How does one know when the chip is being throttled? Is fan spinning up an indicator?
Last edited:
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
First review about sound... fan noise
- Thread starter Cape Dave
- Start date
- Sort by reaction score
Not to be picky, but now it is not winter in switzerland... (it's fall)
Interesting, didn't know that. A MBP M1 Pro my first and only MacOS device, with fans being really OFF most of the time in my case, good to know, that things are different for e.g. a Mac Mini.
Isn’t there a learning period for battery usage, or am I mistaken (I think that used to be a thing).
if it’s truly not an anomaly and it’s draining faster for the same tasks, I do wonder why battery size isn’t paired to accommodate with each chip. It’s probably the middling / low processing intensive tasks that are the marketed usage times
if it’s truly not an anomaly and it’s draining faster for the same tasks, I do wonder why battery size isn’t paired to accommodate with each chip. It’s probably the middling / low processing intensive tasks that are the marketed usage times
that could be the case
the m3 was a refurb so I guess the battery had already been through some cycles
I've never owned a new MacBook so wasn't aware of that
Could somebody confirm what version of “USB 3” marketed by Apple for the two front USB C ports is in fact? Their specs say 10GBps, but it’s not clear if they’re collectively referring to the two ports (so USB 3.0 @5Gbps x 2 ports), or if it’s truly USB 3.1 /gen2 (10Gbps) per port. Their specs should probably have used the decimal-point attribution to avoid ambiguity.
EDIT: I messaged Apple. They confirmed it's USB 3.1 gen2 @ 10Gb/s PER port.
EDIT: I messaged Apple. They confirmed it's USB 3.1 gen2 @ 10Gb/s PER port.
Last edited:
With the greater dependence of P-cores instead of E-cores in the M4Pro (vs the M3Pro), this is not all that surprising. Although not especially relevant to the Mac mini (this thread is in the Mac Mini section) since there was never an M3 mini.
Well they are both 3nm process. The M4 can run much faster with higher speeds than the M3, but you're using more power to get there.
Perhaps not but it is the same chip in question in this thread
And it does have more p cores than the m2 pro
Did you check activity monitor to see what is running? I always get spotlight running like crazy in the background indexing everything for a day or two whenever I get a new mac.
spotlight got all it's indexing done over the weekend while it was plugged in
nothing was running other than what I was using. quickbooks, excel, mail, safari
Honestly? Because you are making many assumptions which are incorrect. For example, I suspect I first learned what latency was around 1981.
Even the base M4's CPU is *so much faster* than your Intel Mac you clearly haven't absorbed what that would mean for you. RAM issues aside, there is no job that can run on your Intel mini, even with difficulty, that could absorb more than a modest fraction of the attention of an M4, much less an M4 Pro.
I truly don't know if either M4 would be a good fit for you, compared to the M2 Studio. I don't know if the load you'd put on your machine would cause the fans to be too loud on the M4 Pro mini. But I do know that you're making a decision, from what you've previously written, based on false "facts", misunderstandings, and/or bad assumptions. Your privilege I guess.
Thank you for posting useful facts instead of misinformed speculation. There's been far too little of the former and much too much of the latter here so far.
If your constant load number (24W) is correct, then I think your conclusion must be correct as well. However, I would want to be very very sure that that number is truly correct for the M4.
If you're willing, I think it would be extremely interesting to see what the performance ceiling is for the M4 Pro, while maintaining fan silence. You can get the fan control software that @nathansz referred to, set it up so that the fan never spins up past the audible threshold, and see how fast the M4 completes the handbrake run, compared to the M2 Mini (and maybe also the M4 Pro without fan limits). That would answer a number of open questions.
What are you using to measure this? It was my understanding that the M4 consumes significantly more power. Does this measurement stay roughly constant over time?
Honestly, I'm extremely impressed it takes 10-15 minutes to clock down! I would have expected that to happen much quicker, given the small heat sink.
If your constant load number (24W) is correct, then I think your conclusion must be correct as well. However, I would want to be very very sure that that number is truly correct for the M4.
Unless I misunderstand what you're saying, I would expect that position to be less optimal than having it standing in its normal orientation, though I would also be surprised if it made much difference. (The part of the air intake/exhaust ring nearest the surface would experience some turbulence that it would not in its normal position.) Still it would be interesting to see if there's a difference in behavior when it's in a normal orientation.
If you're willing, I think it would be extremely interesting to see what the performance ceiling is for the M4 Pro, while maintaining fan silence. You can get the fan control software that @nathansz referred to, set it up so that the fan never spins up past the audible threshold, and see how fast the M4 completes the handbrake run, compared to the M2 Mini (and maybe also the M4 Pro without fan limits). That would answer a number of open questions.
No. The fan spinning up indicates that there's more heat being generated. If the fan can't remove enough heat, then the chip will be "throttled".
It's important to understand that throttling is not generally an indication that the chip is clocked too high, the cooling system is too weak, the OS is misconfigured, etc. It *can* indicate that, but much more commonly it simply means that the manufacturer of the system chose that level of sustained performance. There are lots of reasons why you might, for example, want a laptop that does *not* permit unlimited use of its CPUs at their maximum rated speed. Those include form factor/size, weight, battery, and sound (fan noise) considerations.
Here's something I wrote last spring that will help you understand all of this a lot better.
It certainly can. However, everything I've read so far claims that battery life is even or better, so that seems pretty odd. Maybe check it again in a few days? Or poke around with powermetrics (or just top, or even Activity Monitor) to see if you can spot something sucking up juice.
True, but you are implying things that aren't true, because they are not the same "3nm" processes. The M3 is on N3B, and the M4 is on N3E. Both "3nm class", but quite different. Notably, N3B is a little more dense, while N3E is a little more efficient and performant. But the differences are minor - a few percent each way. N3E is cheaper, too, and NOT in a minor way- it uses only about 2/3 the number of layers, and does no EUV double-patterning.
As it turns out, the M4 is more efficient than the M3 iso-clock, but Apple has chosen to be a little more profligate with energy with the M4 than with the M3, so they allow it to clock higher. At those higher clocks, the M4 is less efficient than the M3 at it's maximum (lower than M4) clocks.
So it seems the m2 mini peaked at 1700rpm while the m4 mini peaks at 2100 rpm per the handbrake test, with the latter being audible (while the former not). All I’ve heard is that folks can’t hear the m4 mini (non-pro) under any conditions, so I’d be interested to know what the noise curve / dB looks like for both under max load. I also wonder if being on its side made it audible (more fan rattle or resonance)
Can confirm at 104C maxed out the fans are pretty loud. I sustained the heat load for about 5 minutes and
CPU was still at a little above 25% usage. Pushed it with A Simulator game. So Graphics got it heated up really good.
Took maybe 15 minutes or so to get there. And it cooled to less than 60C in about 2 minutes so it cools down quickly.
M4 Pro 14 core. So Cpu there is plenty of power left just in my test graphics hit it hard.
CPU was still at a little above 25% usage. Pushed it with A Simulator game. So Graphics got it heated up really good.
Took maybe 15 minutes or so to get there. And it cooled to less than 60C in about 2 minutes so it cools down quickly.
M4 Pro 14 core. So Cpu there is plenty of power left just in my test graphics hit it hard.
Normal Range: Around 30°C to 80°C (86°F to 176°F) during regular use, such as browsing or light tasks.
• High Performance Tasks: Can reach up to 85°C to 100°C (185°F to 212°F) during intensive workloads like video rendering or gaming.
Maximum Safe Temperature
• Apple has not published specific thermal limits for M-series processors. However, modern CPUs, including those in the M-series, typically have thermal thresholds around 100°C (212°F) before thermal throttling kicks in to prevent overheating.
• Prolonged exposure to temperatures near the maximum limit may reduce long-term durability.
• High Performance Tasks: Can reach up to 85°C to 100°C (185°F to 212°F) during intensive workloads like video rendering or gaming.
Maximum Safe Temperature
• Apple has not published specific thermal limits for M-series processors. However, modern CPUs, including those in the M-series, typically have thermal thresholds around 100°C (212°F) before thermal throttling kicks in to prevent overheating.
• Prolonged exposure to temperatures near the maximum limit may reduce long-term durability.
Dude, you're babbling about stuff you clearly know very little about.
There is a difference in CPU demands between running a large DAW project at 512 I/O buffer size vs say 128, which affects latency. Latency is important when playing parts in realtime via keyboards. Then there's running at a higher sample rate with the same number of plugins and software instruments, than what's possible on the i7 Mac Mini etc.
Also, buying an expensive Mac Studio at full price, that's two generations behind the latest Apple silicon, is not something I would do - unless my old computer broke, and I needed a new one ASAP.
Then there's future-proofing, getting a M4 Pro (or more) for sufficient RAM running large sample libraries..etc..etc..
Please keep going if you want, but I'm done replying to this nonsense.
Last edited:
I can't hear you over the coil whine of my PC.As I threw benchmarks, games and encoding jobs at the Mac mini, I was constantly impressed by how much it could tackle without making any discernible fan noise. But Apple can’t escape heat. The Mac mini has a large intake fan along its bottom that sucks in cool air and spreads it out throughout the entire system. However, when I started running Cinebench's multithreaded benchmark, the fan kicked into high gear and the Mac mini sounded like it was getting ready to take off.
It's not an unpleasant sound — it's more like a calming white noise machine than the obnoxiously loud Mac fans of yore — but it's certainly noticeable. If you're typically wearing headphones or blasting music, it may not be an issue, but it could make the Mac mini very annoying in a shared office setting. If you're planning to constantly throw heavy workloads at it in high power mode, you might be better off with the larger Mac Studio, which can better handle heat.
Apple Mac mini review (M4 Pro, 2024): Shockingly small, incredibly powerful
"That's a computer?!" My daughter just couldn't believe her eyes when I pulled the new Mac mini out of its box. It measures just five-inches by five-inches across, and it sits a mere two inches tall. The Mac mini is an adorable box that resembles an oversized Apple TV. But what's more impressive...www.msn.com
I have a mini PC that also has an obnoxious fan on it that fires up every 2 minutes because Windows 11 is running the stupid malware check that consumes so much CPU power for the 3 seconds it runs, it causes it to warm up, triggering the fan (and worst of all, it is set to low power mode).
I'm also a bit deaf from all the R'n'R in my youth.