically nothing compares to Zen2 at 10 watts atm for ppw & that's the reason for the Deck OLED only getting a die shrink ... a more efficient low-power APU simply doesn't exist yet.

[@patrickproctor3462]

The frightening thing about this video (to me), is what it reveals about "where we are today." I JUST finished dialing in my Dad's new build today. His old 4790K (no dGPU) build is used as an office PC and for recording+publishing professional education videos, but it's so old it's practically guaranteed to have a catastrophic failure "soon". The replacement is an R5 7600X (stock, no delid, no lapping) build, and on an EKWB 280mm AIO as case intake, if you let it sit steady state idle in Windows 11with the pump running 60% and the fans 15% as baseline, the CPU WILL reach 60C within roughly an hour (at which point, fans ramp for a bit and we're cool again). This is after locking package power draw to 88 watts and stability testing a -30mV offset. And the sidewall of the radiator is WARM. Like, holy Hell. It used to be you could cool a top-end I5 with JUST a top-mounted 120/240 rad, and it wouldn't need to spin up the fans until you threw AIDA 64 at it. We are at the point where the entry-level overclockable CPUs--AT STOCK--require liquid cooling AND active radiator fans to keep them at a steady temperature AT IDLE?! And the 7600X is one of the most efficient CPUs on the market? WTAF?! Edit: btw, the non-zero minimum front fan speed is because I neither want the RAM (G.Skill 6000 MHz, CL 30 2x16), nor the VRMs (though they don't seem to care even at full load + 15% speed, stuck at 51C), nor the PCIe 4 M.2 getting excessively warm under spike loads such as opening applications, which manage to send to Tdie all the way up to 70C before it collapses right back to 60 +/- 2). And, I triple-checked the mount. Paste spread pattern is practically ideal. Thick IHS or not, the contact is great, the rad is not-quite-hot, and the chip is pushing 60. Good, lord, Ryzen Master is reporting total package power usage of 15-28W, and THIS is the resulting temp? I have no words. If Intel is THAT much worse at efficiency, I don't want to think about building another. Time for ARM/RISC-V

[@CallMeTeci]

Hey, im curious. Could you compare both companies CPUs when it comes to Professional applications? Like how well do they integrate and support the dGPU in programs like Davinci? Or how well can the iGPU transcode/encode and be used as streaming processors? Feels like a topic nobody has ever really touched and the videos that ive found about it all used intel for that, without any comparison to AMD. I probably guess there is a reason for that (AMD not being particularly known for their care of the professional userbase and things like the worst h264 quality encoder on the current market) but it would be interesting to see the differences analytically tested.

[@CESAR_H_ARIAS]

Cad rendering and fusion 360 should be included.

[@df3yt]

" worst h264 quality encoder " If quality is a thing - you don't use GPU encoding anyways...NV and AMD both suck.

[@CallMeTeci]

@@df3yt If performance matters, then of course, you are using GPU encoding. And besides streaming (mostly h264 and max 8Mbps), it doesnt matter at all at this point. h265 as well as AV1 dont show meaningful quality differences to CPU encoding. Doesnt help with AMDs h264 encoding, what is still the #1 reason why most people dont use AMD for streaming.

[@df3yt]

@@CallMeTeci Personally, I'd use AV1, allocate one core and be done. GPU encode can still cause hiccups on setups while gaming. This therefore handles performance and quality. I've setup a fair amount of streamers this way, so if performance matters it's not necessarily gpu. Haven't tested with ARC though. Maybe that works better. As for 265 - not a chance for me the biggest waste of codec. 264 or AV1.

[@Zendien]

As someone that prefers silent focused cases this will also help me figure out what to upgrade to next when the time comes. Less power consumption means easier cooling. Sadly I messed up somewhat when I went for a 9900k upgrade for my Define r5 case the last time I upgraded, lol. I made it work tho (early reviews didn't catch something something bios blasting power for stupid heat)

[@mikfhan]

NR200P with 7800x3D here :P -30 magnitude undervolt and q-fan curve from 19% at 30c to 100% at 75c PBO thermal limit right now, still boosting over 5GHz occasionally and noctua fans are super quiet around 40c idle.

[@tristanwegner]

Power use is not only about power cost, but also cooler size/PC size and especially noise required for cooling.