Updated: BeeLink GTR7 7840/7940 Pro Random Reboot/Crashing Issue (The Reason and a possible fix for some)

I doubt it has anything to do with the drivers as I’ve had the same issue in Linux as well.

After update the drivers (08/21/2023) … the reboot issue remains :frowning:

They could not find a solution for the reboot problem.

I returned my unit to China, and they received it today. They should refund me shortly.

The device is great, but the reboot problems are still not solved after almost 2 months!!

I bought Intel NUC 13 i7 1360P 64GB DDR4 and 2 TB SSD for 993 USD and it is great. Stable and solid like a rock with no reboots and the performance is outstanding compared to the GTR-7 that cost me around 770 USD.

Using 1080p to stop the reboot issue is not a solution. It means that the device is useless since I will be stuck with 1080p forever.

I’m experiencing the same issue here. It’s becoming quite frustrating that Bee-link hasn’t provided any official statement about this problem. I’d like to know if they’re collaborating with AMD to address this issue. It’s a fantastic piece of hardware, but these problems are really irritating. It’s not related to 1080p, video, etc. I’m running Proxmox, and I’m facing the same issues.

One thing to mention: I’ve conducted some tests. I cleared the CMOS, but the issue persists. I also tried booting into Windows with some apps running, but the problem remains, and there are no logs indicating the cause. I removed the bottom plate because initially, overheating was causing the machine to malfunction. As the machine is on my desk, I noticed that when it gets stuck, I can feel the airflow at the back, and it’s hot – very hot. It’s evident that something is happening on the CPU or GPU side.
CPU downclocked as well, same issues.
Was monitoring CPU and ssds temperature and everything normal.
Disabled hibernation on BIOS, same issues.

I disabled SR-IOV, but the problem persists. Apart from the random reboots, I observed a somewhat peculiar behavior on Proxmox. If I start Proxmox without launching any VM, the machine stays online for hours. However, if I start a container or a VM, I experience a reboot in less than an hour. Sometimes it takes 2 or 3 hours, but I’ve noticed that the reboots occur more quickly when these are running.

A day ago, I stress-tested the machine with CPU, memory, and IO on Proxmox for a few minutes, and there were no issues. The reboots seem to be triggered by a very specific reason. It’s frustrating that we can’t access any BIOS logs.

I sent an email to Bee-link asking for the BIOS update that I’ve heard others talk about, but I’m perplexed as to why it’s not available on their website. Additionally, for some strange reason, their support site is inaccessible in Brazil; it seems to be blocked for us.

Let’s see how this issue gets resolved. I can be patient, but it’s currently challenging to recommend their products. I also have a suspicion that this might be more of an AMD issue with Phoenix, which could explain why it took them so long to launch it in laptops.

Hello all,

Just wanted to inform everyone that AMD has released a new updated Laptop\mobile chipset driver. This may help with the C-state issues on the 7840/7940HS chips. As I no longer have mine, I can’t test it out but if anyone wants to give it a go and see if this helps.

The driver is on AMD’s site, just go to Chipsets=>Laptop Chipsets and only one option after that.

1 Like

As weird as it seems, you could try to reach them through their Aliexpress page. I’ve managed to get a recovery ISO for Windows 11 from their chat there. I’m in Brazil as well.

The second option is to try to reach their support website through a VPN (I’ve tried that). In my case I’ve opened a ticket, but never heard back from the support team.

The chat from Aliexpress was effective, though.

First comment here on LevelOne…

hmmm… Just took delivery (rec’d very quickly - ordered Monday, rec’d on Friday (only 5 days shipping time from CN to US) - on a new GTR7 Pro (7950 HS)…

I was VERY surprised that it came ‘without’ a vented bottom plate - as I waited to order for a month hoping the new plates would get into newer orders - nope… so I have sent email requesting a new plate.

Included was a black (not white) power supply - and BIOS firmware per the BIOS screen listed as GTRP01P5C6V28

Vented the bottom to test with 18mm x 2.5mm standoffs to open it up for airflow… if it blue-screens/reboots, it gets a DHL plane ride/returned.

I do wish I had seen/read this thread before ordering. I’m patient but I’m not stupid >> i.e., exactly where is that bottom fan supposed to source airflow from without a vented bottom plate? - oh well.

Hello friends.

Beelink posted in the official forum, that the problem is in the motherboard. They are replacing all units with problems.

In the new units, there are a module to control something relative to energy/memory (as I could understood) .

It is too late, I had the problem for 1 month and no solution had been provided. I returned my unit and got a refund. It has been almost 2 months now

@Jorge_Amadeu Well I will let everyone here know here shortly if the new units actually fixed the issue or not. Beelink just informed me that my replacement unit was shipping out today. So as soon as I get the new unit, I will run it through the gauntlet of tests to see if the random reboot issue is still present or not.

I would like to share my bad experience with returning the computer.

I bought the mini pc from the official Beelink store, on AliExpress. And as I had reboot problems, official support via email advised me to return the computer, through the AliExpress seller.

They said they would send a new unit as soon as the old one arrived in China. Well, the computer arrived yesterday and now they said they will send it to the factory to be repaired and then they will send me the same mini pc again.

Bad experience, I will have to wait even longer to receive the mini pc back.

They are “soldering” an additional board to control something related to voltage. I saw them writing this on the official Beelink forum.

We are waiting for the tests …

I returned my unit and got a refund … I then bought Intel NUC 3 Pro with 64GB DDR4 and 2TB SSD … I found the performance much better than the beelink probbaly because of RAM … but the price difference is huge 993 vs 769 USD

I may buy another unit soon, if the beelink problem is fixed, I will definitely buy it

Yes, they added a small board to fix the issue … so my computer is probably going to be repaired then sold to another customer as a new unit !

Okay, we got our replacement GTR7 unit today.

The replacement unit was a brand new, sealed unit in full retail packaging. It did not come with the new fan grill bottom cover but luckily, we kept ours from the last unit so we will test with both covers.

We will start testing this unit this week to see if the random reboots have been fixed.

Testing will be done with the stock RAM, SSD and Windows Pro install that came with the unit. We will install the latest updates for Windows 11. Drivers used will be the latest AMD Adrenalin 23.9.1 Driver for APU and AMD Chipset drivers 5.08.02.027 from 08/17/2023.

We will be testing idle behavior with multiple monitors, several 1080P HDMI & DP monitors, a few 1440p HDMI & DP monitors, as well as 4K HDMI, DP & DP over USB-C native using an ASUS Pro PA329CV.

Also, will revisit temperature testing with both covers on the new unit in gaming and benchmarking.

If all goes well, we will then swap the RAM back to the faster G.Skill RAM and the Samsung 990 Pro SSD’s with fresh Install of Win11 Pro to see how it runs with upgraded components.

We will post all our findings here for anyone that wants them.

2 Likes

how is the unit performing so far?

Hello! Do you have any news about your mini pc? I asked for a refund and am wondering if I would buy again.

I returned my unit and got a full refund 769 USD
I bought NUC 13 1360p with 2TB SSD and 64GB of RAM for 993 USD

Intel is much better in term of performance

@Jorge_Amadeu Still running through the random reboot testing. As the issue occurs at idle and appears to happen when displays over 1080p are connected. It takes time to test it properly, so we are letting it run as long as possible during the day just running something and then we let it idle through the night to force the worst-case scenario on the unit. If it makes it through next week which will be two 4K monitors as this week was testing with two 1440p monitors connected. Then we will start the temp and performance testing on it and will post everything here when it is all completed.

2 Likes

Got a little update here on the our replacement GTR7 unit.

So after testing the unit on 1080p, 1440p and 4K monitors. We have not experienced any random reboots or crashing issues. The system has been 100% solid and has been beat to death, non-stop for over two weeks now.

We tested the monitors using the native HDMI and DisplayPort as well as USB-C to HDMI and USB-C to DisplayPort adaptors. The monitors were tested in dual and tri-monitor arrangements throwing in a USB-C adaptor or native USB-C connection on the monitors.

During our testing we found that the new fan cutout bottom was much better in cooling the under side components but the unit still required to be propped up about 1/2 in off the desk for the best cooling results.

Quick Example, This was done playing a simple game for 30mins @ 1440p maxed settings (Have a Nice Death)

Unit sitting on the desk with original solid bottom cover.Memory Temp was DIMM1 89C & DIMM2 91C.

Unit sitting on the desk with updated fan grill bottom. Memory Temp was DIMM1 75C & DIMM2 79C.

Unit 1/2in off the desk with updated fan grill bottom. Memory Temp was DIMM1 59C & DIMM2 67C.

So the the new bottom with the fan grill cutout does help quite a bit but still need to have the unit off the desk about 1/2 in for maximum cooling.

If anyone wants anymore testing on temps, performance or Linux testing. Let me know. We will have the unit in our LAB until I can find a place or use for it.

4 Likes