AMD Ryzen 7 7800X3D CPU Burns Up

[str]

Funny thing is I got called crazy for making comments about this kind of stuff earlier in the thread

Some quite weird posts in these threads here and elsewhere when it comes to trying to deflect from Asus, I didn't realise they had such a fanboy following and/or using shills to that extent.

Haha yes. Keep up the good work!

 

[Rawr]

I'm confused why people are advocating for less coverage of questionable practices really. I dunno about anyone else but I'm glad these guys are covering it. They got Asus to walk it back and now cover boards with a beta bios as well. I don't think any other manufacturers will try anything either cos of it. That's a good outcome even if you have a Asus board.

 

[Troezar]

I'm confused why people are advocating for less coverage of questionable practices really. I dunno about anyone else but I'm glad these guys are covering it. They got Asus to walk it back and now cover boards with a beta bios as well. I don't think any other manufacturers will try anything either cos of it. That's a good outcome even if you have a Asus board.

I work on the basis I have no idea who or even what is posting this stuff these days. You may not even be arguing with a real person so I try to limit it, I'm old enough to not want to waste too much time on it I do agree it is a good thing for these issues to be made public but it does get hyped. I just want to hear what the resolution is. I think companies are slowly getting the message that being open and honest and a quick resolution is the best option in these days of transparency and the internet.

 

[Greebo]

Did they though? My understanding is they tried to retrieve the dead parts so they could investigate. GN beat them to it and purchased some of the parts themselves. Things don’t always happen fast in big organisations. They probably were trying to investigate matters themselves before making any kind of public statements.

Not agreeing to a recorded meeting might seem shady but I don’t buy that ASUS were ever being malicious anywhere. I’m glad there was some pressure from GN and others to get the focus on the issue and get it resolved ASAP, as well as forcing ASUS to remove the warranty text. But there does seem to be an element of hysteria and a lot of it is driven from high profile social media accounts.

Hmmm they were only bribing people with failed motherboards who had sold their board to GN with replacement mobo and cpu and a "rog item of their choice" . All the other people with failed boards and cpus who werent sending them to GN got offered nothing. Where was everybody's free choice of an ROG product?

 

[GoogalyMoogaly]

A lot was made of it and I'm guilty of that too. I'm still not sure whether to return my Asus board or not. I've lost a bit of confidence in it/them.
That said, I do have to wonder if such a fuss wasn't made, would we have seen Asus respond the way they have it would we still be seeing beta BIOSes with warranty void statements and warnings around using EXPO?
Maybe it was the only way to make them listen?

 

[rare]

Where was everybody's free choice of an ROG product?

Just how many people were there? Because it’s literally single digits.

 

[Nutty667]

It's probably the best time to own/buy an Asus board. You can be sure they'll have upped their QA in the last few weeks.
And the best thing is there's loads of discounts on Asus boards too

 

[rare]

GIGABYTE's Statement Regarding SOC Voltage on AMD's AM5 Motherboards Under 1.3V - Hardware Busters

Hardware Busters - GIGABYTE's Statement Regarding SOC Voltage on AMD's AM5 Motherboards Under 1.3V - News

hwbusters.com

 

[RBImGuy]

It's probably the best time to own/buy an Asus board. You can be sure they'll have upped their QA in the last few weeks.
And the best thing is there's loads of discounts on Asus boards too

when the news broke I wasnt worried.
all it took was a simple bios fix and asus did that.
I got the 7800x3d and gee that is some great cpu

 

[Greebo]

Just how many people were there? Because it’s literally single digits.

And all of them weren't offered to be bought by GN but it was only them who got offered a free ROG product if they changed their mind. So it doesnt matter if it was 3 or 23 other people affected, why didnt all the Asus customers with dead cpus get the same offer?

 

[rare]

And all of them weren't offered to be bought by GN but it was only them who got offered a free ROG product if they changed their mind. So it doesnt matter if it was 3 or 23 other people affected, why didnt all the Asus customers with dead cpus get the same offer?

No idea but there’s a lot of assumptions being made. Perhaps they did offer other people something?

I think GN did a good investigation, but there’s a lot of hysteria here being pointed at one brand in particular and when you look at the actual details it seems all pretty lightweight stuff to me.

 

[FredFlint]

No one knows how much long-term damage has been done to chips that have been running for 6-7 months with > 1.3v. Hopefully any problem happen before the warranty ends. AMD should extend the warranty by a year, that would a sort it.

 

[Greebo]

No idea but there’s a lot of assumptions being made. Perhaps they did offer other people something?

I think GN did a good investigation, but there’s a lot of hysteria here being pointed at one brand in particular and when you look at the actual details it seems all pretty lightweight stuff to me.

No it was pointed at Asus for two good reasons

1. Asus has a history of pushing the highest voltages through their boards when left on auto in order to be the fastest boards in reviews and when the GN investigation looked into it, Asus on some boards was pumping 1.45v through the SOC, way above AMD old guidance anyway. Yes GB and others were also overvolting and past 1.3 but by a much smaller margin.

2. The way Asus dealt with it compared with the other board makers.

3. What are Asus going to do about all the 1000s of cpu;'s which may well be still alive today but for months may have had up to 1.45v put through the SOC and will have damage to cpu, not enough that the cpu has failed yet but there will be damage and may shorted its life.

 

[rare]

Gigabytes boards were tested at around 1.42v and above. Both exceed the 1.4v limit AMD set out.

I genuinely don’t see how ASUS have dealt with it in a particularly different way. They released a statement quite early on setting out what was happening and begun releasing beta BIOS. The biggest difference was they quite rightly retracted their standard ‘warranty’ text on their bios releases when called out on it. They offered a particular customer additional parts to get hold of their failed bits.

The third point is purely it’s and buts. If a Gigabyte Master has been pumping out 1.42v odd then they might well have a ton of problems on their hands as well. No one will know. Personally, I don’t think it will have done any long term damage.

Gigabyte and others seem to have got away pretty sneakily given the focus put on Asus by a couple of high profile YouTube accounts which whipped up a load of hysteria.

 

[Silent_Scone]

No it was pointed at Asus for two good reasons

1. Asus has a history of pushing the highest voltages through their boards when left on auto in order to be the fastest boards in reviews and when the GN investigation looked into it, Asus on some boards was pumping 1.45v through the SOC, way above AMD old guidance anyway. Yes GB and others were also overvolting and past 1.3 but by a much smaller margin.

2. The way Asus dealt with it compared with the other board makers.

3. What are Asus going to do about all the 1000s of cpu;'s which may well be still alive today but for months may have had up to 1.45v put through the SOC and will have damage to cpu, not enough that the cpu has failed yet but there will be damage and may shorted its life.

1. AMD's old guidance in AGESA was 1.4v. There are videos in this very thread that show how these initial values taken by media were not indicative of what was being received at the die as well as the methodology in taking measurements being invalid. AMD has only in retrospect addressed this limit and lowered this limit to 1.3v. The correct value can be seen by monitoring AMD's SVI3 which is located inside the CPU. It can be monitored in Ryzen Master or HWINFO.

2. GN got caught up on a statement ASUS North America made because it was poorly worded and didn't quite make sense. He opted to probe them on the PR rather than ascertain whether he was measuring the voltage correctly and continued ahead, which ended up with the data being invalid. He then proceeded to make the same mistake again in claiming the SOC was still exceeding 1.3v after the UEFI update was released.

3. Nothing, as none of the CPUs will have seen that much voltage.

Nobody can be exonerated here, the situation was handled poorly by everyone involved.

 

[Greebo]

So some Asus boards putting 1.45v through SOC on auto has now been disproven?

 

[Silent_Scone]

So some Asus boards putting 1.45v through SOC on auto has now been disproven?

As far as I'm aware (and also present in GN's own video) 6000MT auto rules were 1.35v set.

 

[Robert896r1]

So let me get this right. AMD change vsoc limit from 1.3 to 1.4

Chips start burning up. Mobo manufacturer take the L for AMD.

AMD goes back to 1.3 limit as a “fix.”

However no way of knowing long term sustained damage as in accelerated degradation for chips during that period.

 

[Brizzles]

Its looking more like ;

1. AMD lets a rogue/testing Agesa ( with SOC 1.4v as a 'guide' figure maximum ) loose on the board vendors ( hence multiple board vendors invloved )
2. CPUs start failing
3. Asus and others say ' hold on you said 1.4 was a guide figure - not a total failure figure '
4. AMD say ' yeh we kinda got the numbers wrong ' so take the flak for us as we've got more to lose. We'll see you right financially / or we'll drop you from any future board developments
5. AMD release Agesa 1,0,0.7a reducing soc to 1.3 from 1.4
6. Today

 

[jigger]

Its looking more like ;

1. AMD lets a rogue/testing Agesa ( with SOC 1.4v as a 'guide' figure maximum ) loose on the board vendors ( hence multiple board vendors invloved )
2. CPUs start failing
3. Asus and others say ' hold on you said 1.4 was a guide figure - not a total failure figure '
4. AMD say ' yeh we kinda got the numbers wrong ' so take the flak for us as we've got more to lose. We'll see you right financially / or we'll drop you from any future board developments
5. AMD release Agesa 1,0,0.7a reducing soc to 1.3 from 1.4
6. Today

Maybe.