• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

My 9800X3D died

erek

8=D
2FA
Joined
Dec 19, 2005
Messages
17,745
no symptoms or warning signs

we had some lightning rolling through and i gracefully shutdown my pc and switched the psu off completely and left it sit for a day

come back and try to power it on:

dram (yellow) for about a minute
cpu (red) for a few seconds

then stuck on vga (white) indefinitely

i stripped the build all the way down to just cpu, mobo, 1 stick of ram at a time

reseated cpu, reseated ram stick and even swapped ram stick

cleared cmos

bios flashback to latest version

and never gets passed solid white vga light

—-

went to my local best buy and bought a 9700x to troubleshoot with and not expecting much.. was thinking mobo is probably dead


…wrong

the rig POSTed and is now fully operational and within margin of error performance of the X3D

i tried swapping back in the 9800X3D, still doesn’t work


going to RMA
 
1771725098332.png

https://www.3dmark.com/compare/sn/12347512/sn/11766478
 
Yikes! So much for people always saying it's "user error". What motherboard were you using?
asus tuf gaming x670e plus wifi ,


no scorch or burn marks. no bubbling, and also no prior warning signs such as system crashes or anything


just i shut it down gracefully and cut all mains power at the psu

by the next day it wouldn’t POST
IMG_3912.jpeg
 
in cyberpunk 2077 i lost about 5-8 fps but both are averaging 160+, so not noticeable even mins are about 140s to 150s
If you only lost a few FPS there was already something very wrong with your 9800X3D. Cyberpunk is a known game that hugely benefits from 3D Vcache CPUs, I don't think it was the storm that killed your CPU, I think it was already on the way out and it was just a matter of coincidence in the timing of them happening.
 
If you only lost a few FPS there was already something very wrong with your 9800X3D. Cyberpunk is a known game that hugely benefits from 3D Vcache CPUs, I don't think it was the storm that killed your CPU, I think it was already on the way out and it was just a matter of coincidence in the timing of them happening.
9800X3D:
1771728718230.png


9700X:

1771728739735.png


opinion?
 
  • Like
Reactions: xx0xx
like this
Well at 4K with RT Overdrive and multi frame generation muddying it up, it's no wonder you're not seeing a huge difference. It's really hard to diagnose anything when you introduce MFG into the equation, no baseline to reference no idea how many actual frames rendered vs generated.
 
As crazy

If you had tested at 1080p youd see a larger difference in results was what El_Capitan was pointing out I believe.
Yeah, test at 1080P native with no frame gen and no scaling, and no RT Overdrive like CraptacularOne stated. You'll see a vast difference. The CCD X3D isn't going to be as visibly noticeable at 4K, and with the RTX 5090 doing all the heavy lifting.
 
As crazy

If you had tested at 1080p youd see a larger difference in results was what El_Capitan was pointing out I believe.
Yeah, test at 1080P native with no frame gen and no scaling, and no RT Overdrive like CraptacularOne stated. You'll see a vast difference. The CCD X3D isn't going to be as visibly noticeable at 4K, and with the RTX 5090 doing all the heavy lifting.
i understand, but i just prefer to know overall system performance at 4K
 
i understand, but i just prefer to know overall system performance at 4K
Yeah, same here. I mainly benchmark with settings I play at, and then compare the differences. However, when testing or comparing CPU's for gaming, 1080P native is better to gauge the performance difference, and try to leave the GPU as much out of the equation as possible. Which is why I highly dislike it when CPU reviews don't test at 2K and 4K native, because the CPU becomes less dependent and the differences are minimal.
 
I'm curious what your SOC voltage was. It might not matter, but I have read some comments in other places about it. I put mine to 1.1 it was 1.3 out of the box.
 
I'm curious what your SOC voltage was. It might not matter, but I have read some comments in other places about it. I put mine to 1.1 it was 1.3 out of the box.
i am curious too

wondering if my motherboard slowly killed the 9800X3D, but i had no symptoms, crashes or anything

all i did was shutdown the pc, and switch off the PSU for 1 day
 
This is why I generally always keep an eye on the voltages the for the mb/CPU. Very interested what they were previously. Both Intel and AMD run voltages higher than they should be shortening the CPUs lifespan. Try to always run hwinfo so you can see the average and max spikes after gaming sessions.
 
This is why I generally always keep an eye on the voltages the for the mb/CPU. Very interested what they were previously. Both Intel and AMD run voltages higher than they should be shortening the CPUs lifespan. Try to always run hwinfo so you can see the average and max spikes after gaming sessions.
with the 9700X so far:

1771734356559.png

1771734505388.png
 
This is what I was referring to, see that vcore and vid max all the way up to 1.43 volts? To meet that is crazy high. I would keep it below 1.25v to be safe or 1.3v max. Some are totally ok pushing spikes beyond 1.4v but I'd rather play it safe and stay below 1.3 in the 1.275 max area personally.
 
Well at 4K with RT Overdrive and multi frame generation muddying it up, it's no wonder you're not seeing a huge difference. It's really hard to diagnose anything when you introduce MFG into the equation, no baseline to reference no idea how many actual frames rendered vs generated.
This is why 720p/1080p "CPU tests" are useless since that is not how people play their games.
 
This is what I was referring to, see that vcore and vid max all the way up to 1.43 volts? To meet that is crazy high. I would keep it below 1.25v to be safe or 1.3v max. Some are totally ok pushing spikes beyond 1.4v but I'd rather play it safe and stay below 1.3 in the 1.275 max area personally.
Idle vcore is right at 1.4V for AMD CPUs since AM4 when I started using them, and that continues with AM5 across many CPUs (7800X3D, 9800X3D, 9950X3D). It's just what they're run at (mine and friends'), and under load it drops down. Though full disclosure, I do run my 9800X3D at a fixed voltage of 1.23V and 5425MHz.

This is why 720p/1080p "CPU tests" are useless since that is not how people play their games.
It makes a pretty big difference even at 4K in games that are actually CPU heavy. Star Citizen is the one I can speak to, as I've upgraded over the years from an i9 10940X, to a 5800X3D (and the same RAM at the i9), 7800X3D, 9800X3D, and 9950X3D (all with the same RAM as the 7800X3D). Outside of the 9950X3D being a definite sidegrade for gaming, I saw significant average and minimum FPS gains with each upgrade.

Star Citizen is an exception, though, and not the rule. I've heard that MMOs like World of Warcraft is another. The majority of games will see minimal improvements at 4K unless there's other significant issues with the system (e.g. bad thermal paste application / cooler mounting). Although running at 4K but with DLSS performance mode can change that equation a bit, as it will hit the CPU more than standard 4K as it's a 1080P upscale.
 
Idle vcore is right at 1.4V for AMD CPUs since AM4 when I started using them, and that continues with AM5 across many CPUs (7800X3D, 9800X3D, 9950X3D). It's just what they're run at (mine and friends'), and under load it drops down. Though full disclosure, I do run my 9800X3D at a fixed voltage of 1.23V and 5425MHz.


It makes a pretty big difference even at 4K in games that are actually CPU heavy. Star Citizen is the one I can speak to, as I've upgraded over the years from an i9 10940X, to a 5800X3D (and the same RAM at the i9), 7800X3D, 9800X3D, and 9950X3D (all with the same RAM as the 7800X3D). Outside of the 9950X3D being a definite sidegrade for gaming, I saw significant average and minimum FPS gains with each upgrade.

Star Citizen is an exception, though, and not the rule. I've heard that MMOs like World of Warcraft is another. The majority of games will see minimal improvements at 4K unless there's other significant issues with the system (e.g. bad thermal paste application / cooler mounting). Although running at 4K but with DLSS performance mode can change that equation a bit, as it will hit the CPU more than standard 4K as it's a 1080P upscale.
I game at 4K with max settings(path-tracing when ever possible), if you are not setting your games up to be GPU limited, you should buy a console because you are doing it wrong.
Hence why you can upgrade your GPU 3-4 times on the same CPU with minimal issues, but then again I have seen..."people" who buy a 5090 to game a 1080p but hen again, idiots are more numerous geniuses so that is to be expected.
 
Sorry to hear about the cpu. Asus and AsRock have had issues, but haven't seen anything on the X670 variants. Hopefully that doesn't become an issue as well down the line.
 
  • Like
Reactions: erek
like this
I game at 4K with max settings(path-tracing when ever possible), if you are not setting your games up to be GPU limited, you should buy a console because you are doing it wrong.
Hence why you can upgrade your GPU 3-4 times on the same CPU with minimal issues, but then again I have seen..."people" who buy a 5090 to game a 1080p but hen again, idiots are more numerous geniuses so that is to be expected.
Lol. You obviously haven't played Star Citizen with the most powerful GPUs available across several generations. But that's OK, it's just weird that you'd blanket statement like that, saying people are doing it wrong.


Back to the OP's dead 9800X3D - unfortunately, CPU's do just die. Supposedly it's still about a 1.5% rate for X3D CPUs according to various retailers and system builders, but there has been quite the kerfuffle for Asrock and now Asus. I suspect a manufacturing issue with a QA problem over at AMD. I had two bum 9950X3Ds in a row right out of the box to put some perspective on it.
 
Last edited:
@ arestavo, I'm pretty sure littlegreenmenfromhome is here intentionally trolling. I don't know why some kids get a rise from this and make stupid comments like calling industry wide standards for measuring CPU performance stupid and telling people to go back to console gaming if they don't do what he says. Just ignore the troll and don't feed him anymore and maybe we can get back to having an intelligent conversation here.

As for OP, it could have been SOC voltage that killed your 9800X3D. Are you on the latest BIOS for your ASUS X670E? I'm on the same board and CPU as you and I recall them issuing BIOS updates that lowered SOC voltage to stay under 1.275v
 
@ arestavo, I'm pretty sure littlegreenmenfromhome is here intentionally trolling. I don't know why some kids get a rise from this and make stupid comments like calling industry wide standards for measuring CPU performance stupid and telling people to go back to console gaming if they don't do what he says. Just ignore the troll and don't feed him anymore and maybe we can get back to having an intelligent conversation here.

As for OP, it could have been SOC voltage that killed your 9800X3D. Are you on the latest BIOS for your ASUS X670E? I'm on the same board and CPU as you and I recall them issuing BIOS updates that lowered SOC voltage to stay under 1.275v
Do you play games at 720p or 1080p?
No?
Then benching at those resolutions will tell you nothing about performance at 2160p

Even tests will show this:
https://en.gamegpu.com/test-gpu/action-fps-tps/cyberpunk-2077-update-2-3-test-gpu-cpu
1771783377527.png


1771783428159.png


And drop the insults, I am +50 years old and focus on reading up instead.
 
Lol. You obviously haven't played Star Citizen with the most powerful GPUs available across several generations. But that's OK, it's just weird that you'd blanket statement like that, saying people are doing it wrong.


Back to the OP's dead 9800X3D - unfortunately, CPU's do just die. Supposedly it's still about a 1.5% rate for X3D CPUs according to various retailers and system builders, but there has been quite the kerfuffle for Asrock and now Asus. I suspect a manufacturing issue with a QA problem over at AMD. I had two bum 9950X3Ds in a row right out of the box to put some perspective on it.
You are talking about a "game" in alpha, allow me to LOL
 
Sorry to hear about the cpu. Asus and AsRock have had issues, but haven't seen anything on the X670 variants. Hopefully that doesn't become an issue as well down the line.
Lol. You obviously haven't played Star Citizen with the most powerful GPUs available across several generations. But that's OK, it's just weird that you'd blanket statement like that, saying people are doing it wrong.


Back to the OP's dead 9800X3D - unfortunately, CPU's do just die. Supposedly it's still about a 1.5% rate for X3D CPUs according to various retailers and system builders, but there has been quite the kerfuffle for Asrock and now Asus. I suspect a manufacturing issue with a QA problem over at AMD. I had two bum 9950X3Ds in a row right out of the box to put some perspective on it.
thinking about this motherboard if i have to,

https://wccftech.com/review/msi-mpg-x870e-carbon-max-wifi-motherboard-review-maxing-out-carbon/
@ arestavo, I'm pretty sure littlegreenmenfromhome is here intentionally trolling. I don't know why some kids get a rise from this and make stupid comments like calling industry wide standards for measuring CPU performance stupid and telling people to go back to console gaming if they don't do what he says. Just ignore the troll and don't feed him anymore and maybe we can get back to having an intelligent conversation here.

As for OP, it could have been SOC voltage that killed your 9800X3D. Are you on the latest BIOS for your ASUS X670E? I'm on the same board and CPU as you and I recall them issuing BIOS updates that lowered SOC voltage to stay under 1.275v
Latest BIOS
 
no symptoms or warning signs

we had some lightning rolling through and i gracefully shutdown my pc and switched the psu off completely and left it sit for a day

come back and try to power it on:

dram (yellow) for about a minute
cpu (red) for a few seconds

then stuck on vga (white) indefinitely

i stripped the build all the way down to just cpu, mobo, 1 stick of ram at a time

reseated cpu, reseated ram stick and even swapped ram stick

cleared cmos

bios flashback to latest version

and never gets passed solid white vga light

—-

went to my local best buy and bought a 9700x to troubleshoot with and not expecting much.. was thinking mobo is probably dead


…wrong

the rig POSTed and is now fully operational and within margin of error performance of the X3D

i tried swapping back in the 9800X3D, still doesn’t work


going to RMA
I just had a similar event with a 7900x and Asrock B650M Pro RS WiFi. PC is on a UPS and never showed any signs of issues. All I did was a reboot and it will not post anything, all I get is the CPU and RAM error MB lights. My CPU is out of warranty but the MB is not. I'll wager the CPU is dead.
 
  • Like
Reactions: erek
like this
I ran that motherboard for a while with no issues.
Me too, it's a great board and works like a charm on the latest BIOS.
the motherboard itself hasn’t croaked, which im pretty surprised that it fired right up with that replacement 9700X. i was hypothesizing that the motherboard was bad due to subpar VRM components and design since its mid tier

so not sure why my 9800X3D died without prior symptoms

i only had 1 out of the 2 cpu power cables hooked up to the motherboard
 
Back
Top