NVIDIA Advises Intel Raptor Lake CPU Users to Consult Intel for Stability Issues

Published by

Click here to post a comment for NVIDIA Advises Intel Raptor Lake CPU Users to Consult Intel for Stability Issues on our message forum
https://forums.guru3d.com/data/avatars/m/216/216349.jpg
Bad luck when someone buys an high end CPU that has stability issues...
https://forums.guru3d.com/data/avatars/m/239/239175.jpg
I get NetBurst vibes...
data/avatar/default/avatar03.webp
You should never release CPUs that cannot be fully stress-tested under prime95 with all the instructions enabled and without thermal throttling. How are you supposed to know your 14900K is fully stable if you cannot put it through p95 for 24hrs at its max boost speeds without some arcane cooling solution?
https://forums.guru3d.com/data/avatars/m/215/215813.jpg
The instability issues with raptor lake are being caused by motherboard manufacturers enabling their own AI Overclocking features by default. They are using their own settings and over-volting the CPU instead of having intel’s recommended settings enabled by default. There was a video on YouTube recently by Jayz2cents showing the exact issue. I’ve been running my 14900kf with Asus’s AI Overclocking feature disabled and it’s been stable since day one coz I’m using intel’s recommended settings instead.
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Uh.... why is Nvidia saying all this? That's like a car manufacturer telling you to ask your local government to fix potholes in the road. It's not that they're wrong, just rather odd they're the ones making the suggestion.
https://forums.guru3d.com/data/avatars/m/216/216349.jpg
schmidtbag:

Uh.... why is Nvidia saying all this? That's like a car manufacturer telling you to ask your local government to fix potholes in the road. It's not that they're wrong, just rather odd they're the ones making the suggestion.
It`s a marketing move, they want to make sure that their customers know that the problem is because of Intel CPUs and not because of their 80/90 cards. Of course they are throwing Intel under the bus with this move but such is life...
https://forums.guru3d.com/data/avatars/m/248/248291.jpg
schmidtbag:

Uh.... why is Nvidia saying all this? That's like a car manufacturer telling you to ask your local government to fix potholes in the road. It's not that they're wrong, just rather odd they're the ones making the suggestion.
Because people are complaining at the nvidia forums and tech support, but the it's not their fault. It's a problem with Intel.
https://forums.guru3d.com/data/avatars/m/80/80129.jpg
schmidtbag:

Uh.... why is Nvidia saying all this? That's like a car manufacturer telling you to ask your local government to fix potholes in the road. It's not that they're wrong, just rather odd they're the ones making the suggestion.
Well in that analogy the car owner is blaming the car manufacturer for the road being bumpy. Presumably with Intel already having a dedicated support page for it - they have some potholes to fix.
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
Horus-Anhur:

Because people are complaining at the nvidia forums and tech support, but the it's not their fault. It's a problem with Intel.
Denial:

Well in that analogy the car owner is blaming the car manufacturer for the road being bumpy. Presumably with Intel already having a dedicated support page for it - they have some potholes to fix.
Considering how long Nvidia's GPUs have been on the market (and therefore should be pretty solid by now) and how a CPU pushed to its limits is going to be unstable, you'd think this would be obvious, and quite easy to prove.
https://forums.guru3d.com/data/avatars/m/65/65088.jpg
I've undervolted my 13700K from the getgo. Haven't noticed any problems. Runs like a dream.
data/avatar/default/avatar14.webp
schmidtbag:

Considering how long Nvidia's GPUs have been on the market (and therefore should be pretty solid by now) and how a CPU pushed to its limits is going to be unstable, you'd think this would be obvious, and quite easy to prove.
It is not easy to prove when the silicon quality of 13900k/ks and 14900k/ks fluctuates a fair bit, a 13900k can be better then a ks and a 13900ks can be better then a 14900k and........................ Then there are more then 5 different motherboard brands that have 5 different "default" settings that rarely follow intel spec. Then there is the memory speeds some use on these chips. On top of that there is the cooling and power requirement when running 100% CPU load during shader buildup. (at least that is what I am seeing on my 7800X3D)
https://forums.guru3d.com/data/avatars/m/297/297804.jpg
r3claim3r:

I've undervolted my 13700K from the getgo. Haven't noticed any problems. Runs like a dream.
i limited my 13900K in BIOS. Temps dream, pulls max 190W under extreme load, which never happens while gaming. So i do for most PC´s i build and never experienced any issues.That one setting saying: CPU can pull up to 4096W in BIOS makes me laugh each time 😉
data/avatar/default/avatar30.webp
Just my luck... Got myself, new PC for work, 14900k instead of AMD because of potential issue with DDR5 mem stability (mostly hear about that from one YT persona complaining and then some more on net). Also I thought 14x series is basically refresh of 12 and 13x series, that thing has to be rock solid and stable by now... At the get go I was unable to run GSKILL @ 6800 due to memory issue (memtestx86 detected memory errors) got myself second pair and that one worked fine. Using custom liquid cooling, silent profile oriented build, running that chip without any limits and ... compiler errors during, well, compilation. That usually indicates memory issues. Other errors during compilation caused by some CPU related issues. Recent BIOS and running that CPU @ stock resolved issue for me, at least for now as I didn't noticed any crashes. Why oh why did I go with Intel ... (was running 3960x for 4-5years before that, no issue whatsoever stable as rock, and that with 2 non matching RAM kits 4 x 8GB and 4 x 16GB). The QA nowadays is just as joke. Same for GPU. Used NV GPUs for like 20 years now, not even once did I encountered crash related to GPU (driver or otherwise) during that period, after getting myself 4090 ended up with scenario that system hung after using vulkan application ...
https://forums.guru3d.com/data/avatars/m/297/297804.jpg
RavenMaster:

The instability issues with raptor lake are being caused by motherboard manufacturers enabling their own AI Overclocking features by default. They are using their own settings and over-volting the CPU instead of having intel’s recommended settings enabled by default. There was a video on YouTube recently by Jayz2cents showing the exact issue. I’ve been running my 14900kf with Asus’s AI Overclocking feature disabled and it’s been stable since day one coz I’m using intel’s recommended settings instead.
I saw that video too cause Jayz2cents shows things which other´s not even mention about. Had once an issue with an 13700K on an ASUS Z690 MoBo which oced the CPU automaticly. Insane temps and crashes were the result.
https://forums.guru3d.com/data/avatars/m/79/79740.jpg
r3claim3r:

I've undervolted my 13700K from the getgo. Haven't noticed any problems. Runs like a dream.
Same with 13600k. Over a year of running with rock solid stability. I agree with others saying default mobo settings are ridiculous and needlessly over-volt when left on auto.
https://forums.guru3d.com/data/avatars/m/246/246171.jpg
TLD LARS:

It is not easy to prove when the silicon quality of 13900k/ks and 14900k/ks fluctuates a fair bit, a 13900k can be better then a ks and a 13900ks can be better then a 14900k and........................ Then there are more then 5 different motherboard brands that have 5 different "default" settings that rarely follow intel spec. Then there is the memory speeds some use on these chips. On top of that there is the cooling and power requirement when running 100% CPU load during shader buildup. (at least that is what I am seeing on my 7800X3D)
It is very much easy to prove if your CPU is the problem - run memtest86, prime95, cpuburn, or a few cinebench runs. If it ever crashes, then you've just narrowed down that either the CPU or RAM has a problem. Perhaps your PSU is the problem too. From there, start underclocking a bit to see if the problems go away. EDIT: Even if your CPU's silicon quality isn't the problem, so long as those CPU stress tests fail, what matters is the GPU is not the problem and that underclocking (and undervolting) are likely to solve your problem. If those apps don't crash then run GPU stress tests like Furmark or Gravitymark and check for artifacting or crashing.
https://forums.guru3d.com/data/avatars/m/198/198862.jpg
Has anyone even being remotely surprised by this? Intel cpus are pushed to the max. Core clock, voltages, power usage its all on the edge of not being stable enough for a daily use or gaming. Like Steve from HU said it reminds of Pentium vs Athlon and Intel desperate ways to compete.
https://forums.guru3d.com/data/avatars/m/246/246088.jpg
Undying:

Like Steve from HU said it reminds of Pentium vs Athlon and Intel desperate ways to compete.
Yeah, then they came back with the C2D architecture 😛
https://forums.guru3d.com/data/avatars/m/198/198862.jpg
pegasus1:

Yeah, then they came back with the C2D architecture 😛
True. In this situation they need another C2D becouse Zen5 is coming.