VehiclesFashionRecipesBlogsHuntTravelsSportFunHandmadeITEducation
Mini-Games
x

x
zakruti.com » IT - Software » Gamers Nexus
We Exploded the AMD Ryzen 7 7800X3D & Melted the Motherboard

We Exploded the AMD Ryzen 7 7800X3D & Melted the Motherboard

FBTwitterReddit

video description

Rating: 4.5; Vote: 2
This video benchmarks and tests various points of failure for the AMD X3D CPUs in particular, but also just generally AM5 / Ryzen 7000 (Zen 4) CPU failures. The most eye-catching failure mode is catastrophic: Users have posted photos of exploded CPUs with blown-out motherboard sockets and melted plastic. We bought those parts from the users to perform a failure analysis, then set forth testing our own AMD Ryzen 7 7800X3D CPUs to see if we could make them explode. Spoiler alert: It worked. We were able to melt new AMD X3D CPUs in the socket of an ASUS X670E Crosshair Hero motherboard, but we also found severe BIOS bugs in Gigabyte's BIOS. SOC voltage isn't the only concern, here. It's also poor OCP implementation on at least the ASUS board we used, plus a combination of the erosion of thermal and over-current protections. As stated in the video, no test of this type is fully conclusive: These are complex issues with functionally infinite variables between users (and possible error), manufacturers, BIOS versions, and AMD itself. As such, we can only speak to our findings in some instances -- but we have some that do point to specific issues. We killed 2x 7800X3Ds and 1x ASUS motherboard, but we also had 2 failed CPUs and 1 failed board from viewers. We will have at least 1 more part following this video up. We've sent parts out to the same failure analysis lab we worked with in our NVIDIA 12VHPWR content - it'll take a few weeks to put together.
Date: 2023-04-30

Comments and reviews: 14


i recently build a new pc with the 7800X3D CPU and Asus ROG Strix X670E-A motherboard.
i am by no means an expert on all these things, and alot of the stuff mentioned in this video i do not fully understand.
so i am hoping someone can clarify abit for me, and see if i understood the situation correctly.
the different brands asus/msi/gigabyte etc, is pushing out bios updates to limited the SOC voltage(this i do not know what actually is) so the CPU doesn't get fried under startup?
although this wont happen for everyone it seems like it is common enough to have some kind of concern i assume? after i build the PC (not currently running because i had to RMA the GPU) the boot time was INSANE long which from what i can understand is some memory training, which apparently takes many minutes to complete for the first many months, i assume this can contribute towards this issue when the CPU gets over voltage everytime? my AIO did run during boot times though.
agian there was alot i didn't truly understand part because of technical words and english not being my main language, is there anything us as consumers can do at this time? or is it just a waiting game to see if this happens and hoping bios updates help?

reply

Truly astounded and horrified by Asus deliberate negligence . By a company such as this.
There are 2 types of mistakes.
The, i did all i should have done and the unexpected happened , that is the unforeseen,
and the, i did a sloppy job and i know it, to rush the product unto the shelves and subsequently peoples' computers, and pockets, hoping it will all work out and i'm not gonna get exposed , that is deliberate.
All this is basically the legendary gigabyte exploding psu all over again.
But (as all good horror sequels) now, it's with a different twist.
It's the motherboard version , times more the f up, and a different, bigger, brand.
Asus, had a few mishaps in its past, but never something like this. This is utterly unacceptable. This is pitchfork,tar and feather stuff.
Asus is dead to me after this.
I really like my Prime B550 Plus, solid build, works great, does it all, great value, but that's the end.
Never ever gonna buy something of theirs ever again. To the black hole they go, next seat to where Gigabyte is sitting -since old.
A BIG THANKS to GN for all the hard work, time and money they put in, to make this piece and others like it!

reply

Anyone more knowledgeable then me pn this point, could u please guide me what to do? My enviroment: i'm running a 7700x with a gigabyte x670 aorus elite ax wifi, cooled by a 360 aio. Running fine for 3ish months now.
I did not update bios since purchasing the mobo, since i've read some horrorstories, and combined with gigabytes legendary after sales service, i didn't bother as it just works. Their page with bios states this aswell, don't update if everything works. After this issue coming to light, i've checked hwinfo while gaming, all cores have a nice clock of 5540mhz almost constantly with a temp mid 50's, the thing concerning me is that the SOC voltage sits continuously at 1,39xxx. Is this a problem on the regular x variants? What should i do. Thanks
Thanks

reply

I noticed your MSI board had SOC voltages below 1.2V, even with EXPO. On the other hand, my tests of my MSI Pro X670-P WiFi board (using HWinfo) showed an SOC under 1.2 stock, but that immediately went to 1.36 when XMP (it can run both XMP and EXPO) was enabled. This was prior to the newest BIOS; I haven t had the chance to test that yet.
One thing I couldn t help but notice in the motherboard chart late in your video was that all the boards that didn t push SOC over 1.3v were B650-based; while it appeared that all X670 boards had that problem. Is it possible that this only affects X670 designs, while the lower-priced B650 platform was assumed to not need extreme performance that might lead board manufacturers to play fast-and-loose with voltages?

reply

Hi, I have Asus PRIME X670E-PRO WIFI MOBO (bios ver BIOS 1414 - latest ) and ryzen 7 7700 CPU . (since 3 weeks)
Memory kit EXPO - 6000 mhz voltage is 1.4 ( Corsair 32GB(2x16GB) Vengeance RGB 6000mhz CL30 DDR5 AMD EXPO Ram (CMH32GX5M2B6000Z30K) )
Now Expo is turned of to jedec 4800 mhz 1.1 v settings.
3 questions here ;
- Will I ever can turn on my EXPO - 6000 mhz option safely again ?
- Almost 2 weeks used with EXPO on , does it make any damage to my system or CPU ? (haven't noticed anyting at all)
- Will AMD , motherboard and memory kit vendors would accept their fault off this EXPO disaster and give me my money back ?

reply

Dear guys,
I just wanted to express my gratitude for posting that informative video about the current issues with CPUs and motherboards. Your video helped shed light on the problems and allowed us to make informed decisions about when to purchase these components.
I'm glad that we now know to wait until these issues are resolved before investing in new CPUs and motherboards. Your dedication to providing valuable information to your audience is truly appreciated.
Once again, thank you for your hard work and for keeping us informed.
Best regards,

reply

These issues are not good for AMD and its board partners. They are charging more for CPUs and motherboards but the quality of design, testing and quality assurance is not like what it was before. GN is 15 years old and I am not sure if Steve has seen this kind of sloppiness in the industry. Pay more, get less. Isn't this encouraging people to hold off their purchases until we see some kind of V2.0 products on the market?
Although this kind of problem is hard to reproduce, no one wants to encounter it by accident. One such failure will be once too often!

reply

so i just ordered 7800 when all this started coming out...its sitting in my living room, still in the box. are the bios updates coming out actually a good enough solution to safely use the cpu without worry? or im i going to have to sit around praying my computer doesnt die at any moment?
dont want to spend 750 (cpu+mobo) to live in fear of my money (possibly literally) going up in flames at any time.
i guess we cant really know until these new-new bios updates drop and see....

reply

So, my 7800X3D had arrived on the 18th and was put into the Asus X670 Pro Gaming Wifi on the same day. For the first few days, and some gaming, i had put the RAM to the AMD ANNOUNCED SWEET SPOT of 6000. Which set the SoC V to 1.36 or thereabout, maybe 1.368...
Then, once the reports came out, I went back to auto/capped it at 1.25, later 1.2V. And barely used the PC, aside of some benchmark stuff. Currently on Auto again, waiting for the AGESA Bios update.
Should I worry?

reply

AND these motherboard manufactures want more and more for their junk.
These boards are not worth the price
Where is the R&D from these manufactures? Right, they don't do that anymore and just design from specs
Asus Auto voltage has been buggy like this for a while.. I will never use auto as its almost certainly higher than what is shown in BIOS. This is from my own readings going back years, almost a decade..

reply

And what if unaware customers do not upgrade their bios ? I mean many just look into these type of content only when they are building their PC and after that they just do not continue to follow up. This is a poor handling of the situation. Motherboard vendors and AMD should at least work with Microsoft to show a warning message inside windows about a critical bios update to those affected.
reply

ido not buy hardware at release anymore, unless gamersnexus had it in their hands. and i dont buy a game anymore, unless digitalfoundry had their hands on it. i cant live without techtubers anymore because everything is scammy or combusts shortly after release. timebombs. always best to let something sit for a long time and let the kinks get worked out by betatesters. i swear.....
reply

As someone with a 7800x3D, an Asus x670E-A Gaming and a lot of anxiety at the moment... what is it that I should do in order to run EXPO on some Z5 RAM at 6000mhz CL30? Do I just get the latest BIOS and turn on EXPO? Or do I still need to manually set the SOC and other voltages to specific numbers to prevent the CPU and MOBO from effectively killing themselves?
reply

Nvidia 12 pin failure: It's COMPLETELY USER ERROR, with most users being just that. Individuals that are NOT professionals and prone to do something wrong.
AM5 socket failure: It's AMD's fault, because it doesn't add a huge list of restrictions to motherboard makers, motherboard makers who ARE professionals with decades of experience and highly experienced.

reply
Add a review, comment






Other channel videos