14" M3 Max 14 Cpu 30 Gpu Throttle Test
HTML-код
- Опубликовано: 16 ноя 2023
- Seeing that the 14" M3 Max 16 core & 40 Gpu core are throttling under full load, I ran benchmark tests on the binned 14" M3 Max 14 core & 30 Gpu core version.
Please view my updated video for better tests • 14" MBP M3 Max 14 cpu ...
This is the type of video everyone wanted.... not the shitty videos where people are shouting and using weird thumbnails
LOL. The enthusiasm trend is tiring.
Man, I hate those thumbnails with the cringe face and all the arrows pointing to no where!
Thank you got the video. I've been looking to upgrade my 2017 MacBook Pro 15inch with a M3 Max 14 inch base model and was seeing lots of clickbait videos but I realised it was testing with the 40 core. Cannot wait to get my M3 Max now!
good video! thanks for this!
Not a fancy video, but it really gets to the point. Great work! It seems the cooling system worked really well on the 14 inch max, you barely heard it when it was not in heavy load.
just wanted to whip something up to do a personal test but noticed RUclips was full of comments asking about this model. This 14" is silent unless you really start to push it. Seems it stays on the quieter side if you load up the GPU. I'll do some more videos that arent simply screen recording.
@@LucasCreativeTech Thank you!!! This is all i wanted to know. So many people are boasting about the high fan speeds under load and the thermal throttling, but no one ever talk about every day performance.
@@CloroxBleachOfficial watch my newest video also. No fans for everyday use. Only fans under heavy load. I just ordered the Base 11 core to test it against this
Do you hear the fans whike gaming? E.g lies of pi or baldurs gate 3. I also prefer the 14 inch form factor but still want some decent performance when needed.
Thank you for this! I am hoping you don't mind offering advice- I want to upgrade to the M3 Pro or Max 14. I do light video editing (Final Cut) but I would really like to be able to play FFXIV (casually) on it as well. Of course the Max is better, but do I really need it? Thank you!
By tonight I will post a video comparing realistic video editing on this M3 Max vs an M1 Pro so you can see if you really need the Max
What about fan noise? Do the fans kick in often during normal operation? - how loud are the fans if you do something like a Final Cut export? - I really love how quiet my current 2020 MacBook Pro 13” M1 is.
I will do another video putting Final Cut to a hard core test. I was exporting this video as ProRes with a different resolution and did not hear the fans.
Do you want a new laptop or is your m1 hanging and getting delays? Unless you export videos multiple times in the day I am not sure a couple minutes makes much of a difference for the fan noise in the m3s.
@@timbukstu9861 the M1 struggles with 6k and pretty slow with editing 4k videos. It's also pretty slow in CPU performance, running Rails specs takes 15 minutes when on the M3 Max (14/30) it should take around 4 minutes which would make a huge difference for faster iterations. I have 16GB of ram on the M1 which often runs low. Not to mention the screen is a lot better on the 14", etc, etc.
FCP exports do eventually have fans but it's not bothersome. Check out my video showing this ruclips.net/video/jfWsIqxQNK8/видео.html
Cool video, looking forward to that real world FCP video you said you're making :) I have a chance to get the 14" m3 max for the same price as a 14" m3 pro 36gb 1TB so same specs, but I'm wondering if the fact that it will be basically overkill for me at this moment will make it deplete in battery fast, be noisy etc. Would love to hear your thoughts about it, I will be doing some video editing and coding but nothing intense
stay tuned. Definitely need to cover the battery drain test
Sounds like the M2 pro is your best machine for your Use. The max is overkill given your use case so I would get a pro and the M2 Pro is a better choice more than likely. You can always sell it later and get an M5 max. If you’re not doing hard-core video extracts all day long and working with 8k for multiple hours a day, I don’t think it’s worth it.
We use Final Cut Pro on an m1 air for occasional videos and run a small graphics design business (photoshop and Indesign all day) using m1 airs. I am upgrading my lead designer from a 2016 i7 mbp to a M1 Pro next month.
So the m1 8gb/512 air is excellent for our use case. For software development I have a 24gb/1tb m2 air which is more than I need as a dev. Alex ziskind has a review on the m2 air for developers. m.ruclips.net/video/A1IOZUCTOkM/видео.html
@@timbukstu9861 I traded in my old Intel and was going to order the 14” M3 Pro 36RAM 1TB 12/18 CPU/GPU for $2550 as a student. So I was between that and refurbished M2 pro for $2250. But my uni is offering a Black Friday deal for the base M3 Max for $2600 and I rather just future proof the hell out of my work for that little. They have a M3 Pro 12/18 1TB but only 18gb for 2000 so I honestly may do that. I for sure need at least 16-18 RAM for my workflow. I’m using my girlfriends M2 air 8gb and it’s not enough, keeps crashing during premiere renders. An Air with 24gb and 1 TB 15” is basically 2k, and at that point the difference between everything is 500-600 so not a deal breaker for me. What do you think?
I know I can get an M1 but I rather fork a little more for better performance and a bit more longevity. I’m hoping M3 becomes even more optimized on the software end, increasing that performance gap.
So would you say it would be worth me returning my 14inch for 16inch for more performance. I do music production and programming, some light gaming
What's with the core temp readings bouncing down to 2 degrees like that?
Thanks for the test and the information. The question is how do the m3 max in an 14 pro behave in real world tasks like Video editing especially in Final cut pro such as rendering and exporting in h.264 or HEVC. Will the fans kick in and how loud. Several tests on several projects in fcp short and long videos an so on. Thanks in advance.👍
im was wondering myself so im working on a video where ill use the S5iiX h.265 6k footage in a realistic editing situation on multiple timelines. Play test and export test. Fan speeds and testing how much of the GPU is being used. Stay tuned. Will be out before Monday. Will also test it against the MBP M1 Pro to see if its worth upgrading.
Still working on the video but can say that what you are asking about runs smooth as butter. No Fans. Silent editing. Exporting will have fans.
@@LucasCreativeTech
Good news in general. Still waiting for the video. I appreciate your efforts. Thanks a lot.
What is the name of the app where do you check system info ? Thank you.
MX Power Gadget
Can you do the same test with low power mode ? Will this reduce the fan noise but still give better performance than top m3 pro?
Curious about how the Max is in low power mode too
I don’t pretend to know much about computers, but it appears to me the fans are doing what they are supposed to do and they are not nearly as loud as the fans on my 2021 Intel iMac. They don’t sound that loud to me on the video. I think what noise they do make is the trade off for a 14” computer that can run heavy workloads smoothly. You also have the automatic performance mode option which I assume will produce performance and noise level no worse than the non-max pro chip in the 14”.
was this done on High Power mode or Balanced?
Hi Lucas, thanks for the tests. On the RUclips is no videos with tests m3 Max base model. I would like to upgrade my m1 machine with MBP 14 M3 Max base model.
Could you please run the NeatBench 5 benchmark and share the results here or in new video? It will show how the video denoiser performs on your machine.
You can find the benchmark no the neatvideo website in the download page. Thanks
If it’s possible run the benchmark on M1 Pro MacBook to compare the performance.
Thanks!
@@Stockvideomaster yes I have a 16" MBP M1 Pro I will test.
@@LucasCreativeTech Might be useful to someone: MacBook Air M1, NeatBench 5, 13.7 frames/sec with Frame size 1980x1080.
Waiting of tests on MBP M1 Pro and M3 Max Base model.
Thanks.
Lovely videos ,man!
You rock!
I would use it for music production.
Could you do tests or let me know how's the battery?
Thanks!
I have GPU 12450 on same binned m3 max 14'. Also same cinebench 2024, power mode, without charging. I wonder why there is such a difference with your 7693 test? My CPU multicore I have 1448. I make 10 minutes tests
I used this video to make my purchase decision after being on the fence about sticking with an M3 Pro or getting the base Max. I am now returning my M3 Pro and getting the base Max. The Pro is great, but just a tad less than I want since I need some additional GPU performance especially.
thats awesome. Im currently testing this Mac in Final Cut Pro. So far it can handle anything you throw at it until you insert tracking motion FX plugins or denoising. Then I have to switch it from best quality to best performance to get smooth playback. This is of course with 6k h.265 footage and grading. But it is eating the GPU to 100% in this case. I'm still uncertain about battery life with the Max chip.
uncertain meaning battery life is poor?
@@bladerealm124 I will release two videos shortly. One will show battery life while working with FCPX and the other will show general use battery life. I will compare these to an M1 Pro. For now I can tell you that pushing this 14" to its limits at high power mode while editing in FCPX will drain 20% in 37 minutes.
@@LucasCreativeTech thanks for the info
@@LucasCreativeTechthat's the reason I'm upgrading to m3 max. For motion vfx tracking which is very slow on my m1 pro 16 gb ram. Is it worth upgrading as I'm using 5k raw 12 bit apple pro res files.
Hey guys, I have an updated video up ruclips.net/video/jfWsIqxQNK8/видео.html
Throttling will not affect how much % of a cpu is used, it affects the frequency at which it is running. In order to test throttling, we would need to be see the frequencies of cores, and show if they are downclocking at the t-junction (max temp of 101/102) after a given period of time.
see my new video. I redid the test showing frequency.
Awesome, Can't wait to check it out!@@LucasCreativeTech
OK guess this doesn't tell the story sigh. Will wait for a re-do with all the dynamics as well as explanations for the vast variance in performance.
For comparison:
Cinebench CPU (14“)
M3 Max (binned): 1350
M3 Max (unbinned): 1510
Cinebench GPU (14“)
M3 Max (binned): 7693
M3 Max (unbinned): 13032
So there is a huge difference in GPU performance here
Those GPU scores are strangely low. The binned 14 core M3 Pro scored around 5500 in Max Tech's video.
ruclips.net/video/aQvsZQ3QBiU/видео.htmlsi=pbWq1GjHFP2B8Bfu&t=920
I hink something is wrong here can’t be such a big difference. Same cinebench version used for comparison?
@@alexpixeli well the main difference between the two M3 Max Chips are the GPU cores.
M3 Max (binned): 14 CPU Cores, 30 GPU Cores
M3 Max (unbinned): 16 CPU Cores (+14%), 40 GPU Cores (+33%)
CPU Scores: 1350 vs 1510 (+11%)
GPU Scores: 7693 vs 13032 (+70%)
So yes, the difference in GPU performance seems to be unreasonably high. But you can see the results of the M3 Max (unbinned) on the 14“ in this video: ruclips.net/video/E9IJ7nOAZyM/видео.htmlsi=Wrxmj0or1jcXR85M (at 8:50)
Same Cinebench version as far as I can tell. Maybe a different kind of GPU test?
@@FliptricksftwdudeCheck your math again. You are wrong. The 30 GPU is 75% of 40. The that would mean if the 40 GPU scored 13,000 then the 30 GPU should be 9,750. Maybe the 7,200 was a yes ran on low power mode?
@@AK-qh4yymine just did 10400 same configuration. Something wrong with how he ran it.
Strange, I have the exact same configuration and my GPU score was significantly higher over 10300. CPU score pretty much identical.
Did yours throttle? I am trying to figure out if there was throttling or not so I can either keep this one or go with a 12/18 pro instead :)
@@minstrelcaslightly but nothing that will matter to anyone in real world. Even with heavy programming, photo editing, or video rendering. I think this is the sweet spot for pricing. I didn’t buy this for gaming, so I couldn’t care less about the extra GPU if I want to game I’ll use my ps5.
@@BIGWEAV Thanks for your insights.
Hi Lucas, thanks but i don't understand ... for me throtteling is the computer decreasing the frequency/clock of its CPUs to cool down itself. Thus, bench scores could be not as good as the 16" model.
I don't see frequency check in your video nor bench comparison.
As for me : 14" M3 max 16c 64Go 1T
In my test, i recompiled Webkit : 1st pass = 10.29 sec, 2nd pass = 10.03 seconds. Other youtubers on this test :
Dave 2D 16” M3 Max (…ed out) 128 Gb ram : Webkit : 11,18 min ( ruclips.net/video/qie7cfjnLAY/видео.htmlsi=s2ly27UXa6eD07nt&t=173 )
Max tech on the 16 too = Webkit 10.13 min.
On Xcode Bench :
M3 Max 16" Xcode bench = 79 seconds in this Max Tech review : ruclips.net/video/BWeuhxnWDl0/видео.htmlsi=qBqpcNJKIH80P9ft&t=393
M3 Max 14" Xcode bench = 71 seconds in this Luke Miani review : ruclips.net/video/BWeuhxnWDl0/видео.htmlsi=qBqpcNJKIH80P9ft&t=393
73.9 seconds for me ( while working ! with loads of things in the background )
Dude has no clue about throttle, no power/clock visible so useless
My god the fans are loud. I am torn between top m3 pro 36gb 2tb or base m3 max 96gb 2tb for the 14 inch. I definitely want the 14 inch and not the 16. Any suggestions guys? I dont want my device to become hot, throttle or have loud fans.
it didnt feel hot during testing. The fans kicked on due to the CPU being pushed to its limits. I will do more real world tests
@@LucasCreativeTech thank you. So I am guessing the m3 pro variant won’t kick the fans into overload mode, if only Someone with that configuration can give confirmation.
If 36GB RAM were enough with the M3 Pro, why would you opt for 96 GB with the MAX?
@@bmakszim if I get the m3 max then a whole new range of possibilities would be unlocked. I can then do more with the machine. For that I would require RAM more than 36gb. Unfortunately, Apple does not give 48 & 64 memory options with base m3 max, which is bargain. So, I would be forced to go with 96gb ram. Since those extra gpu cores use shared memory, 96gb will be worth it.
If you buy a high performance small laptop, under load it will throttle and it will get hot and the fans will be audible.
There is just no way around it, there isn’t a small powerful laptop in existence that doesn’t behave this way haha.
You can try cooling pads, propping it up etc you can do things to mitigate it but a top spec small laptop will most definitely be affected by all of those things.
looking at utilisation doesn’t tell anything about throttle. It already decreases clocks in the first 30 seconds and probably never gets them back up like a 16“ this review is basically worthless
PS: also may reduce power before clocks but neither is visible
@@blackandcold thanks for the tip. Is there software you'd recommend to monitor clock speed of CPU?
So it is or is t throttling ?
it is on long running tasks@@minstrelca
@@LucasCreativeTech btop or btop++ also get asitop since asitop also shows a throttle flag and power usage. when power dorps because temps are too high - it throttles :)