Yes, show some stats and pictures.
what kind of graphic card do you have in the enclosure?
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.
(2019) 16" is HOT & NOISY with an external monitor! :(
- Thread starter Appledoesnotlisten
- Start date
- Sort by reaction score
what kind of graphic card do you have in the enclosure?
Yea that's not good. How much power is going to the graphics card? You might try Volta (or voltage shift) to limit the wattage that the processor uses. You can set it pretty high but it might help to avoid needless power usage that raises the temp. If you have istat, post a screenshot of the drop down menu when you have it set up and running as in the video.
I've been testing out an eGPU configuration as well (Razer Core X + 5500XT). It works great, in my daily usage fan speed never gets higher than 3000 rpm now and usually averages around 2000-2500. Before, the fan was completely maxed out at 5000+rpm.
If you are okay with spending the extra and have space on your desk for it I don't see much drawbacks to the eGPU solution. I would get it over a 5600M personally - that GPU is an improvement over the 5300/5500M but it still uses decent amount of wattage so will heat up the internals more compared to going the eGPU route.
Monitor is a 38GL950G running @ 85hz
With lid closed:
Lid opened:
If you are okay with spending the extra and have space on your desk for it I don't see much drawbacks to the eGPU solution. I would get it over a 5600M personally - that GPU is an improvement over the 5300/5500M but it still uses decent amount of wattage so will heat up the internals more compared to going the eGPU route.
Monitor is a 38GL950G running @ 85hz
With lid closed:
Lid opened:
I ended up returning the thing. Will opt for the lower 2.3Ghz CPU and the 5600m Radeon see if it handles my external monitors better. If not, I'll return it again and wait for the next ARM models.
I've posted performance stats from my Razer Core X Chroma + RX 5500 XT with both 2019 16" i9 2.4GHz 5300m and 2019 15" i9 2.3GHz 560x setups. But here are some more stats and commentary.
The first three images show CPU utilisation, frequency, and left side fan speed under real-world everyday usage.
First (on the left) is me compiling 24 builds of a game (4 variations, 6 platforms) in Unity. This does not hit all CPU cores, but it does sustain close to 4GHz with minimal fan noise (momentary peaks at or just under 3000 rpm).
Then there's a period where the system was idle (I went away and came back) before (on the right) you can see me compiling 4 builds in Xcode. This DOES hit all CPU cores, with a pause in between each build step. This still sustains an amazing level of CPU performance at around 3.75GHz, with very high fan speed (but not pegged at max).
The next three images show performance under sustained synthetic heavy load (Cinebench R20 on loop). In this extreme test, the CPU can sustain 3.3GHz indefinitely with the fans pegged at max. It's also important to note that I can concurrently run the Luxmark GPU stress test on loop at this point with absolutely ZERO impact to CPU performance.
That's a key benefit of an eGPU over dGPU (even the 5600m). GPU heat and power consumption are OUTSIDE the machine's thermal envelope.
So in summary, IF you can live with having to click a "disconnect" before unplugging and having a few apps quit (only the apps that are using eGPU), an eGPU is an EXCELLENT way to workaround any heat/noise issues AND get the absolute best CPU performance possible from the machine.
It's cheaper than a 5600m upgrade (but still expensive), and can be added on after the fact, and upgraded separately.
The Razer Core X Chroma also gives you a "one cable" docking solution. I've heard reports that the ethernet is flaky, but I don't use that and I haven't had any issues connecting SSDs to the eGPU's USB ports for backups.
It also provides a full 100W of power to the machine which is important to prevent battery drain/degradation. Most monitors and TB3 docks only provide 87W or even 60W. The 16" MBP comes with a 94W power supply, but the next image shows the total system can use >=110W in bursts.
I don't buy into the argument that disabling turbo boost or setting a low total power limit with Volta and Voltage Shift has little or no impact on performance.
Obviously you won't notice it in all but the most extreme cases of slowdown, because you're too busy actually working. It's generally hard to notice the absence of something, anyway.
But from my own personal experience, I often DO notice that the performance IS available and is being used. I am frequently impressed that things just happen quickly, all the time. Not only massive compile jobs. Everything. Even everyday single core medium load tasks are happening at close to 4GHz. That's a 66% increase over the base frequency.
The first three images show CPU utilisation, frequency, and left side fan speed under real-world everyday usage.
First (on the left) is me compiling 24 builds of a game (4 variations, 6 platforms) in Unity. This does not hit all CPU cores, but it does sustain close to 4GHz with minimal fan noise (momentary peaks at or just under 3000 rpm).
Then there's a period where the system was idle (I went away and came back) before (on the right) you can see me compiling 4 builds in Xcode. This DOES hit all CPU cores, with a pause in between each build step. This still sustains an amazing level of CPU performance at around 3.75GHz, with very high fan speed (but not pegged at max).
The next three images show performance under sustained synthetic heavy load (Cinebench R20 on loop). In this extreme test, the CPU can sustain 3.3GHz indefinitely with the fans pegged at max. It's also important to note that I can concurrently run the Luxmark GPU stress test on loop at this point with absolutely ZERO impact to CPU performance.
That's a key benefit of an eGPU over dGPU (even the 5600m). GPU heat and power consumption are OUTSIDE the machine's thermal envelope.
So in summary, IF you can live with having to click a "disconnect" before unplugging and having a few apps quit (only the apps that are using eGPU), an eGPU is an EXCELLENT way to workaround any heat/noise issues AND get the absolute best CPU performance possible from the machine.
It's cheaper than a 5600m upgrade (but still expensive), and can be added on after the fact, and upgraded separately.
The Razer Core X Chroma also gives you a "one cable" docking solution. I've heard reports that the ethernet is flaky, but I don't use that and I haven't had any issues connecting SSDs to the eGPU's USB ports for backups.
It also provides a full 100W of power to the machine which is important to prevent battery drain/degradation. Most monitors and TB3 docks only provide 87W or even 60W. The 16" MBP comes with a 94W power supply, but the next image shows the total system can use >=110W in bursts.
I don't buy into the argument that disabling turbo boost or setting a low total power limit with Volta and Voltage Shift has little or no impact on performance.
Obviously you won't notice it in all but the most extreme cases of slowdown, because you're too busy actually working. It's generally hard to notice the absence of something, anyway.
But from my own personal experience, I often DO notice that the performance IS available and is being used. I am frequently impressed that things just happen quickly, all the time. Not only massive compile jobs. Everything. Even everyday single core medium load tasks are happening at close to 4GHz. That's a 66% increase over the base frequency.
The sapphire pulse rx 5500 xt fans are almost always completely stopped. The stock razer core X fans were audible at 2 feet but way less loud and less annoying pitch than maxed MBP fans. From memory I think the stock fans were approx same level as MBP at 3000 rpm and noctua fans are approx same level as MBP at 2500 rpm. But in both cases, the pitch from the eGPU is better than the MBP at the same level.
Some users had a weird mechanical noise with two units (maybe a bad batch). Replacing the fans is not hard and does not require any permanent modification so you can undo and return for warranty claim.
I would consider the eGPU route, but based on some googling it appears that LG 5k TB3 monitor cannot be connected via an eGPU with its full resolution. eGPUs have HDMI/DisplayPort outputs and dongles max up to 4k. Does anyone know otherwise?
Also, I've got two 2560x1440 monitors on both sides of the 5k, and would like to connect all of them to the same eGPU. I haven't found an option that lets me connect these 3 monitors to an eGPU. I'm all ears if someone has made something like this to work.
[automerge]1595666101[/automerge]
I tried it for a day, and based on no measurements, wattage or otherwise, fans seemed more relaxed during regular browsing and non-taxing use. But this is pure anecdata, could even be a placebo effect. I'd be curious to see iStats, if someone wants to take a closer look. DP1 was predictably buggy as a dual boot daily driver and messed up iCloud on my Catalina installation, so I haven't been keen on trying DP2/3 yet.
Also, I've got two 2560x1440 monitors on both sides of the 5k, and would like to connect all of them to the same eGPU. I haven't found an option that lets me connect these 3 monitors to an eGPU. I'm all ears if someone has made something like this to work.
[automerge]1595666101[/automerge]
I tried it for a day, and based on no measurements, wattage or otherwise, fans seemed more relaxed during regular browsing and non-taxing use. But this is pure anecdata, could even be a placebo effect. I'd be curious to see iStats, if someone wants to take a closer look. DP1 was predictably buggy as a dual boot daily driver and messed up iCloud on my Catalina installation, so I haven't been keen on trying DP2/3 yet.
Last edited:
I think you could use the Blackmagic eGPU Pro. But they might be hard to find now, I'm not sure if they're still being manufactured and sold. The Blackmagic eGPUs were the only ones that could drive a TB3 display, and the Pro also has 1x HDMI and 1xDP for your two side displays. The standard only has 1x HDMI and the TB3.
It is expensive and might be hard to find now, but it is reportedly absolutely SILENT. You're probably better off just getting the 5600m.
I hate to be the bearer of bad news but after testing my new 5600M for the past week or so it seems to be only marginally better than my previous 5500M. The heat isn't as much of an issue and I haven't seen Radeon High Side go over 12w but the fans still ramp up when doing simple tasks like watching a full screen Youtube video or using Ableton Live. With Ableton Live on a basic project without many plugins the fans easily reach 100% and it gets very loud albeit not as hot as my 5500M.
So I'm very disappointed having spent all this money to upgrade to the 5600M thinking it would solve the problems I had with my 5500M. Granted I'm using a Pro Display XDR so it is a lot of pixels for the GPU to push but the GPU is never even really stressing. On activity monitor it's barely using any of the GPU's capacity so I have no idea why the fans need to ramp up so high.
It definitely runs less hot over all and never turns into a toaster oven like the 5500M but it still gets just as loud even when not under a heavy load. I'm not using turboboost switcher or any other workarounds because I didn't think I would need to with the 5600M. That was kinda the whole point of upgrading but I think I may need to in order to keep the fan speed down :/
So I'm very disappointed having spent all this money to upgrade to the 5600M thinking it would solve the problems I had with my 5500M. Granted I'm using a Pro Display XDR so it is a lot of pixels for the GPU to push but the GPU is never even really stressing. On activity monitor it's barely using any of the GPU's capacity so I have no idea why the fans need to ramp up so high.
It definitely runs less hot over all and never turns into a toaster oven like the 5500M but it still gets just as loud even when not under a heavy load. I'm not using turboboost switcher or any other workarounds because I didn't think I would need to with the 5600M. That was kinda the whole point of upgrading but I think I may need to in order to keep the fan speed down :/
Still pondering the 5600 upgrade myself. Maybe it's worth waiting a few months to see if they upgrade the CPUs to 10th gen and hopefully fix that issue by then?
What was CPU usage like playing video? What is your ambient temp? I tried a 4K Youtube video today full screen on one of my 4K monitors in Chrome - CPU didn't exceed 150% and fans peaked at 2650 RPM. My machine is 2.3Ghz 5600M. Ambient about 23 deg C. Had a few other things open (Outlook, Macs Fan Control, Activity monitor) and other Safari windows with multiple tabs.
I really don't know if it will be better with this move to ARM, especially in software - drivers.
The last samples from Apple, in software development, are not so encouraging (i.e. Catalina and iOS bugs, mail, music and other apps problems etc).
I hope that it will be great but I'm still sceptical about the end results.
Does iStats Menu work with the 5600M model yet? I do not see any sensors data, except the SSD temp. Do I need to do anything else to get all the temps and power usage stats?
[automerge]1595846530[/automerge]
My 5600M runs really well. No noticeable fan sound even after watching constant youtube 4k videos on the external for an hour.
[automerge]1595846530[/automerge]
My 5600M runs really well. No noticeable fan sound even after watching constant youtube 4k videos on the external for an hour.
You need a beta version of iStats I think. I believe there was a link in this thread a few pages back.
Edit: Yes, you guys were just discussing this above.
Last edited:
This is disconcerting. I had watched this video on the 5600M and assumed that the problems were fixed:
I had also come across this video where the person provides prescriptions for the other non-5600M GPU Macbook Pros:
I'm running a 5600m connected to a Dell 4k for the first time. The Radeon high side is around ~5.5 - 7.5 watts, my fans are not kicking on, and my mouse is much more responsive.
Edit: to complete the experiment, plugging back in 5500m into same monitor, same background processes (clone), radeon high side is consistently ~20 watts, my fans are blaring, and my mouse has a little lag. Scrolling seems a little choppy though.
Edit: to complete the experiment, plugging back in 5500m into same monitor, same background processes (clone), radeon high side is consistently ~20 watts, my fans are blaring, and my mouse has a little lag. Scrolling seems a little choppy though.
Last edited:
I suspect that using the Pro Display XDR is the culprit. I'm not experiencing massive fan ramp ups in Ableton projects using a 1440p 165Hz monitor on my 5600M model. Even medium load projects don't reach anywhere near 100% fan speeds (not using any turbo boost switcher or similar alternatives).
Yikes that’s what I was afraid of. Yet Apple says these machines can run 2 XDR displays. I’m beyond tired of Apple at this point. Everything buggy or just doesn’t work how it should lately
I connected my i9/5500M today to my friend's dual-monitor setup (x2 Philips 32" @ 1440p) via a single TB3 cable, and the Radeon wattage was still ~18W. That is literally the same exact wattage as when this MBP is connected to a single Dell 25" 1440p monitor. It's not supposed to be that way.
Clearly, driving an additional display should raise the GPU wattage even higher, but that's simply not how this is playing out. It is ridiculous that Apple would consider this behavior 'normal'!!
Clearly, driving an additional display should raise the GPU wattage even higher, but that's simply not how this is playing out. It is ridiculous that Apple would consider this behavior 'normal'!!
Strangely I noticed netflix is choppy and the quality is lower on the 5600. This is typically a bandwidth issue. However I tested it against the 5500 which is on the same exact wifi network, it has no choppiness and full resolution. Any ideas?