Apple Shenanigans 2.0 - This needs to stop!! 🤦‍♂️ | M1 Ultra vs NVidia RTX 3090 Performance/Watt

[@kb420ps]

Apple's marketing department is a BEAST! Like I said before, Apple's marketing department never lets the TRUTH stop them from marketing their Macs. If Apple marketed a 4 cylinder car, they would praise it for it's fuel efficiency and then they would create some misleading chart that says it's more powerful than a Bugatti Chiron. Then, somewhere in the small print, it would say "we removed all but one of the spark plugs from the Bugatti's engine for our benchmark". Technically speaking, they wouldn't be lying, but it would be very deceptive, and not representing real world results.

[@theTechNotice]

Yep, well said ha ha

[@megadeth8592]

stretching the truth is just another form of manipulation and lying.

[@gopalakrishnanr525]

That is not BEAST marketing, that is called cheating

[@ThatNorma]

@@gopalakrishnanr525 its still beasts marketing because regarding marketing whats important are the results. In that regard apple marketing team is quite kind as u can see the shitloads of sheeps that still buy their products blindly nowadays.

[@norwegianblue2017]

PC user: Explains how the numbers used in Apple's advertising is misleading and false. Apple user: I want the white one.

[@theTechNotice]

🤣

[@tobytoxd]

LOL 😂😂😂

[@SuperRJ2012]

😭😭😭

[@lemao851]

🤣😭💔

[@temporaryblip4309]

Problem is, his testing is horribly out of whack. Says 81W at the wall then assumes that all that 81W is the GPU and divides that by the Blender score? LOL. The M1 Ultra is a 60W part TOTAL - for both CPU and GPU. That's just the first thing incredibly messed about this 'analysis' - and you swallowed it whole.

[@Thevikingcam]

And the funny thing is, you can undervoltage the 3090 to 200W and still have 90% of the power.

[@Real_MisterSir]

Imagine undervolting it to the level of the M1 Ultra and still beat it by a landslide at its own power rating XD Apple is such a scam company its hilarious

[@Thevikingcam]

@@Real_MisterSir i just tried it with my new 3080 that was 750€. Hehe. Pulling 125w and still beating M1 ultra, not much but still. But with full overclocking at the 3090 level pulling 320watts. So i can go full on war or just hang around with 25Db aka 100% silent PC. With water cooling CPU and no fans running. Cos my background noise levels are 32db.

[@Real_MisterSir]

@@Thevikingcam Exactly. And beating is beating - plus you're doing it for far cheaper. Imagine if AMD came out with a new gpu that was both cheaper, more efficient, and more powerful than NVidia - people would cream their pants. So in this regard Apple is truly beat on all metrics. It's crazy what a marketing team can accomplish where their engineers fall short.

[@Ramattra-ow]

Hey bros can you tell me if we undervolte a card it will be reduce its life or not?

[@Thevikingcam]

No problem to it. It will actually give more life time to the card. Just dont over do it.

[@gopalakrishnanr525]

Apple - No-one is gonna read that huge document and know the truth Tech Notice - May I come in 😎😎😅😅

[@theTechNotice]

🤣🤣

[@HDRPC]

Apple also lied when compared i7 1260p and m2. They shows full tdp of 1260p but according to max tech m2 CPU draws 24 watt in cinebench test and gives 8600 multi core. 1260p gives 8871 in 28 watt power consumption in cinebench r23 test. 1260p gives 10700 if we choose higher tdp laptop than 28 watts. Even m2 MacBook can't control heat as per tech notice channel. It goes beyong 102 degree Celsius and gives 8600 multicore cinebench r23. Why everytime they do this. We all know Apple chip are so much good. But don't they throw fake information about intel chips

[@HDRPC]

@Muhammad Ehtasam yes in late 2023, Intel will lauch 5nm 14 gen core i9, i7, i5

[@HDRPC]

@Muhammad Ehtasam ok

[@HDRPC]

@Muhammad Ehtasam for me it is very easy to believe that Intel and nvidia is industry leading whatswhy except so call youtubers, all the vfx cgi 3d studios, 3r designers use windows.

[@RiceCubeTech]

@Muhammad Ehtasam nm isn’t industry standard. They’re not comparable in any way on a per nm basis. Some 10nm processes are leagues better than other smaller architecture. Intels 14nm was still competitive with AMDs first 12nm processes. NM was used as a metric for marketing. There is no real use for it to actually tell true efficiency or performance.

[@mtmmf7632]

Love your videos. Properly detailed benchmarks unlike other channels that just take Apples word for it…

[@daveh6356]

GFXBench 4K & Affinity Photo Benchmark (where was that by the way) prove Apple's hardware capability. The rest will come - but not on this channel. As discreditation becomes harder to reach for - watch how he squirms.

[@temporaryblip4309]

This video has the worst methodology I've ever seen. Even dodgier than Apple's dodgy graphs.

[@adelmf6134]

Apple have always been very vague with their marketing esp. these M1 Ultra and M2, they are disappointing compared to M1 and M1 Pro/Max

[@DragonOfTheMortalKombat]

yes, M2 is straight lie, 37% more power for 13% better performance? M2 is literally an overclocked M1 with 2 more GPU cores and higher price.

[@brahmasthaandipramana7591]

@@DragonOfTheMortalKombat more heat also

[@cavernadigitalpro]

I’d buy mac studio just for gain space and a more quiet environment, sometimes my tower pc gets full fan noise just for opening chrome

[@uribak9144]

Even when you prove to Apple fans that Apple is making fun of them, they will still tell you that the Mac Studio is more efficient and cost-effective. So instead of performance per watt, let's say less performance per dollar, because by Apple's logic, power consumption is more important than performance per dollar paid by the user. As usual excellent video.

[@brahmasthaandipramana7591]

All fanboy is actually dumb.

[@alpenfoxvideo7255]

man I can’t imagine doing these tests myself, I would burn out after few minutes! that is an insane amount of work

[@daveh6356]

Why would anyone run so much unoptimised software (unless they were proving a point)

[@mortenthorpe]

What’s even worse, is that the huge gaps your test show in performance ( raw performance), means that the 3090 will complete actual tasks about 3-5x faster than the M2… which means that the collective power draw on the M2 will last for 3-5 times longer, making the collective power usage 3-5 greater than any of your benchmarks reveal

[@yogiwp_]

As always, assume they lied until we've got evidence. Thanks for confirming. Apple silicon is a great progress and can stand by itself without these nonsense, which actually dissuades people from taking part in the ecosystem.

[@Real_MisterSir]

but they have to keep the lie alive for all the people already trapped in their web. That beautiful lie, the influencer hype, the superficial fluff that makes up Apple's entire brand image. It's all based on a shallow facade - like someone wearing 10 layers of makeup to always look artificially good for everyone's eyes.

[@hughhepker4897]

Another Pc Fan boy , windows still sucks that won’t change in 50 years .

[@FBL9982]

one of the key proofs is cinebench that you see the real performances ... unlike geekbench which is fake and Intel has practically buried M1 and from time to time, put your heart in peace Apple user and above all you will always remain only a miserable 8% worldwide

[@Waldherz]

@@hughhepker4897 Just because you never understood how to use another operation system doesnt make it a worse system. I will rip appart any of your work times on any of my Windows systems.

[@AA_OnTheEdge]

It's like saying a Toyota Yaris is more fuel efficient than a 911.

[@mtmmf7632]

And saying in the first 0.2 secs of the foot hitting the throttle the Yaris makes the same power while delivering more fuel economy… 😂. Dammit I own a Mac Studio…

[@mikaelbiilmann6826]

Not only that, now Apple's new M2 base model Macbook Pros's SSD are slower than the M1 SSDs....

[@eylonemuskson4177]

Ever since I saw the AppleSilicon release, I've been praying for a chip manufacturer to produce a desktop ARM/RISC-V/MIPS processor. AMD, Intel, hell...even HiSilicon in China. ANYONE. Because as much as I love MacOS, and the Mac design language (Thank you Jonny!) ....their anti-consumer bulls--t really riles me up.

[@thestroller7142]

Another great video as always but most importantly....almost 100k subs!!! Happy to see the channel growing :)

[@entumonitor]

Using afterburner in that way kills the gpu's performance to drop to that voltage. Not the best way to get the best performance with lower voltage. You have to use the curve editor. The difference would have been greater.

[@redbunnyclassic]

...they might have as well measured it in coconuts.🤣🤣🤣

[@theTechNotice]

🤣🤣

[@SamPandiangan]

I can not imagine placing that GIGANTIC PC on my desk. Compare the size of that PC's fans alone with the size of the Mac Studio.

[@Tigerex966]

Apple is saying if you use their wattage which you can't, they will perform as listed but will not tell you what test they use.

[@blion3d]

the are so many companies that uses racks to put those mac studio or mac minis, and for those companies and technitians the performance per watt, space and stability are the most important thing

[@brahmasthaandipramana7591]

What are you talking about? Less power doesn't mean more better at performance per watt.

[@Nicc93]

I think they used a gigabyte gpu. Here at my local computer shop all gigabyte products massively on sale lol. many issues with getting low numbers in watts etc

[@brown2840]

You're first 8k upload? I can't experience it, but love it! :)

[@rodhester2166]

I have no idea watt apple is doing over there with that chart.. .. lol

[@theTechNotice]

Ha ha

[@Serendipity4477]

About the GFX score where the Mac actually beats the 3090 it woud be interesting to know how high the GPU usage actually was. There is a point where things like the API/CPU and other things are just hard limiting the GPU and at nearly 500fps that is probably already the case. But on the other hand something like Tomb Raider seems to run surprisingly well on the M1 so maybe there is something bottlenecking the 3090 in some games.

[@aerofart]

In general, I applaud Tech Notice testing here. HOWEVER, the use of a consumer-grade POWER METER for capturing PEAK levels is dodgy. There is no way that device is sampling power at a high enough rate to capture Apple’s elusive “At its peak” claim. However, the reviewer dutifully observes that the overall power levels “don’t even come close.” While TN’s testing does not prove Apple’s claim to be false, it does reveal that average power consumption is nowhere near Apple’s peak value claim. Again, this assumes that the consumer grade power meter is capable of the accuracy required to dispute lab-grade power measurements. And yes, I am biased to believe that Apple may be “capping” 3rd-party software, or reserving access to its chips highest performance tiers for its in-house developed software (e.g. Final Cut Pro, etc.). Overall, great evaluation here. Ciao!

[@chrisk6474]

This has to be Apple using their software to complete their stated “benchmarks”. That’s the only thing that makes sense. So they are using Final Cut to achieve the results. Let’s be honest real world and manufacturers stated performance are skewed if you are aware of the parameters that they are using. It’s like the mpg benchmark in cars or the phone benchmarks that are heavily leveraged towards a custom rom running on the phone. GPU manufacturer do it all the time with custom bios that are only available to reviewers. Can these bench marks or something similar be run in Final Cut or anything else in the OSX eco system? Otherwise it’s apple marketing shenanigans.

[@theTechNotice]

Well, if they compared it in FCX then how did they get that running on PC? I doubt that's what they used for this graph...🤔

[@djthrottleboivoltagecontro2145]

@@theTechNotice maybe they're running hack os x Catalina and are benching the 3090 lolz

[@valimalidudu7991]

Like people actually spend 4000$ to save wattage.... I think that the macbook is unbeatable in the laptop market... but this mac studio... what is that??

[@mtmmf7632]

I love the design and from a CPU perspective I’m glad Apple (and AMD) are pushing Intel. But the Apple “marketing” is terrible… borderline lies…

[@Real_MisterSir]

PC: Spend $1000 less, get far more raw performance, also get better performance/watt if you undervolt components, get better airflow, get more options, get the possiblity to upgrade down the line. Apple: But ours is small and made out of Aluminium please buy it Me: If the Studio costs 4k, is small and has stupid low tdp and raw performance.. and its main selling point is its small size and build quality but it's locked away and can't be serviced/upgraded.... Why am I not just buying a laptop with the same hardware for less money?

[@huy285]

let's start a class action lawsuit. they cannot get away with this shitty marketing graph lol

[@DailyCorvid]

That Apple power graph is absolutely ridiculous, it's not even a proper graph it's just a styled graphic of a graph lol.

[@theTechNotice]

LOL

[@DailyCorvid]

@@theTechNotice This fish represents a Windows PC, and this cruise missile represents the Mac Pro. And this small fishing vessel here represents your workload! * Cue animation and music. Motto? Apple gotta Apple. Thanks for the video, very interesting to hear. I was an Apple guy once upon a time...

[@veikkaheikkila7568]

What's even more dodgy to me is that you are supposed to use a graph with absolute performance and absolute power usage, using wattage as x and performance / wattage as y is going to get really skewed results

[@theTechNotice]

🤯🤦

[@veikkaheikkila7568]

@@theTechNotice Yup, it's, like, exponential results or something

[@adamsaran9033]

Thank you for all your detailed comparison videos. Just found your channel as i was looking for PC builds. I have been using Macs for over a decade now and just got fed up with them after I purchased the Mac Pro two years ago. Biggest waste of money. I was about to purchase the new MacBook Pros but seriously thinking of moving back to windows after watching your honest no BS videos. Where as Apple are literally full of BS. Their brand is based on feeding their BS and I feel for the BS.

[@denvera1g1]

22:45 I bought a 3950x and undervolted to about 70w because it is better performance per watt in handbrake than a 3700x/3900x Basically, unless you're using a laptop, on battery backup, or are concerned about the power you're drawing from a circuit, you probably only care about performance per watt in the summer

[@johnnyanton305]

For creators I always say the following "Apple is ready to go out of the box, if you don't mind tinkering building a pc or even buying one will run circles around it for the same price." This is for lower level video creators, and all music production creators. Obviously a beefy PC is going to better at a higher level video (i dont mean resolutions, i mean effects, anything else really) than Mac.

[@theTechNotice]

I'd agree with you, but on the audio side of things, a lot of high end audio guys are having issues with the m1 ultra and the new silicon, have a look at Neil Parfitt's video: kzbin.info/www/bejne/hKXUkH6wg8qkptU

[@haysoos123]

If you’re talking 3D vfx currently, then yes, Apple is still behind here. But the media engines of the Apple silicon line has no equal in the PC world, especially when leveraging prores. So in high end video, macs still dominate.

[@djthrottleboivoltagecontro2145]

@@haysoos123 hackintosh much?

[@denvera1g1]

"At peak power" Basically what they're saying is that their GPU is being throttled to not allow it to reach these power levels. A factory undervolt if you prefer that terminology. If you take a 3090TI or 6950XT, and then undervolt them you dont lose a whole lot of performance, but you do gain a whole lot of efficiency. The same is true for my 5700XT and GTX 1080, undervollting my 5700xt only saw a ~10% performance loss, but there was almost a 40% reduction in power.

[@giorgosstavrou8515]

Interesting experiment and definitely worth mentioning attempt to delineate the issue. However, reading the text on dodgy graph, I would like to mention the following: it does not say anywhere in the text that the watts on the X-axis is power of the GPU. It could equally be the GPU performance per M1 Ultra device watt. If you possibly set the total power consumption of the PC to the same power that Apple is consuming, then it would be interesting to see the results. If someone notices that the consumption on the PC is more because it is using other parts that consume energy, the answer could be yes because you need all these parts to make the PC work. This is a difference on the architecture that needs to be taken into account. However, the problem here is that I am not sure if the PC will be able to work with the same watt. Nevertheless, nice try.

[@yannisgk]

pesta re Giorgare!!! (vgainei o ka8e sterhmenos kai leei ta dika tou!!!)

[@giorgosstavrou8515]

@@yannisgk e ma. Exei paraginei to pragma. H sugkrinoume sosta h den to kanoume katholou.

[@NicBam]

I wouldn't be able to use a loud PC in a studio... My maxed-out Mac Pro is a silent beast. 28-core, 1.5 TB of ram, Two W6800X Duo with 64GB. I mean, come on! I can edit in three DAWs simultaneously, each with 40-120 tracks, and mix them into my video editing software simultaneously to produce all the versions of files I need in one move. Before this Mac Pro, I could NEVER do that with a machine this silent or this fast. I'm bouncing down features in 2 hours (practically real-time.)

[@denvera1g1]

I'm a neither fanboy. But i am amazed that AMD can some so close to Apple's 5nm performance-per-watt when comparing the M1 in my Mac Mini to the 7nm 4750U in my work's thinkpad. Considering Apple has a 50% node advantage(according to the company that makes both processors) I was a bit upset that my work's laptop, was not only faster than my M1 Mac Mini, but also used less power to do the same task sometimes. I asked premising to bring the laptop home just so i could be sure there wasnt something weird about having 3 phase power at work causing the power meter to read differently, but to transcoded a recorded news broadcast from MPEG2 to H265, too 4 minutes longer on my M1 while the M1 used 18% more power. My M1 used 33Wh to complete the transcode in 1H9M, where as the work laptop, which also had to power the screen for part of the test, as well as the much smaller, much louder fan, only used 28Wh and completed it in 1H5M

[@TheGameBench]

It's User Benchmark... def THE definitive industry standard.

[@HDRPC]

Sir when comparision introduction started by describing specs of PC, i love that bgm in this video

[@nyambe]

I thought the M1 Ultra was an integrated SOC, in which case comparing it to a dedicated GPU brings out some questions. How would a 300w M1 Ultra perform? Can the RTX be limited to using only 160W?

[@kuyt20]

He's not comparing the M1 to a dedicated GPU to simply make Apple look bad for the sake of it. He did it because that is what they did in their claims. And you can rest assured that if the 3090 has a better performance per watt than the M1 in general, its will most likely be better too at 300 or 160 watts.

[@nyambe]

​@@kuyt20 I think those chart are mostly for developers. It's showing potential. Current software is made for 3090 type power hungry GPUs. No worries as developers will soon make multi core GPU benchmarks. As I said, its like comparing film to digital or Gas cars to electric cars. When comparing, if you measure from what film or gas car can do.. the race is close, but if you measure from what digital photography and electric cars can do, it's a no contest. He is compering everything from where the 3090 shines.

[@kieran9882]

@@nyambe Are you under the assumption that a 3090 has only one core in it?

[@nyambe]

@@kieran9882 no

[@kieran9882]

@@nyambe What do you mean developers will soon make multi core gpu benchmarks?

[@salteveline]

Apple's marketing team needs to be let go 😆😆😆

[@rydeentokio]

Kinda reminds me of Sega's infamous Blast processing. The Genesis had faster CPU than that of Super Nintendo, but lacks flexible 2D transform aka Mode 7. It was just the difference in architecture.

[@stefankarlsson8215]

Electricity is getting more and more expensive . . .

[@yannisgk]

😉(exactly!!!)

[@athalsteinn3745]

Excellent video. Thank you for doing all the testing. I think overall it might have made more sense to have the total power draw from the PC in your initial calculations since you are not able to distinguish the watt usage pattern on the M1 Ultra and thus compare with its total power draw. I would like to see those numbers. I do not however disagree with your main points. Apple marketing 101 is clearly on display with this product.

[@Jerryossai]

I think their argument is that the M1 and M2 chips are SOC so you should take the power draw as one. But you are right to say it's totally misleading since it doesn't come close in terms of raw performance. Great video.

[@8bitbender495]

My new Intel 12600k with IGPU pull only 81,38 watt in GFX benchmark in HWmonitor

[@enigmawstudios4130]

Apple's entire website is dodgy but pretty... Pretty Dodgy

[@techlitindia]

Hey Tech Notice, I've been watching your channel lately in search for the answer that which computer will give more value when compared to performance per dollar (rupee)? I want to get a computer for video editing which I can use for the next 7-8 years and your channel has cleared all my doubts so far, your in-depth understanding and comparative study is amazing, it helped me a lot. My good wishes are with you, I hope you will continue to help in increasing the knowledge of people in this way. You are so good that you might be the next "Linus Tech Tips"

[@slipoch6635]

Geekbench is pretty lopsided in favour of Apple. p.s. I love the music yyou chose over the adorama spot, who is it?

[@RiceCubeTech]

Other than video editing and music, the new chips aren’t anything special. And music is only great on the new chips because it can use Logic Pro which is Mac exclusive. So really other than video editing if you don’t use logic pro , there is no real advantage for these chips. If its a desktop power draw shouldn’t be that big of a deal. The real advantage to this architecture is only for laptops. They’re cool products but for their price, it doesn’t make sense for people to switch. Plus, the moment that intel supports quick sync on their new ARC GPUs and offers a 4:2:2 encoder/decoder, PC will be competitive in video editing again. 3D rendering, machine learning, virtual environments are all still faster on Nvidia based GPUs so for MOST professions that require powerful hardware, PC is still the better buy. photo editing, word processing, etc. can all be done on cheaper devices anyways so it shouldn’t even be in the conversation.

[@pham3383]

bruh,that is just video playback,wait until you add a crap tons of animations,editings,stablization,btw if you want good music,get a dac from senheisser

[@TheErraticTheory]

But how many watts did it cost to do a task. With energy prices on the rise and they may ration your power in the future what can you get done

[@twoeggcups]

It's quite possible that this ISN'T a GPU specific benchmark in the first place - the M1 Ultra has such a lot of high performance CPU resources, that there's a good chance that Apple's graph DOES stack up if the whole GPU/CPU system is considered. I think people are getting fixated on the 3090 and ignoring the 12900 CPU in this graph. This TERRIBLE and MISLEADING graph.

[@theTechNotice]

But still, what benchmark was it that uses both? And why taking about GPU power? Is like doing a 100m race and counting who does less steps? 🤯🤯🤔

[@twoeggcups]

@@theTechNotice I’d say there’s a pretty good chance it was an Adobe focussed benchmark, as the mix of M1’s abilities seems to be finely tuned for typical GFX workloads. The Apple notes show that there are two windows systems benchmarked, and they are named after the GPU used.

[@twoeggcups]

@@theTechNotice You are literally going to have to ask Apple, as it was they that conducted the tests. The graph itself remains an atrocity.

[@brahmasthaandipramana7591]

@@twoeggcups Ask Apple? Like they will answer your question.

[@jerryren748]

Apple be like: I'm the industry standard, so imma just gonna compare 3090 on apple exclusive software and compare it against M1 ultra on apple exclusive software. Because I'm the ONLY industry here.

[@temporaryblip4309]

This video is hilariously bad with shockingly poor methodology and totally lacking in common sense. The M1 Ultra is a 60W part inclusive of CPU and GPU. It is faster than Core i9-12900K - a 125-241W part - and it is on par with an RTX 3070 - a 250W part - in raw compute. So it doesn't take Einstein to work out that it takes nearly 500W of Intel and Nvidia compute to match an M1 Ultra at 60W peak. I agree Apple's charts are confusing, but the fact is they are actually being conservative in their performance per watt claims.

[@galaxiusstudio9544]

preformance is preformance but the most powerful system here in this earth is our brain

[@denvera1g1]

21:00 At 130w from the wall, i'd estimate that the system is internally using 108w split across the GPU, CPU, RAM, SSD, fan, networking, and board transmission losses.

[@denvera1g1]

@Muhammad Ehtasam Yes this does make it harder to judge how much power the GPU alone is using

[@stephenpainchaud1099]

I have not read all the comments, so maybe what I am about to say has already been noted. Nowhere does Apple explicitly say they are comparing GPU vs GPU. It is implied, but maybe it is not what they are doing. Perhaps they are comparing the internal video processors vs. the 3090. BTW, like the video, good job.

[@theTechNotice]

Thanks :) Well actually they specifically say M1 Ultra GPU, that to me is pretty specific of the GPU only performance. If they would've said M1 Ultra Mac Studio, yeah that's the whole computer... that's what puzzles me, why go after the GPU claim? Just state M1 Ultra and the mac studio and compare video encoding, coz that would've been the biggest performance leap from Apple's side.... 🤔🤯

[@Stouzough]

The reality distortion strikes back at apple. I switched to windows this year.

[@jnbennett6496]

thx 4 exposing the rotten Apple, great post

[@theTechNotice]

😇

[@AldoV]

Wow this is ultra scam by Apple !!! Sadly apple marketing lie worked I and most fell for it. Great work!

[@summertyme5748]

What's the 1/4 mile time tho?

[@MoreMeanThanAverage]

Ngl but any product which performance is measured in coconuts is a straight buy from me dawg.

[@MacinMindSoftware]

No big disagreements but just providing some of my own info. Apple's graph is misleading and useless. I did my own tests using what I have here. Used terminal, sudo powermetrics on Mac to get GPU wattage, GeForce Experience to get wattage on GTX 970. I've confirmed the Mac Studio M1 Max GPU power using another utility and using the delta of an outlet power draw meter. Geekbench 5 Compute OpenCL approx: M1 Max 10W 60K (Metal 69K) GTX 970 100W 30K. The power consumption varied a lot on both so that's an average and both are low. They both can draw 2-3X more wattage with other software. M1 Max 20x better performance per watt vs GTX 970 in this benchmark? Extrapolating a bit, the 3090 is 4x faster than the 970 and draws 2.3x more power so 3090 is perhaps 1.7x more efficient. That's not going to make up the difference even when factoring in that as more power is used, both get less efficient. But there are so many factors and use cases that will demonstrate different results so take it for what it is. I still game more on the PC but I also do save $10/mo on power with the new Mac vs 10 years ago because that's what I do work on. And battery life and cool performance is unbelievable on MacBook Pro M1 Pro. Better performance per watt will come to the rest of computers and most people will be quite happy when it happens. Apple's done well enough to just be more upfront with this. Thankfully there are a lot of people like you to present more information.

[@ЙорданКирачен]

How can I dislike this video more than once?!

[@theTechNotice]

course you can, you can hit the dislike button twice :)

[@ЙорданКирачен]

@@theTechNotice the reason I saying this is because you was so “rude” in this video and it was like a rant. At some point it was annoying. But good job. Just don’t be like Karen.

[@maharaj8460]

@@ЙорданКирачен Seems like our apple fanboy got offended

[@ЙорданКирачен]

@@maharaj8460 I just don’t like his energy and way of speaking

[@finestPlugins]

The closest to an industry standard benchmark for GPUs that I'm aware of would be SPEC ViewPerf. Which doesn't run on MacOS and would also be mentioned explicitly.

[@theTechNotice]

🤔🤯🤯

[@Hjominbonrun]

I underclock my 6700XT and my 5800x specifically for cooling and power consumption cos my electricity bill is taking a jump in last few months. So I would buy the fastest GPU I can get, that doesnt exceed a certain power consumption. I have a desktop. So I am that guy buying for efficiency and not just to burn coal and increase temp of my room.

[@DerKek89]

i think the biggest problem is lack of Apple silicon support as always. As far as i know Blender doesnt have an official stable version for apple silicon. So you are basically comparing a simulated software which cant utilize the cpu/gpu to its full potential (thats why it doesnt draw even close to 110 Watts). Its not to defend apple but this way of testing is maybe as dodgy as apples marketing.

[@Byronic19134]

I mean I don't own any Apple products at the moment but to be honest, the amount of processing they are able to achieve at such a low wattage is utterly insane to me, and as somebody who pays an electric bill every month I would like to add I am all for it.

[@Real_MisterSir]

Why not just get a 3090 and undervolt it to the same level? You'll get far better performance, it will cost you less, and you will get a much more capable platform should you ever wish to up the voltage again for certain performance-heavy tasks. The biggest problem with the Mac Studio and the M1 chips, is their cost/performance. It's attrocious. Their only upside is the size and its easy implementation into existing Apple ecosystems. That's literally all it has going for it compared to the main hardware competitors.

[@mohanddarraji8285]

@@Real_MisterSir undervolt it to the level of m1 ultra ? Are u serious ??

[@brahmasthaandipramana7591]

Man, what kind of professional would think about electric bill? M1 ultra is 5 times slower than 3090 and consume 5 times less power. So 3090 is having a good performance per watt, and the electric power consumption is actually the same between 3090 and M1 Ultra.

[@brahmasthaandipramana7591]

@@mohanddarraji8285 You can undervolt 3090 at least around 200 watts, and it still faster about 4 -4,5 times than M1 ultra. So if you think M1 ultra performance per watt is the best, think again.

[@markdalbey]

Apple is a 75% marketing company, a 25% engineering company, and a no percent manufacturing company. It must work for them because they had nearly 350 billion in sales and over 94 billion in net revenue in 2021. It is completely expected for a marketing man to state a number that makes his product look good. They have hung their hats on performance per watt. This may be a relevant figure for laptops, but we are talking about desktops here. The average price for electricity in the world is 14.1 cents per kwh. We are taking about 100-200.00 savings per year. When you are spending 4-5,000.00 for a computer the cost of electricity is not the relevant factor. I noticed that you have started using an ergonomic mouse. I edit volume photography and have not suffered from carpal tunnel yet. I would like to prevent it though. Is an ergonomic mouse worth buying, and which one should I get?

[@Real_MisterSir]

All you need to do to make a good product, is convince people that it is good. That's literally all that Apple is today. A company built around a hyperinflated brand image. They don't have to make good products, they just have to keep the hype rolling - even if it comes at the expense of lying about their products. Most of their customers would rather believe the pretty lie and live happily ignorant to the scam, rather than knowing the truth and living with the fact that they are being lied to and actually aren't as special as they think they are.

[@realramo]

Absolutely.

[@sulaymen]

I wonder what results you would get I you would limit the RTX to 80-100W. Great video BTW. Chanel is growing - Great Job!

[@brahmasthaandipramana7591]

You can undervolt RTX 3090 to 200 watt and still get the 90% of performance. It means that it still faster about 4,5 times than the M1 ultra. So what kind of performance per watt did apple mean?

[@honestcomputing]

I'm sorry but your testing is flawed man. You're comparing the entire Mac's draw from the wall which means everything in the computer (fans, motherboard, memory, cpu and power supply inefficiencies when drawing from the wall) That means if apple's gpu wants 25watts it will still pull more power from the wall to supply each component inside the pc with power. Let's say for example the pc needs 100 Watts then the psu depending on it's power output (efficiency curve of the load) and the level of it's efficiency (80 plus gold, platinum etc.) will draw more from the wall. Not writing this comment to defend apple or anything but you have to give it to them that even though the m1 is not as powerful it's much more efficient just because it's a custom design. Squezzing everything together (RAM, cpu, gpu) in a single die is actually very efficient because of the time they take to talk to each other. Whereas on a PC everything is modular but further away from each other which makes it require more power to communicate with each other. It's kinda impressive that nvidia is actually beating apple even though they are on a less efficient architecture (samsung 8nm) and having to talk to the cpu through Pcie (which again is not optimal considering that apple has all components very close to each other) That is the beauty of creating something that is fully custom and not meant for modularity like PC's and I'm ok with it. If you want to be locked in one ecosystem that's fine but you also get major benefits. This is not new companies have been doing this from the invention of the first transistor. Just take a look at the datacenter market. All in all good video but if you want to do it properly you should measure the PC from the wall as well. :)

[@Technocrat.]

Glad I cancelled my M1 Max 10 core 32 core order as well Apple Studio. I instead ordered a refurbished 2020 iMac 8 core 5700xt 16gb 1TB.

[@daveh6356]

"I've got nothing against this" - yeah right! Your own testing shows the M1 beating the 3090 on GFXBench 4K Aztec Ruins (offscreen) 260fps to 223fps. What about that is difficult to understand? The takeaway is; as software is optimised, it proves Apple claims. Apple's graph was about hardware performance not software optimisation so you're a .1 release away from being totally wrong. You can use macOS powermetrics or install asitop (command line) for more accurate component power draw reporting. The M1 is better performance/watt end of story, whether that's important to you or not (love the way you said it 'could' be for laptops😆.

[@uribak9144]

Typical iSheep

[@daveh6356]

@@uribak9144 for pointing out he's contradicted his own argument?

[@nicktayloriv310]

At 10:53 I agree...

[@felixkwankangwei3088]

measure with cococut~hahaha

[@theTechNotice]

🤣🤣

[@mumbakayoba3082]

lol bro apple probably hates you

[@theTechNotice]

😬🤣

[@8bitbender495]

i sold my M1 mac mini after 1 year i need more power

[@tommycarter723]

Great video 😃👍

[@GroundGame.]

Remember, when the BS is thick, the snake oil is slick. 🐍 💰🧐

[@hassanr4887]

Blender is not fully optimized for apple silicon and the metal support uses a translation layer that they gonna improve in the future. wow would be a good example

[@uribak9144]

Yes, my car has not been optimized for a highway ... LOL

[@supermegapowerg1188]

@@uribak9144 not the same lol not just your car anyone any creatures any object exist can travel on that road. Arm designed processor cannot properly process translate x86 app.if there is not translation layer it would not run at all.

[@arthurchristmas4112]

@@uribak9144 no, not being optimised is literally the equivelant of a lactose intolerant person drinking milk. x86 and ARM are completely different architectures, just like your car won't run on kerosene

[@uribak9144]

@@arthurchristmas4112 Blender written in open source Software suitable for all platforms. Apple, itself has invested a lot of money in the software, so, enough request with the excuses it also does not interest the customer who paid 5k to Mac Studio.

[@supermegapowerg1188]

@@uribak9144 That's for previous Intel mac's not for apple own slicion m series line up it working through Rosetta now

[@marcotngsn]

the relative performance metric is comparing its performance compared to their M1 (100%) HOWEVER if you look at the graph from M1 Max, relative performance is HIGHER than M1 Ultra. Whats going on apple?

[@theTechNotice]

🤯🤯

[@drli8765]

Great video to reveal the low class misleading ads from Apple, The smart one will choose the pc after seeing this review, The not so smart will remain with Apple because they don't understand what this review is all about, I own Nvidia, AMD, and Apple stocks. I own Nvidia and AMD stocks because they are great companies with great innovation for the future, I own Apple stocks because a great majority of people are gullible,

[@Gray-Today]

Apple makes its money on imagined exclusivity. They always have.

[@rmanuelb]

They are not saying that the M1 Ultra can be compared with the 3090. My understanding is: in a “controlled environment” - aka lab, where we pushed the M1 Ultra GPU (giving it 120w), it reached the same performance of the RTX3090 with 320w. We will not be able to do this. They don’t sell it with those specs, so this is just marketing. They could have avoid it. But it what it is. Don’t like it? Ignore it. Don’t buy it. However, that much power, with low consumption and low noise, I believe it’s awesome. My RTX2070 Strix blows at my ears when it is pushed…. If they haven’t put it that way, there would no videos in the web…. and there are a lot!!! How many have you made around Mac Studio? 😂

[@mtmmf7632]

Comparing the M1 Ultra to a PC with a RTX3090 was exactly what Apple were doing when they said…. M1 Ultra offered more performance than an RTX 3090, while using 200w less power at its peak, M1 Ultra delivers faster performance than the highest-end GPU available while using 200 watts less power. Sure… Tesla can get 2000 miles in a lab with a 200kwh battery but I guess that would be misleading advertising that… 🤷‍♂️

[@rmanuelb]

@@mtmmf7632 Opinions vary. I was just stating mine.

[@djthrottleboivoltagecontro2145]

Put a fullblock on your strix. It'll shut up

[@costaliberta5969]

in apple apple users trust

[@JayzBeerz]

Adorama is so over priced.

[@evelynfive5684]

Don't get me wrong, apple product are good and I like some of their products, but I do agree that they always use these kind of grey words marketing which we need to focus with Lawyer's accuracy to understand what they mean on their boards. But sadly many people always fall for it heh heh.

[@evelynfive5684]

??

[@vojtechrichter8436]

Why wouldn't you just disconnect these fans tho?

[@russquest8161]

What about the DRAM scam?

[@graphiydesign]

The industry benchmark from apple… is not our standard benchmark… The gpu in M1 Ultra is locked from apple… Thanks for testing again to find out the true, so we can choose to buy it or not…. Great video

[@DayLateGamerWill]

Good ole crApple

[@HDRPC]

Sir i am waiting for rtx 4090 vs 3090 in your channel this year 2022.

[@Technocrat.]

Apple probably restricted power usage on retail units. ON their test units, they most likely removed power consumption limits, so the units were probably burned out after all the testing. If theres a way to change available power usage lets say to 110W

[@theTechNotice]

If that's true that's really bad, I really like to believe that it's not true!

[@mtmmf7632]

I think they will use the same ultra design, perhaps add a extra one or two processors in an actively cooled chassis, up the wattage, then call it a Mac Pro…

[@Real_MisterSir]

110 watt probably cooks them just to the brink of throttling - in a tiny case with no airflow they know this can't work for consistent-use consumer variants. But for their internal testing benchmarks I don't see why they wouldn't just run the chips at 11 / 10 even if the chip would fry itself after a month of such use or less. Aint their problem if the chips don't see any further use after the testings. It's like a dragster having to be rebuilt after each race session.. Easiest way to avoid this would be to just lower the real life tdp and not tell anyone about it.

[@1ahmadkalim]

hmm seems like marketing strat, just increase the power consumption limit and call it m2 ultra - build the hype and the apple fanboys upgrades -stonks!

[@melvinastrahan]

the RTX card alone is about the same size as the entire Mac Studio - its the form factor that matters

[@Real_MisterSir]

Then why does it cost more for the Mac Studio then?.. The form factor means literal ass to actual professionals. The Studio seems to just be for youtubers with money burning a hole in their pocket - or for people who are already deep in the Apple ecosystem (or someone who exclusively edit videos. Again, KZbin crowd comes to mind). With the Studio you pay more than a 3090 rig costs. You get far worse performance, and you even get worse performance/watt in gpu intensive tasks - which is usually one of the main reasons to go for such a tier of computer hardware in the first place. It's classic Apple. Overcharge for something that looks neat, and can be marketed as "better than competition" by scamming the living shit out of anyone gullible enough to believe the Apple marketing team.

[@silklion7375]

Give your videos 5-10% less saturation on the skin. Just an idea. Would be more pleasing I think. Otherwise I like it

[@theTechNotice]

yeah, there's a slight export issue with a batch of videos, I'll have to check that out... :)

[@TheAndymuns]

As of jan 2024 new computers will limited to 300w total power draw in the EU

[@JonnyPink65]

I really enjoy watching your videos. I have a new Dell 5820 workstation that is about the slowest piece of garbage i have used over the years. I "upgraded' to keep up with my editing of larger files, but this is a joke even with 128 GB or ram and and RTX 4000 and..... I am going to switch to the Mac Studio and that is how I found you. I NEED something that will handle 5, 6, & 8K video and this thing just shuts down. It will be a Goodwill donation when I do get my new set up. I dont expect it to be perfect, but I do expect it to simply..... work. I dont work on computers or building them, I use them. Past changing RAM, I know nothing... which is why I go duped on this Del garbage.

[@ashishpatel350]

apple always lies. it's just no one calls them out on it.

[@uribak9144]

Exactly, they did it in the era of PPC processors by claiming that G4 processors are faster than the Pentium at less clock speed. The G4 did not even reach the 1GHZ barrier while the Pentium 4 crossed the 3GHZ mark.