• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Ryzen 7 7800X3D CPU Burns Up

The trouble is, Asus make damn good boards and have some nice BIOS extras that make it worthwhile IMO. They do have a habit of throwing too much voltage though.

I was not aware of this but saw it posted elsewhere it looks like Z690 had issues on Asus too.

The z690 hero was incinerating itself and the z690 apex ( I had that one), you know the 2 dim mobo that's meant to overclock memory? Yeah, it failed to boot at anything above 5800mhz!! Asus was claiming it's samsungs DDR5 that is the problem and not the motherboard (back then no one knew any better, it was the first days of DDR5) but soon - after lots of testing - I found out that they messed up the first dim slot. 2nd slot was running 7000+ MHZ, first slot was stuck at anything between 4800 and 5800.

No official recalls or anything, if you were lucky and persistent enough they would just rma. Problem is, the issue was so widespread on the apex that lots of people ended up with a problematic mobo even after the rma. Lol

With that said, Asus ai oc is actually insane, it does manage to oc and fine tune voltages better than I can manually. It's absolutely bonkers
 
Last edited:
It seems Asus aren't alone in having issues though, based on the GN video. Asus was the worst of the ones tested, but they only seemed to test Asus and Gigabyte, so who knows what ASRock and MSI are like if put through the same testing. They also made it sound like there is some fault at AMD's end, so they'd all be affected by that.

I'm sure you can find someone that dislike each of the brands and probably someone that swears by each of the brands...

It's a minefield with possibly no right answer.
Well we know the accuracy of SoC voltage values of other manufacturers from the testing Wendell did in that GN video. ASrock was also off by quite a bit compared to what it showed in the bios though not quite as bad as Asus, but MSI seemed pretty accurate/close.

OCP it's hard to say without more testing.
 
The trouble is, Asus make damn good boards and have some nice BIOS extras that make it worthwhile IMO. They do have a habit of throwing too much voltage though.

I was not aware of this but saw it posted elsewhere it looks like Z690 had issues on Asus too.

If you're into memory tuning then Asus have shown for a good few generations that they tend to be head and shoulders above the rest. This might be one of the root causes for their prevalence in these current issues and trying to maintain that perceived superiority.

I also remember that reversed capacitor on the Z690 which was first spotted by Buildzoid. This was apparently just one batch due to lapse quality control at the height of Covid and came to light and rectified within a ~month of Z690 release.

This issue with AM5 is more universal as DerBauer showed with the tell tale bulg on his non X3D 7900X in combination with a Gigabyte motherboard.
 
As someone who only built their first PC a few days ago, I am wondering whether I should leave my system at stock until the overheating issue shakes out. I have EXPO turned off. Is it safe to stress test my system?

I have the following: 7800X3D, Gigabyte B650 AORUS Elite AX, 32GB Corsair CL30 6000 Mhz RAM

I am happy with my system at stock (for now). I can run my favourite game at maximum settings 4K/120. The only thing I would like is slightly less fan noise. I intended to do a PBO undervolt, but I am now concerned about things like Cinebench and CoreCycler burning out my CPU. Am I worrying needlessly? I am sorry if these are stupid questions, I learn as much as I can from these discussions, but I am still very new to this.
 
Last edited:
Good god @ the behaviour of the latest Gigabyte bios which is *supposed* to have the SoC fix in Buildzoids video.

Buildzoid shows that manually setting the SoC voltage under the main tweaker menu, doesn't have any effect if it's not also set under the AMD overclocking section.
Then he shows that setting auto on and off for voltages can start changing settings elsewhere, in the AMD overclocking section with janky manual values.

Worse, just flicking between auto and manual for the SoC voltage in the bios can suddenly result in the SoC jumping to 1.35v (despite it supposed to be limited to 1.25v)
Even worse still , manually setting the memory voltage to certain values can get the bios in a complete mess and it suddenly starts outputting 1.41v for the Soc voltage!

These Bios's are just riddled with bugs, and no doubt have contributed to the CPU failures out there with them having configuration traps in the bios that will start setting random voltages without telling you.

This is why you need to at least confirm with something like HWiNFO that the voltage is being set to what its supposed to have been set (yes HWiNFO is not 100% accurate but you can at least tell if the bios setting matches roughly what its supposed to be and is not miles out because it set something completely different all on its own!)
 
Last edited:
Even worse still , manually setting the memory voltage to certain values can get the bios in a complete mess and it suddenly starts outputting 1.41v for the Soc voltage!

I forget specifics now it was so long ago but there was a bug with a DFI motherboard many years ago where if you set the RAM voltage above a certain level then tried to drop it to any lower value by any means i.e. reset or manually decreasing, etc. it would instantly put the max voltage through the board chipset killing it dead.

EDIT: DFI Infinity 975X/G - I killed 3 boards before narrowing down what it was - nasty brick bug.
 
Last edited:
I forget specifics now it was so long ago but there was a bug with a DFI motherboard many years ago where if you set the RAM voltage above a certain level then tried to drop it to any lower value by any means i.e. reset or manually decreasing, etc. it would instantly put the max voltage through the board chipset killing it dead.

Like stepping on a land mine.....seems the BIOS mine fields are back.
 
Like stepping on a land mine.....seems the BIOS mine fields are back.

Gigabyte was always a little rough around the edges BIOS wise but with some decent core features, but development quality across the board seems to have dropped off a cliff of late.

Shame because back in the day they really put some effort into their boards.
 
There was another reported dead CPU on reddit last night. 7900X3D on an Asus board with a bios from March.

Ive got my 7800X3D on an MSI B650 board currently running a bios that was released on the 14th of April,this bios has since been pulled and there is a newer one. Ive only used my system a small amount a few Cinebench and 3D mark runs and a few minutes in games. Going off HWINFO my SOC voltage is just under 1.2vs and never goes above it and thats with expo enabled. Im wondering if I should leave it as is at this point or flashh to the newer bios.
 
There was another reported dead CPU on reddit last night. 7900X3D on an Asus board with a bios from March.

Ive got my 7800X3D on an MSI B650 board currently running a bios that was released on the 14th of April,this bios has since been pulled and there is a newer one. Ive only used my system a small amount a few Cinebench and 3D mark runs and a few minutes in games. Going off HWINFO my SOC voltage is just under 1.2vs and never goes above it and thats with expo enabled. Im wondering if I should leave it as is at this point or flashh to the newer bios.
1.2v is probably fine, but it's always advisable, for both intel and amd, after you enable xmp or expo to start lowering vsoc / sa voltages to see what's the minimum volt you are stable at. But, again, 1. 2 is on the low end, so you are fine
 
Good god @ the behaviour of the latest Gigabyte bios which is *supposed* to have the SoC fix in Buildzoids video.

Buildzoid shows that manually setting the SoC voltage under the main tweaker menu, doesn't have any effect if it's not also set under the AMD overclocking section.
Then he shows that setting auto on and off for voltages can start changing settings elsewhere, in the AMD overclocking section with janky manual values.

Worse, just flicking between auto and manual for the SoC voltage in the bios can suddenly result in the SoC jumping to 1.35v (despite it supposed to be limited to 1.25v)
Even worse still , manually setting the memory voltage to certain values can get the bios in a complete mess and it suddenly starts outputting 1.41v for the Soc voltage!

These Bios's are just riddled with bugs, and no doubt have contributed to the CPU failures out there with them having configuration traps in the bios that will start setting random voltages without telling you.

This is why you need to at least confirm with something like HWiNFO that the voltage is being set to what its supposed to have been set (yes HWiNFO is not 100% accurate but you can at least tell if the bios setting matches roughly what its supposed to be and is not miles out because it set something completely different all on its own!)

 
He needs to do another video for people who don't use expo, see how the soc voltage behaves then, because I set everything up manually, ram related, voltage, timings, speed etc, I've left the AMD overclock menu well alone, but it seems I need to go into it and duplicate my voltages in there too to make sure the volatges stick, at the moment I've only set them up in the AI tweaker menu, I'm got an Asus boards, but I serious doubt it's behaving any differently to the gigabyte board buildzoid was using, it coming down to the new agesa from AMD at the moment I expect.
 
He needs to do another video for people who don't use expo, see how the soc voltage behaves then, because I set everything up manually, ram related, voltage, timings, speed etc, I've left the AMD overclock menu well alone, but it seems I need to go into it and duplicate my voltages in there too to make sure the volatges stick, at the moment I've only set them up in the AI tweaker menu, I'm got an Asus boards, but I serious doubt it's behaving any differently to the gigabyte board buildzoid was using, it coming down to the new agesa from AMD at the moment I expect.
I've always thought the BIOS's behaved strangely and left alone what I wasn't sure of. Glad it wasn't just me! It should be either/or though, if you enable another menu it should disable the other one and not have competing settings.
 
You see my response to that would be, is, don't buy Asus motherboards.

I've been saying that for a while.

That is too parochial a notion as based on my experience over the last several years I could easily say don't buy Gigabyte but I would never give that out as general advice. You have to try and assess each individual component on its merits as pertaining to your requirements and hope that initial reviews are thorough enough to highlight any pitfalls.

Though if you're an early adopter then it can be a bit like Russian roulette, while you pray to the tech gods you come out the other end unscathed.

As already stated, it can be a bit of a minefield which is why my motherboard policy has been the same as my OS one; wait ~1 year for most major bugs and issues to come out in the rinse.
 
Last edited:
Looking at building new AMD system, is the problem only with 3d chips or is it rest of AM5 chips as well ? Just want to put system together without overclocking or fiddling with bios too much.
 
Good god @ the behaviour of the latest Gigabyte bios which is *supposed* to have the SoC fix in Buildzoids video.

Buildzoid shows that manually setting the SoC voltage under the main tweaker menu, doesn't have any effect if it's not also set under the AMD overclocking section.
Then he shows that setting auto on and off for voltages can start changing settings elsewhere, in the AMD overclocking section with janky manual values.

Worse, just flicking between auto and manual for the SoC voltage in the bios can suddenly result in the SoC jumping to 1.35v (despite it supposed to be limited to 1.25v)
Even worse still , manually setting the memory voltage to certain values can get the bios in a complete mess and it suddenly starts outputting 1.41v for the Soc voltage!

These Bios's are just riddled with bugs, and no doubt have contributed to the CPU failures out there with them having configuration traps in the bios that will start setting random voltages without telling you.

This is why you need to at least confirm with something like HWiNFO that the voltage is being set to what its supposed to have been set (yes HWiNFO is not 100% accurate but you can at least tell if the bios setting matches roughly what its supposed to be and is not miles out because it set something completely different all on its own!)
Yeah this is what is more concerning to me. Vendors need a good whipping. Whipping incoming from youtubers and the community :)
 
Back
Top Bottom