Who REALLY Runs Hotter? AMD vs Intel
ฝัง
- เผยแพร่เมื่อ 15 ก.ย. 2019
- We have two 8-core CPUs: One at 105W, the other at 95W - But those numbers are completely arbitrary. Let’s test them and see for ourselves which will heat up faster!
Buy a Core i9-9900K:
On Amazon: geni.us/A7fMnMq
On Newegg: lmg.gg/8KVmd
Buy a Ryzen 7 3800X:
On Amazon: geni.us/c8jRoGK
On Newegg: lmg.gg/8KVmy
Purchases made through some store links may provide some compensation to Linus Media Group.
Discuss on the forum: linustechtips.com/main/topic/...
Our Affiliates, Referral Programs, and Sponsors: linustechtips.com/main/topic/...
Get Private Internet Access today at geni.us/7lLuafK
Displate metal posters: lmg.gg/displateltt
Linus Tech Tips merchandise at www.LTTStore.com/
Linus Tech Tips posters at crowdmade.com/linustechtips
Our Test Benches on Amazon: www.amazon.com/shop/linustech...
Our production gear: geni.us/cvOS
Twitter - / linustech
Facebook - / linustech
Instagram - / linustech
Twitch - / linustech
Intro Screen Music Credit:
Title: Laszlo - Supernova
Video Link: • [Electro] - Laszlo - S...
iTunes Download Link: itunes.apple.com/us/album/sup...
Artist Link: / laszlomusic
Outro Screen Music Credit: Approaching Nirvana - Sugar High / approachingnirvana - วิทยาศาสตร์และเทคโนโลยี
![[FULL EP.69] #เอิร์นจิรวรรณ เปิดบ้านหมื่นล้านทุกมุม ครั้งแรก! l เคหสถานบานปลาย l One Playground](http://i.ytimg.com/vi/_J4_nw8c64o/mqdefault.jpg)
AMD outputs less heat than intel.
There, I saved u guys 20 minutes.
In 20 minutes... Because apparently both sides have different goals in mind. One going full throttle, the other slow and steady.
THANK YOU!
Thank you xD
Thank you webperson!
Thanks man
Anthony: why am i wasting my skills with this
Linus: im paying you
Anthony: makes sense
why do you people think Anthony is some sort of genius god 0.o it makes no sense
@@frosty9392 It's because Anthony is an everyman. No one thinks he is some MENSA genius. I for one like seeing him in more videos. He doesn't have to be all svelt or have eight PHDs to be entertaining, more so than Linus at least.
ah these comments are a bit much we all love linus and anthony but same jokes everytime is tiring and a bit cringe as well
Kevin Connolly many of those in MENSA are not even geniuses. MENSA is pretty useless, except for those who like boasting about their IQ, which is meaningless if they’re not doing anything with their ‘intelligence’
@@springbok4015 yes indeed these brainlets probbaly perform experiments from reading a manual and not do something fun by thumselves
“You know what you should do?”
Me: Skip 10 seconds?
😂😂🤣
I do this exact same thing every time! 😂😂😂
Legit man i do it in his every video's..😂😂😂🙌
KINGU CRIMSON
Yes.
Hotter, more expensive, and less performance yet they're still "the best". Now you see why branding, and advertising is so important.
Less security too, don't forget that.
its not just advertising, AMD only catching up lately, and Intel been doing much better than AMD for years
But it's still have better performance. These things are not related. Anyway, this test is absurd.
@@rodrigomendes3327 yeah this "Test" is just an entertaining video at best, not a reliable test at all.
AMD only outperforms intel in workstation use and multitasking. Intel is still better for gaming and desktop use.
1:04
Wait for it...
Wait for it...
Wait for it...
SKIP SKIP SKIP SKIP SKIP
The fact that LTT hearted that means that they are aware
Double tap on phones and double right arrow key on pc. The only youtuber that makes skipping ads that easy
don' t you mean LL?
@@applejambers7674 All their views are probably 10sec shorter than the video, lol.
csgo vids aswell :)
This video is why the scientific method needed to exist
@NigelBasset so go find some bloody scientists to handle technology that will go WAAAY over their heads. Good luck!
At least they attempted to make this as fair as possible. You want them to do this test several times and create a population of results? This one test has already created such a huge disparity - and Intel will never accept their results, anyway. AMD will harp on it for sure. And who has the time? Maybe you. What have you done lately brother?
Consciousness Vivid you don’t need to be a scientist to follow the scientific method. But yeah, I agree that it’s not suuper necessary for this type of thing as it’s more of a disproof attempt rather than a proof.
Honestly their method seemed OK. They tried to limit the number of external variables etc.
@@Mandragara Go wiki the scientific method
@@Mandragara for me, they are doing stupid stuffs :D not a scientific stuffs
Intel: We're still the fastest!
AMD: For frying eggs and bacon
Intel: *cries in higher heat output and pricing*
TFW heat output doesn't matter and never did as long as you're within reasonable temps on your cooler. Crying about CPU heat on the faster CPU is like complaining that an engine in a car with 1000HP puts out more heat than a luxury sedan. OFC it does. It's running faster...
no lawsuits against intel pale, you get what it says on the box, cough cough zen 3000 NEXT LAWSUIT INCOMING LOL
@@mmatous I own cpus from both companies, moron.
lol gtx 480...
@@TreyM_channel This is less about it mattering practically I would say and more about Intel bullshitting their marketing to make their numbers look better to consumers.
I trusted AMD when ryzen 1gen came out.. Now im proud
Dawio89 I'm with you man 🤙
I pre-ordered a: 1700x back then.. Still rocking 😂
U will be even prouder now
(I am talking about asus G14 laptop
Same man. Bought Ryzen 1600 when it came out. One of my best investments in computers ever made.
Got the 1700 when it came out, I still love this cpu running at 3.9 GHz and 1.33 volts (full load voltage). I had a choice between the 7700K and the 1700 back then, it was pretty much a no brainier.
Get an AIO, remove the fans, set the pump to max, submerge it in a measured portion of mineral oil and calculate how much time it takes for the oil temp to rise X degrees.
Sounds good!
That seems like the best route, probably do it in a small insulated cooler to keep from losing too much heat from from the mineral oil during the longer run.
@@CondoreComputing Yeah, you can control so many things...like weigh the thermal paste before application or even use those graphite thermal pads. Furthermore, use torque wrench for same pressure on the cooler. The ultimate goal here is have as many constants as possible.
@@kostasbezaitis2695 Id love to see GamersNexus play around with making this a standard test procedure for CPU's
This, at least makes more sense. Having hundreds of uncontrolled variable doesn't help.
Buys alex a workshop, still goes through the trouble of doing it janky
Linus's version of "AMD is the clear winner when it comes to heat output. Oh and AMD did finish the rendering while Intel did not!":
_That's rough_ .
@@Skeletoncross so, where did you get your assumption?
@@Skeletoncross wait 16 y.o are smart enough to use false age. I was 13 when I create my first paypal
@@Skeletoncross your comment makes no sense... What are you on about?
@@Skeletoncross ah, okay I see what you mean. Yeah that is most people. Ngl I haven't built my own but I'm saving up for build Myself. But I do know people who are "posers"
@@Skeletoncross You are one pretentious kind of moron tbh.
I'm glad this is a Canadian channel and they use the right measuring system.
No they didn't, they used the metric system instead of the US Customary units. :)
Bruh is this sarcasm? Thats what he meant lmao
Intel & Nvidia SUXXZZZ!!! We actually use both systems. It’s completely bs just switch us over to metric instead of half and half
@@ToGrimmToWin I'm referring to the mainstream people. If you used both, then good for you.
Yeah i am happy too since i am Europian
ALL HAIL TO THE FX 9590, LEGEND SAYS YOU COULD MAKE AN ENGLISH BREAKFAST ON IT....
i need that
guess ill grab my eggs
Hmmm ... (old)TopGear style: "Some say that it's the reason of global warming or glow hot red when on full load. All we know is ... it's called FX9590"
AMD : FX what?!
Move on and change your calendar to 2019.
I love LTT's science videos because basically every scene has something going on that's contributing to error in the final results
This is the real MVP comment right here.
Ya see, now that's some real world performance. Glad LTT is keeping Intel's claims up to snuff
Yeah for a well known tech channel they sure do f*ck things up quite often when wanting to do tests like these....
2:39 "Ryzen" temperature
Wow, that was an EPYC choice of words.
Melting Ice i see what u did there lol.
Nice one
Oh, the insane puns that AMD produces. Too bad Intel doesn't do that though. I don't think their Core was Intel-ligent enough.
sorry I'll see myself out now
LOLMAN 9538 lmao
Your jokes are ThreadRipping mate
Linus: Has rtx 2080 ti s on both pc's
Also Linus: Both pcs have 8 gb ram.
In single channel no less
Lol
Would ram more than 8gb required when you are playing high end games? I have rtx 2060 6GB VRAM but still have 8gb of RAM due to local RAM shortage. Is it really necessary to give it another 8gb?
@@snowmanleblanc6053 some games consume upwards of 10 GB of system memory
@@snowmanleblanc6053 in high settings and frame rate 16 is usually recommended on 3a titles
Missed opportunity: Should have used Madrinas coffee as the liquid, that way you could have made a pot of coffee,
have the caffeine help with overclocking and plugged a sponsor all at the same time.
Except that would diminish their thermal mass over time... One cup of coffee at a time.
This! I'd watch this!!!
:) brewing a coffee lake
@@charlesleonitol.iringaniv8320 Now that, ladies and gentlemen, is a missed opportunity.
1:04 *skips 10 seconds..*
Got on the ad
ME: bru
It really cockbocked me.
“35 is way to high”
Me: idles at 37 degrees 😳
That's cool tho
That's cool tho
Me: idles at 45-50 on my Ryzen 7 3700X
Me idles at 52 degrees
T-Shirt: 20$
Water Bottle: 30$
Noice Linus. Plz explain
well you cant live without water for a long time
Cotton VS high technological materials to keep things very cool or hot
I can explain: T-Shirt is made at sweatshop factories by child slave labor, but the Water bottle isnt.
because water bottle is more expensive to create what is their to explain their?
Weaved cotton vs vacuum insulated stainless steel..... what’s hard to get?
Good idea but that execution... science was thrown away fast af, damn.
@@geiokamihasuki5998 What Bias lol
Since the imperfections were same for both the processors, the results can be validated (more or less)
Linus with AMD bias? LMAO that's a first.
As of lately everything they've done is half-assed to say the least just for the sake of views. Just read through the comments in their video "this coolers draws 545W" that video is also full of mistakes too... Glad I unsubbed a while ago. Shit is full of shameless plugs and fuckery, he's slowly becoming the Logan Paul of the tech world...
@@geiokamihasuki5998 You recently bought an Intel CPU or what, lol?
They should've just used one of Gamers Nexus's Copper Cold Plate MUGS.
I was expecting them to pull one out.
They're probably having beefs
@UC5pc3n6DISo9XuaekKionbQ it's a INSULATING water bottle...
EDIT: Ey, look at that! TH-cam's being TH-cam again! Which means something in the comment section is broken XD
Of course this @UC5... gibberish was meant to link to paul andrew Apolinario's comment, lol.
They need that sponsor integration lmao
@@a.ffirmative No they're not, Steve was a featured guest at LTX just a couple of months ago, and there hasn't really been any interaction since.
Solution:
List both the TDP at base clock and the TDP at boost clock.
It also depends heavily on the workload. Just compare IBT with Cinebench. IBT can draw more than 30% more wattage at the same frequency.
@@JirayD what they should do is show the maximum tdp at running a specific Programm ..i mean intell is not wrong ..its like saying that the dodge Camaro runs 70miles per gallon and I'm sure it can run 70 miles per gallon the question is if that usage is realistic
@@DrSmugface Well, then we get to the specifics, at which temperature are those chips running, etc... The point is for Ryzen 3000 a CPU will always run above the base speed if the cooler can dissipate heat output in TDP. Testing my 3800x with prime95 Small FFT (AVX2, worst case) @ 105W PPT-> 3.9 GHz all cores. Cinebench R15 under the same conditions: 4.1 GHz.
Of course the Ryzen CPUs will try to run faster if the cooler, and the PPT, allows it to dissipate more heat.
AMD: We have efficient CPUs that output less heat than yours!
Intel: At least we save you electricity since you dont need a separate room heater during winter!!
The_DankSmith and that 'AIO winter room heater' continues to 'heat you up' even in summer 😏😏😏😛😂
Funny thats been AMD's trademark up until now "expensive heatcannon that gets tired after 1 year of heavy useage".
I still havnt seen how their Ryzen holds up after 5 years of useage. Dont get me wrong i think its about time they got up to speed with all their "this time we have intel performance without downsides", except the one that only time will really tell.
@@Rawsawn have been using ryzen 2600 . its far better than my pc with i7 7700k. It runs cooler too. Gives me better performamce too. And less power consumption too. The most of the work i do is photo and video editing. And gaming very very similar performance
By definition every watt of power that goes into a space heater turns into heat, and every watt that goes to a computer tower becomes heat too.
So it only saves you money on a space heater
On this episode, we watch paint dry, in the most convoluted way possible.
This video started off so promising and scientific lmfaooo
IKR!!!
I believe this is why "reproducible" results are so important in science experiments. Although, it's kind of fun watching all the problems pop up (fun for us, the viewers, at least).
"And you know what YOU should do?"
Yes. *skips 20 seconds*
2:37 I see what you did there. " RyZeN Temperature"
Ow my ear holes
Ow my butt holes
@@chinchy111 plural? How odd
you know amd could have used the line radeon has rizen(ryzen) when announcing the radeon 6000 series😂
they used the rtx 3000 series as a fucking benchmark
Cooling a 8 cores cpu with a cup of water, you guys are mad.
Waiting for the 3950X video of this technique.
der8auer did a test with a delidded 8700k without a cooler and it didnt even turn off it just throttled down to 800mhz and stood there on 100°C
you must be new here
@@LcFan96 so you basically were running at a atom core
@@lockinhinddanger934 pretty much
0:03 You missed chance to say “If you believe Intel’s intel...”
Cyberman3245 I appreciate the unbiased testing. It's refreshing.
Apply for a writing job at LMG.
after AMD liars, lawsuit and low boosts i find this comment ironic
+
I like seeing parts made, you guys should include this in more videos. Really cool, guys.
Not gonna lie, he got us with that typical pause for sponsor
2:09 "IHS Solder Quality" wink
nice.
I saw that, didnt quite know what he meant by it.
J
O
Y
.
2:38 - Missed the opportunity for saying that you were gathering intel on the ryzen temperatures!
The way he said it at first, I thought he was, but then he just continued, and no, he didn't :(
@@shahmask yup 😂
1:04 Linus is staring into our souls...
Linus Starring Into Souls Tips
"IHS Solder quality" *Winks at intel*
As an engineer that takes these exact measurements, this video is on par with the Verge building a PC video to me. This really hurts to watch.
I guess they aren't making a tutorial so it's a bit better. A bit... What is the proper method they should be following?
@@Edwinthebreadwin
It is worse than verge pc. At least verge got their shit of run, this produce no usable data, like zero usable data. My middle school lab has more rigid process ffs.
The process they should follow? Google up how to perform an experiment is a good start, they did literally nothing right.
Yeah it was awful.. :D
@@jintsuubest9331 no the verge claimed to make a tutorial knowing what they do, this clearly wasn't engineers doing any sort of those claims
It feels like they see the audience as dumb people that just want to have a laugh at Linus's supposed stupidity, not people that want to have access to information
Previous video: "This is the sketchiest thing we have done"
*Litterally the next video:* So we are going to just fill the cup with water and place it on the CPU!
Nathan Gehman we have a saying in our lab, the more ghetto it looks the more accurate the results are
Dont forget balancing two heavy weights on a piece of glass ontop of the cup!
7:03
Translation: "I didn't sign up to this, but I damn enjoy it."
@7:35 - Should have used a digital scale to weigh the mass (water. Zero the scale with the cup on it, then add the water / liquid / solution)...rather than eyeballing the volume.
But all the major players in the game do have the same philosophy which seems
to be "The nice thing about standards is that there are so many to choose from."
You are to stick with the one your competitor, you and several other companies agreed on.
@@yumri4 then explain why Nvidia's tdp definition is way tighter than everyone else lol
@@DSDSDS1235 As no one sticks to one definition everyone has a different one. Usually this is used for the benefit of the company selling it. I am unsure why nVidia would have a higher TDP than i can find a reason for that is repeatable on multiple rigs with multiple different set ups.
I guess marketing messed up as the max is a little over the TDP, the min is alot under the TDP unless you are in very unfavorable to running a computer conditions, the other ways of measurement used of 0.1% low, 1% low, avg, and median while in a normal 70F degree temperature controlled room do not go together with what nVidia says either.
Maybe my testing is done wrong BUT most likely nVidia like AMD and Intel are lieing about how much heat their GPUs give off. Thus it would be better if they all agreed on one way to measure it. At least then the consumer can get a guess of how much the company usually lies to know how much cooling they will need for the chip, card or board.
@@yumri4 nah. Nvidia explicitly very bluntly just don't go above the tdp as total board power. You literally needa mod the boards to get over that.
Linus, may i suggest you invest in a BTU meter,, they measure flow of water, and the delta of the radiator, so you can run MUCH LONGER, tests and work out system variable.
onicon, sells BTU meters, any proper BTU meter will have a flow meter and 2 MATCHED probes, plus a control module which will tell u BTUs, and WATTS of heat output.
Thats the proper way.
So, many unnecessary, commas, it’s hard, to read.
Just use two thermocouples and flow meter. Regardless, you need to calibrate anything.
@@z1xems "How MISGUIDED are you? I feed you a couple of bullshit legal precedents, and there you go - you jump on it like a bitch in heat." - Law Abiding Citizen (2009)
Linus'es Schtick: during next wan they will tell you that they purposefully didn't use a btu meter to make a more entertaining video.
4:27 linus perfect Zoidberg impression
:) stay away from the nuclear plants! Good show.
This experiment had way too many inconsistencies and uncontrolled variables
Edit: I'm not saying the conclusion was wrong but I expected better from these guys
I welcome you to Linus tech tips
Well, it is known that a Ryzen TDP is measured at boost clockspeed, so 105W at boost for 3800X
For Intel, TDP is measured at base clock, so 95W at 3.6GHz, a 9900Ks actual TDP is around 150W with boosts enabled.
#outofideas . LTT needs to leave real science to youtubers with college degrees.
@@curt8806 bruh, Linus has college degrees, so does Luke, it's just they are not actual Scientists, this is Entertainment in the end, not Gamers Nexus
That coil would waste so much juice
Isn't Intel's TDP measured at base and AMD at boost?
Edit: They said it in the end
If wanting to demonstrate a calorimeter (device that measures heat), the best and easiest way is to build a water cooling loop without a radiator, just have a large reservoir. From there. it's just the heat equation: specific heat multiply by volume multiply by the change in temperature multiply by the time it took and you get the wattage. You need a stopwatch and a thermometer.
Note: The calculated wattage is not from the CPU, it's only what energy entered the water to heat it up over time. This is because heat is absorbed throughout the path from the die, to the internal TIM, to the IHS, to the applied TIM, to the CPU block, and to the water. It's the "thermal resistance" spec of a loop that takes the block and the TIM application and pump and rad into account for heat absorption.
For instance, say your whole water loop has a thermal resistance spec of 0.1W/K, this suggests that at 200W, the change in temperature from the CPU IHS to the ambient air temperature entering the radiator is 20 Kelvin, or 20 Celcius. But if you measure the temperature of the radiator, liquid itself, the block or the IHS-- they may be slightly different numbers. It's supposed to be like that.
In science, you create consistency, simply pick a reference and stick with it, i.e. whatever you want to measure for every single test, do the same thing every time. Don't change where you measure temperature every time, etc. Be consistent, but also share what you are doing (don't assume) and be open to criticism (not get defensive, the idea is we all need to be on the same page).
Truth is, one could calibrate a standard air cooled heatsink to calculate CPU die wattage, as long as the test conditions are all the same. It's just fun to do it with liquid. You can even use a Peltier, the power level changes with the load as long as the hot side is kept in check. Or get a big block of copper, mount it on the die, and measure the change in temperature over time. It's the same concept; there's lots of ways to estimate load through a calorimeter demonstration.
Great video. You guys should show how much i9-9900k is drawing from the VRM. Regulated: 145W. OCed to 5.0GHz - 185W. 5.1GHz - 230 -245W
when all cores are in the AVX 256 / 512 with -2 steps in multiplier.
Hey look, it is a calorimetry experiment! Let's see how this goes.
Things gone wrong, Gone sexual
It's cool to see the channel develop. They generally improve quickly/well with critiques. I'm sure some of the crew at LTT learned a lot about the scientific method from doing this, and from doing the slo-mo work they did a little while ago.
It's an LTT video. They pretend they care about doing things right (they clearly have the right knowledge, based on what they talk about) but Linus doesn't have the patience to do things correctly (measure EVERYTHING precisely, isolate as much as possible, etc.) so they never quite get actually useful results.
@@Chaosdude341 They didn't learn anything about science or testing with any real accuracy. Linus uncuts any scientific process of technique prior or during "testing." This is a high school level chemistry lab, fail. The Peltier cooler videos are a mechanical/electrical engineers bread and butter, failed.
@@Mezuzah87 Even if he did not follow a precise scientific procedure, he still emphasized that no matter the testing and because AMD and Intel test differently, no CPU can be accuratly judged without a direct comparison. This I respect as the content of the video was to show not only how Linus and Anthony tried doing this comparison on their own, but also wanted to get the message across.
Did anyone else see Linus's little wink there at 2:10 about solder quality? Lol.
Yeah... What was that about?
@@warpedweirdo shintel bad
toothpaste
KawakazΣ and Rogeh are right.
And that's really sad.
@@warpedweirdo intel used to use shitty thermal paste instead of directly soldering their processors to the heat spreader
I am so glad you guys are doing computers and not, like, anything else that require precision measurements
1:05 glitch in the matrix, eyebrow buffer overflow.... hehe great vid thanks!
This is the most unscientific experiment l have ever seen in my life.
I hope this is bait
i sadly have seen even less scientific experiments than this
Well he used a kitchen jar instead of a measuring cylinder which ticked me off the most.
@@crowmane6306 You can go all ham with experiment to make everything lab conditioned, but they are not in a lab and I highly doubt there will be a much different results; equalize starting temperature, add inlet and outlet temperature, a flow meter and you are good to go as far as I concerned, a few ml of water differences is not going to increase the overall BTU of water by any significant amount.
Principled Technologies called, they have something they'd like to show you.
Madrinas Coffee : "Hello, here's a gift from us."
Linus : "Oh cool ! Our new cooler component !!!"
Madrinas Coffee : "...It's for drinking coffee."
Linus : "And to make a cooler ;)."
It's funny cause madrinas is so gross you wouldn't consider drinking it!
@@devins4686 why is it gross tho? Is it like a coffee concentrate or something 😂?
Kinda feel bad for them ^^'
@@silverhawkroman it's like drinking coffee mate from the bottle
LOL. I see your video on my recommended, entered to watch, and right before the video starts there is a commercial of Monitor systems by Linus
4:27 excellent zoidberg impression
Intel or AMD, my cat will sit on a computer no matter what team it is.
Freeziepuff!
Cats are as indifferent to CPU architecture as they are to human suffering. As long as it's warm and has are flow they are content, also having a servant to bring them food and pick up their droppings.
Until you get a blower fan videocard like the infamous FX 5900.
As soon it kicks in, your cat will learn instant transmission.
@@dan_loup lmao check out this video speaking of that card😂 th-cam.com/video/H-BUvTomA7M/w-d-xo.html
Cats are the only real unbiased actors among us that can't be bought....
Yet give them some food & they'll chose which ever PC you put it on.
Too bad your mounting system on the Peltier project was only for Intel CPU. You could use the Peltier to keep a precise temperature and measure the amount of power draw by the Peltier, which is proportionate to the amount of heat generated by the CPU.
We use an equivalent technique to measure with high precision the amount of heat generated by a reaction. (Organic Chemist here).
As long as you had a high quality Peltier unit that could actually keep up with the heat output. And i think the biggest issue they had with their setup was not pulling heat from the hot side fast enough. Dissipating 600 or so watts from such a small area is pretty difficult. Im not sure how much heat your dealing with in your case but if its small scale it's probably a bit easier to do accurately then it would be with a high power draw CPU. Using a power generating Peltier would probably change things a bit too since a unit designed for power generation behaves a lot different then a peltier designed to just pump heat.
I've been playing around with the cheaper medium power peltier units and they seem to take as much power as i give them no matter the temp differential. Im working on a power controller to efficiently run a peltier at a set temperature while never turning off completely so that heat isnt just transferred right back to the cold side immediately.
Or you just use Ice melting at a constant melting temp...
You know sometimes you get along without high tech. Only potential problems are ensuring the ice mass is the same before the test really starts, and that the ice can melt fast enough (but that's a question of surface area).
Also the time ratio will not precisely describe the power output ratio, because of the constant heat source of the surroundings. But with some math and a comparison test without a heating CPU you can actually calculate that out of the equation. Meanwhile your suggestion will have a similar effect unaccounted for, as your situation is... More complicated.
You don't always need to get expensive equipment just for some small tests, and for scientific accuracy for a singular test a simple setup is often better than fancy expensive stuff. It might be more comfortable for you chemists and actually more repeatable once it's set up in your labs, but these Peltier elements come with a huge setup mess that LTT would never get right.
The calculations for the ice melting meanwhile can be done with school math.
@@CondoreComputing if there shoving excessive amount of power in to the peltier than it wouldn't be so bad, a chip rated for 200W doesn't get any colder if you give it 600W more of power, it just makes it more effective at making heat. in actually they would need a much better peltier than they had, and not use chips that power hungry in the first place.
the point of the video was never to make it work, it was just a fun idea to laugh at.
@@BloodSprite-tan word homie
Intel: Our TDP value is based on the recommendation that you should never try to do anything really strenuous with our chips.
AMD: Our TDP is based on the fact that you can turn our chips up to 11 and have a good laugh at Intel.
This is just one of the reasons why I just switched back to AMD for my latest build. Don't get me wrong, my i7 3770K is still running strong and is arguably the best Intel chip I've ever owned (over the course of 3 decades), but IMHO Intel just don't make them like that anymore, and it's nice to return to the 'awesome bang-for-buck' feeling such as I had with the Athlon before it.
you are comparing intel based on the 3770k? shit the 10900k is a huge upgrade next to that
@@psttech4290 I'm still using my 3770K (as I write this, in fact) due to the fact that it's just so damn robust, and nothing Intel-wise (up until possibly this new 10K series) had any truly noteworthy performance boost for the workloads I have that was worth the excessive amount of money required to upgrade it; at least with AMD I can still upgrade my chip by a fair amount without forking out for a new motherboard again. It's not about how much of an upgrade something is on paper, it's about real-world bang-for-buck, and in this regard Intel has failed for years.
Well Intel didn't really have a competition untill AMD brought their first Ryzen series, so Intel could afford keeping us at barely 4 Cores and only bring some small to unnoticeable upgrades here and there because of the lack of competition.
Intel gen 11 and 12: *sure Jahn.*
@@Hardcore_Remixer yet back then, you barely even needed more than 4 cores. That’s why 2nd and 3rd Gen i7s were the kings. Meanwhile there were other variants that were B-side but still did good for what they were.
Even now, you probably only need a 6 core CPU now, for most higher spec games anyway. People buy AMDs stuff more for workstation stuff, mostly thread ripper and epyc stuff, and sometimes the Ryzen 9 cpus. I barely even see anyone talk about how the new Intel Xeon W chip is great. In fact…. No one besides the top line companies like Microsoft or Disney would use then old Xeon CPUs that go above 10 cores or whatever.
Anthony's facial expressions at several points during this video were pure gold! GOLD I tell ya!
I think Anthony and I are on the same page when I say, "This video was a complete waste of time."
that's like 95% of videos on this channel.
Spooky711B - I disagree. We learn something new with every weird experiment. You are clearly not an engineer.
@@andrepretorius6127 yeah lol.
Andre Pretorius well then I think that the team needs to put more into their concepts before making a video of it
Having over 9 million subs there is no argument...Linus is doing something right. And if you watch Linus enough, you'll know that he is about having fun with crazy experiments "in addition to" actual, useful info. It is what makes Linus, Linus. ;)
So many holes in the thought process for this experiment
theyre a bunch of college drop outs. what do you expect?
@@curt8806 that doesn't make sense at all.
@@curt8806
But it is still fun though
ilantzur02 true
Yeah they said that in the fucking vid ya dingus
When you win this battle, could you get amplifier manufacturer's to do the same please. Thanks.
I solely watch LTT vids for Linus to randomly make his “LTTstore.com” pitches. Dude also has a knack for sliding many other subliminal things in their videos too! Hahaha keep’em rolling!
You might get kidnapped by Intel again after this lol :)
I could hear this strange, howling scream from somewhere in the distance. I think it was Gamers Nexus.
As for me, I would have built a closed loop, with a set amount of water, and then have a temperature probe monitor the temperature over the course of a set workload. Assuming the loop is the same, and you make sure you run the tests at the same starting temperature, you should be able to measure any difference in heat being dumped into the loop over time.
Yeah, it's probably GN... because there is some serious scientific error in this video... Linus, please call Steve from GN and ask him to explore this in more depth, maybe on a collaboration video.
EXACTLY.
Yes. Collab with Steve. In Steve we trust.
But this is more fun.
I can hear Steve all the way up here in Canada
8:40 "Once we get coolant flow, the CPU heat will cause more convection" No. What Linus essentially did was mix up the cold coolant at the bottom with the warmer one at tge top. Convection works well enough on it's own. Granted, it makes temperature readings more accurate, but it makes the CPU reach high temps faster, because it now has a 40-50 degrees coolant at the base of the plate instead of 20-30 degrees coolant. Here it didn't make much of a difference because Intel got smashed anyways, but if the race was more neck at neck, that extra time you would have had if you let the coolant sit would be the game changer, because the more time they'd have to output heat, the bigger the difference would be.
The waterbottle scene was probably one of the best segways from Linus recently.
there is a ryzen temperature
no pun intel ded
Intel TDP'd their pants
Perfect pfp
*GROAN*
Best pun ever.
I can see so much scientific methods are introduced into this experiment
That 10 second wait where linus seemed to forget the sponsor placement, i kinda thought they were going to surprise us without one, just the awkward pause cause he forgot to say the words out loud.
fun stuff! Years ago I tried to overclock an amd cpu and I could not get it stable...after years of trying with amd I got a 3570k, I was able to OC that cpu to 5 ghz on air and it's still running to this day...I am thinking about building a new pc and it wasn't until recently that I'm thinking about going AMD once again. Videos like this are helping me make that decision.
I like how i got the notifcation for this video when i am watching it
I got the notification 18 mins late
Easy TH-cam skills
Stonks
You know what else we should do besides raiding Area 51?
We should raid youtube headquarters
Same lmfao
Remember guys (Take notes LTT Team):
Planning is half the work
I am sure, derbauer wont loose a word about that ;)
Nah. Watching the tests and failures are interesting for me as well
Just as a general rule, gloves should never be used while operating any rotating machine.
such as sanders, grinders, lathes etc. They are much easier to get caught and pull you into the machine.
That segway was pretty good.
5:14 'our coolant is a bit chilly'
I sure hope so
Please wear dust masks when sanding things like that.
Hi sarge
+1
Haha that wink on IHS soldier quality line ;)
It is no longer 'linus tech tips' now it is 'linus adventures with tech'
Imagine what mankind could learn from these kinds of experiments if Linus and team would actually prepare.
I love the concept of this video but unfortunately, you absolutely butched the execution 😔
I think they should've poured more dye in. That must've been the problem
Wasn't that the point to begin with? Running this test with proper methodology is quite boring (just measure the power consumed via EPS cable) and is barely videoworthy.
@@user-qc6pg6ef6h I beg to differ, that sounds way more interesting to me.
Edit: I'd like some real data to interpret trends. If the data is junk it's just reading tea leaves.
@@user-qc6pg6ef6h Just for curiosity :
What are you assuming to accomplish by measuring only consumed power ?
How would you find about the correlation of the consumed power turned to actual cpu cycles (and in more depth of core efficiency) ?
@@tracer822 Power equals heat. Wasn't that the idea? Now great does not necessarily show how hot the CPU gets but again there are much simpler wats of measuring that. Depending on what exactly indeed is selling to find out some of the variables could be controlled, be it frequency, workload, time under load or whatever else.
great testing !
I'm gonna build a new PC soon and as a CPU I pick the AMD Ryzen 7 3700x which has a 65w TDP as they advertised. The biggest reason for choosing this instead of the 105w ones is the cooling solution, I want the wraith prism cooler and I wasn't very sure about its cooling performance compared to after market coolers. But this ones TDP is quiet low and I'm pretty sure the stock cooler would be enough. Also yes, Intel CPU's, even the older ones (I have a i5 2500 right now and have 95 w TDP) have higher heat outputs, I never managed to use a stock intel CPU with a stock cooler. I use a big ass after market cooler and even with that, at a full load during a test or gaming, it easily reaches 70-75 degrees Celsius (in the summer). I'm talking about a 9 year old CPU with a huge cooler top of it.
I've been loving all of these videos with Linus and Anthony! Just great back and forth commentary between the two. Keep up the great work guys!
I can really recommend today's sponsor. Revived a 500 GB drive after accident formating.
It took hours, but it worked 100%
Any brand drive can fail.
The look on Anthonys face when he realize the caps arent the same size
Anyone else love how Linus shamelessly plugs in sponsors and his own products in his videos? haha
LTTstore.com
Do you say no to a company if they offers you thousands of dollars just to show their ad in your video for just 10-20 seconds?
They need ur support during this crisis
Go away Linus with your alt account
LTTstore.com
Anthony is the MVP of Linus media group
Watching you guys struggle with this makes me feel more secure as a human being.
My heart jumped to see the sight of Anthony in the background.
I can honestly say I've never used TDP to make any decision on what processor I buy, all that matters to me is out and out processing performance and price, not how hot it is.
Well if its very hot then itll be hard to touch
This video was more painful than the sponsor segments, congrats
This testing methodology seems overly complicated, and very unscientific. I'm not sure the result mean anything.
Yeah. Just take a big reservoir of water at defined temperature and pump it slowly through a cpu block while running on 100%. Once everything is in thermal equilibrium just measure the temperature difference between water going in and out and if you know the flow you can easily calculate the heat output (ignoring heat loss through other channels).
Or even easier is just using a loop in equilibrium, however, would be better if both systems would run on the same water temperature.
You must be new here
It can't due to to many parts being different even they said in the video they are changing the speed to give AMD a "fair chance" to that they should have done it with both speeds with both sides then
Dont worry they dont
sorry guys, but this time you really did a bad job
There are too many variables. Very difficult to test these two different CPU architectures side by side. However if you could have put in a copper spacer block (with the same mass) between the cooler on each CPU and glued a temp sensor to it I think that may have worked a little better?
I’m looking at both of these processors, would a ENERMAX ETS-T50 AXE be able to keep both of these cool?
I really like that awkward pause. It's so professionally done.
Watches Linus’ segway, skips 10 seconds , sees no LTT intro!?!
I did actually skip back to see if I missed the intro + regular "quote", but I didn't. The editor must have forgotten it :-D
@@wildonemeister me too, got too used to that, it felt wrong lol
Isn't it spelled "segue" or am I misinformed?
Man, you can feel that wind-up in Linus' sponsor pitch a mile away.
wow so much fun, thanks for including all the 'challenges', am considering running a rad in a water vessel that I can fill with ice blocks.....😜
you hv actually sorted something out for me too re: Ryzen pression boost, cheers! o7