About 14th intel CPU's quality issues

[xeonhsu]

I am a 14900K user. Since I bought the CPU last November, I have replaced a CPU due to CPU failure. However, the CPU after replacement failed again.

I would like to ask if the process problem has been solved. Otherwise, what is the point of applying for RMA again to exchange for a CPU that can only be used for a few months? I still have to assemble it.

I disabled a performance core and continued to use it.

The 14900K with 7 performance cores is even weaker than the i7 processor.

Please tell me if you have solved the process problem.

[Westfield60]

You are absolutely correct. BIOS updates have now begun to be released (MSI released their today) so if you have received an RMA replacement, or purchased a new Intel prebuilt or purchased a new CPU you should minimize use until you can apply the BIOS and microcode updates. These are supposed to prevent further degradations by adhering to Intel specs, but also to stop degradation on any newer CPUs that are not showing any signs of degradation.  That is why a few people I know are waiting for the patches to be released before they send their damaged ones back. 

[Eisbar]

Which core did you disable?

Was it P-Core 6?

Failure rates of specific performance cores by APIC ID taken from over 100 user reports of failed CPUs.

• 0 - 0%
• 1 - 0%
• 2 - 16.7%
• 3 - 22%
• 4 - 39%
• 5 - 33%
• 6 - 50%
• 7 - 0%

Core arrangement is 1,3,5,7 and 0,2,4,6 

Core 6 is most likely to be defective on the RPL 8+16 die, followed by it's neighbour Core 4. 

Other than the anomalous core 7 failures decrease in distance from Core 6. 

Core 6 and core 7 each sit next to an e-core cluster.

Due to the cores being flipped on the opposite side the layout is reversed. Either Core 6, or Core 7 has its power gates in between it and an e-core cluster, the opposite core does not and will be sitting beside an e-core cluster with no space.  Making it the only point on the die where two cores are not separated by power gates and I don't think it's coincidence that the anomalous failure point seems to occur there and fan out.

I don't make semiconductors, this is non-technical information based on available data and conjecture. Intel has said they are going forward trying to turn their company around using honesty and transparency but as of yet they've not done that. So people are left to try and ascertain for themselves if they believe buying an Intel product is safe.