MacBook 16-inch Fan Noise

mamahtehok wrote:

it doesn't matter how many customers are happy with this machine, with this price and product positioning ALL customers must be happy.

That will never, ever be the case, no matter what the product.

No product is perfect for everyone, especially one with the limitations inherent to any laptop.

r-mon wrote:

Sorry William but this laptop has been super disappointing from a UX/UI Designer stand point as well.

That's too bad, but my machine runs great with dual 4k screens. I run Affinity Designer/Publisher/Photo, ADobe Indesign, and much more all day long. Many times it's running FCPx too at the same time. This is a great machine. I would hope you can get that under control with what you have loaded, because something is causing your GPU to be way more than it should.

Look, this also happens on battery.

I have a brand new base model 16" with the 5300M, and if I go into a meeting room on battery power and add an external display via any wired method (USB-C to USB-C, DisplayPort adapters etc, multiple combos tested) and the power consumption of an *IDLE* machine goes to 40-50 watts. The battery is only 100Wh --> total battery life on idle is less than 2 hours. If I start to use any actual app and actually do anything with the laptop, the battery life is around 1 hour - for example a Teams call without video raises the average power consumption closer to 80 watts with an external display attached.

Closing the clamshell with the external display still connected immediately drops the power consumption back to the 10-15 watts enabling the promised approx. 10 hrs of battery life, even when presenting something trivial such as a PowerPoint deck.

Avai1 wrote:

Pretty sure it's literally just the dgpu saying " I NEED 20 watts" without reason.

Once again, it’s not without reason.

The GPU needs to use that kind of power to run the GDDR6 VRAM at full speed to avoid flicker.

USB-c to DisplayPort does only help if the GPU does not start. Otherwise the defect of the power consumption due to too high vram refresh rate let the heat come up unnecessarily high. Resulting in fans spinning up. This leads then to a bad user experience.

Avai1 wrote:

Do you happen to have a 5600M?

I have the 5500 with 8GB of video memory and 64 GB of system ram. You can look through the thread as I was here from near the beginning. The screenshots show that it generally runs at 7-8 watts on dual 4k screens.

EngagingExposures wrote:

I have the 5500 with 8GB of video memory and 64 GB of system ram. You can look through the thread as I was here from near the beginning. The screenshots show that it generally runs at 7-8 watts on dual 4k screens.

Care to share your screenshots again? AFAIK, no one shared a screenshot of low dGPU usage using 2 external displays with a 5300M / 5500M.

I have a MBP 16" with 5500M Pro 8GB, 64GB RAM, 1TB SSD and i9. I had some issues at the start an went also to an Apple Store to let them identify the issue. No result at all and they were unable to understand the issue.

After searching and looking for solutions, the best I've found so far is a USB-C to Displayport cable to my LG 43" external display. Since then, most of the time, fans are running between 2300RPM to 2800RPM and are not noisy at all (without external display, it runs around 1700RPM). I'v split all connexions with external display connexion on the left and other connexions on the right (charger, converters, etc.). Previously with USB-C to HDMI or Displayport to Displayport thru an external box, I had issues with fans running high at 4500 to 5000RPM. Not sure if this can help but the USB-C to DP cable can be worth trying for others with the same issue.

Please also post resolution and monitor. If I use 1600x1200 on a 1920x1200 EIZO MBP uses 4..5 Watt. With 1920x1200 consumption goes up to 19 Watt. Regardless which port on MBP and display. To use 1920x1200 with two displays I need an external USB to double HDMI adapter to avoid the fan noise. I could not have confcalls with the loud fan noise when the monitors are attached direct due to the GPU driver defect

Cool, so what monitors are you using? Exact resolution + refresh rate + connection type will be nice to document how to get the "low power dGPU mode". And I assume this is for clamshell only right?

Thanks!

But why, after paying nearly 4k on this MacBook Pro, is it that my mom's MacBook Air can run my 2 monitors (1080p 24 inch, 1440p 27 inch) with it's lid open and my 16 inch MacBook Pro can't run a single monitor with the lid closed without it reaching 90c? I bought this instead of a Mac and a small MacBook for it to be the all in one, full sized desktop and laptop, final solution, but instead, at this rate, it may die in a year or two.

First, the “Turbo Boost” on Intel CPUs is designed to “overclock” the CPU until it has to ramp down the clock speed due to heat.

Second, because the GPU in the MBP 16 is far more powerful than the one in older MBPs. I’ve used this analogy before but it’s like buying a full-sized pickup truck and complaining that your old compact pickup got better mileage with its V6 compared to your new V8 that weighs 800 lbs. more.

Third, operating at internal temperatures of 90°C or higher is completely normal for modern laptops.

Finally, concerns about longevity due to heat are pure speculation on your part; if Apple felt they’d self-destruct they wouldn’t allow you to purchase AppleCare coverage for them, they’d lose money.

itunestux wrote:

USB-c to DisplayPort does only help if the GPU does not start. Otherwise the defect of the power consumption due to too high vram refresh rate let the heat come up unnecessarily high.

That is completely incorrect.

1) Any time an external monitor is connected, the dGPU is used. The integrated GPU has no physical connections to the USB-C ports.

2) It’s not “unnecessarily” high according to AMD, and once again mine can run at 21W of power draw all day and the fans never go over 2500 RPM. People asked for data, I posted it and yet this comment keeps being made.

3) It’s not a defect.

Maybe someone in engineering has decided to tell the world why the GPU uses so much power when it is connected to an external display with the lid open.

The VRAM mapping table hypothesis is worth pursuing, is it not?

@Grant Benett-Alder, does that make sense to you?

dcristof wrote:

Maybe someone in engineering has decided to tell the world why the GPU uses so much power when it is connected to an external display with the lid open.

It's been explained; under some, perhaps even many circumstances the VRAM needs to be driven at its full refresh rate to avoid potential flicker on external displays.

It's the power draw of the GDDR6 VRAM - which is known for being fast but for being a heavy power draw - that's largely to blame.

Again, this was AMD's comment on a slightly different GPU but it is likely applicable here (emboldening mine):

I checked with the product team and their feedback is that depending on specific display configurations (resolution and refresh rate combinations) and background tasks, RX 5000 Series GPUs may maintain memory frequency to ensure an optimal user experience. This behavior is expected and does not impact the RX 5000 GPU in any way.

https://community.amd.com/thread/251357#comment-2980117

Older MBPs with GPUs that use HBM VRAM may in fact be more power efficient than those using GDDR5 or GDDR6 VRAM, but lack the performance of GDDR6.

That's why new higher performance and more power efficient HBM2 VRAM was such a step forward with the Radeon Pro 5600M GPU - though it comes at a higher price.

What's kind of neat is this third-party article comparing GDDR6 and HBM2:

Techsiting: GDDR6 VS HBM2 Memory

It states that GDDR6 VRAM consumes 3.5x to 4.5x the power of HBM2 VRAM but is available at a lower price.

That's almost precisely what some have been reporting for power draw with the Radeon Pro 5600M.