PSA for 13th/14th gen Intel CPU owners

[dfonda]

Intel has found the problem their fix will be coming mid-august in the form of microcode.

July 2024 Update on Instability Reports on Intel Core 13th and 14th Gen Desktop Processors

*Update 7/29 regarding customer support process guidance (included below as well): https://community.intel.com/t5/Processors/Clarification-Update-on-Intel-Core-13th-14th-Gen-Desktop/m-p/1618462 Based on extensive analysis of Intel Core 13th/14th Gen desktop processors returned to us due to...

community.intel.com

Intel Addresses Desktop Raptor Lake Instability Issues: Faults Excessive Voltage from Microcode, Fix Coming in August

www.anandtech.com

Intel finally announces a solution for CPU crashing and instability problems — claims elevated voltages are the root cause; patch coming by mid-August [Updated]

Patch is on the way.

www.tomshardware.com

Speculating that it is anything else is just that speculation. I am aware that one of the YouTubers has published a video claiming that there could be a manufacturing defect related to "Oxidation" of copper through silicon vias. A member of Intel staff has commented on this on Reddit. What they said was "...We can confirm that the via Oxidation manufacturing issue affected some early Intel Core 13th Gen desktop processors. However, the issue was root caused and addressed with manufacturing improvements and screens in 2023. We have also looked at it from the instability reports on Intel Core 13th Gen desktop processors and the analysis to-date has determined that only a small number of instability reports can be connected to the manufacturing issue".

 

[EarthDog]

Intel has found the problem they're fix will be incoming mid-august.

Awesome! Fix via.........?

Link(s)???

 

[dfonda]

Awesome! Fix via.........?

Link(s)???

Sorry I was on my phone and tied up...adding links to my post. (needed my laptop)

I may have to backtrack. I wanted to make sure these were not the early false reports of a fix that were reported.

 

[Kenrou]

Awesome! Fix via.........?

"The company will issue a microcode update to address the issues by mid-August... will be distributed through BIOS updates from motherboard OEMs and via Windows updates, so the timing for end-user availability could vary."

 

[EarthDog]

"The company will issue a microcode update to address the issues by mid-August... will be distributed through BIOS updates from motherboard OEMs and via Windows updates, so the timing for end-user availability could vary."

LOL, thanks... I read that in the links when he added them.

 

[Flamethrower1972]

GN has a new video. Nothing at all confirmed as far as root cause, but they've been given some "tips" as to the potential issue. A new one that hasn't been discussed before is essentially that there is a manufacturing defect that went unnoticed starting with 13th gen CPUs Namely, the TaN layer not being applied properly and so the metal lines are quickly oxidizing, leading to increased resistance and early electromigration (degradation) problems. If true, Intel should be able to resolve the issue in the fab but the problem is, that also means that tens of millions of CPUs are effected. Since these are effecting the highest value SKUs the most, you are looking at billions of dollars worth of damaged goods. The scope of the problem is further confirmed by OEM contacts GN talked to, who reported between 10% - 25% failure rates just during their build time QA checks. The large range is believed to be due to the difference in how thorough the QA checks are between OEMs. In terms of raw numbers, one large OEM said they believe that 8 million CPUs that they've received from Intel are problematic. Add to that that all of their OEM contacts are reporting issues, and the total number is going to be quite large. At least one of their OEM contacts also told them that it's not just i9 CPUs effected, but even i5 K series processors are showing issues too (e.g., 13600K), but it is to a lesser extent than the high tier SKUs.

Again, nothing in terms of root cause has been confirmed. To try and verify the manufacturing flaw theory, GN is looking to send multiple degraded CPUs for analysis. This is a slow and costly procedure, so GN is not sure how many CPUs they'll be able to afford to send, but they will send as many as they can once the final quote comes back. They are also asking that anyone with a degraded CPU to fill out some info in a spread sheet they have posted online to see if they can narrow down a manufacturing date range of CPUs that are effected.

You know what this guy does a great job to an extent but sometimes I am just sick of hearing what he has to say IMO.

 

[]-[itman]

Intel has found the problem their fix will be coming mid-august in the form of microcode.

July 2024 Update on Instability Reports on Intel Core 13th and 14th Gen Desktop Processors

*Update 7/29 regarding customer support process guidance (included below as well): https://community.intel.com/t5/Processors/Clarification-Update-on-Intel-Core-13th-14th-Gen-Desktop/m-p/1618462 Based on extensive analysis of Intel Core 13th/14th Gen desktop processors returned to us due to...

community.intel.com

Intel Addresses Desktop Raptor Lake Instability Issues: Faults Excessive Voltage from Microcode, Fix Coming in August

www.anandtech.com

Intel finally announces a solution for CPU crashing and instability problems — claims elevated voltages are the root cause; patch coming by mid-August [Updated]

Patch is on the way.

www.tomshardware.com

This is the communication @mackerel posted about yesterday. I am still a bit weary that this will indeed be the final fix until we see some independent testing, but hopefully this does solve the issue.

 

[]-[itman]

New Buildzoid video where he monitors voltages on a 14900K with an oscilloscope. This isn't with the new microcode Intel said they will release in August, but is the previous one that Intel said fixed the TVB voltage problem. It's a very long video but the summary is that the CPU is requesting 1.5+ V to reach its boost speeds. He believes that anything over 1.4 V is too high and damaging to the CPU. His speculation is that in order to bring the voltages down to safe values, Intel would have to drop boost speeds and lose performance, so instead, they will be dropping the voltage with the new microcode just enough to get most of the CPUs to last past their warranty period. Basically drop the voltage enough to slow, but not stop, the damage so that Intel won't have to do a mass recall/RMA program.

We obviously won't know until the new microcode drops and lots of testing happens, but it paints a pretty grim picture of the situation. Hopefully he's wrong and Intel has a better fix in the works.

 

[mackerel]

I saw that earlier but aborted when I saw how long it was. Could have used a summary version.

Turbo has always been a bit of an awkward area of CPU specs, since it is a maximum, not minimum, and is not guaranteed to be hit as it is opportunistic and depends on the right conditions being present. I guess you could argue it should be able to be hit in some load condition. So my question then is, what is the best case load to achieve the advertised turbo clocks? On 14900k, 6.0 is TVB so there is an additional thermal element over the Turbo Boost Max 5.8 GHz turbo - for selected best core(s) only. Looking up the boost table for 14900k, it looks like it is 6.0 up to 2 cores, then 5.7 beyond that (P-cores only) - thought it should be 5.6 based on spec but I've seen 2 references to 5.7. What voltage is needed to hit that with best case load? Normally I'd say if you're loading all cores, power limit would kick in, but for all but heaviest loads the 253W limit still gives each core a large bucket of power allocation.

If the higher voltage is genuinely needed to reach those higher clocks, then it could reduce the frequency those clocks may be attained. Intel's statement could also be interpreted as excessive voltage beyond needed was applied. If so, and they're just correcting for that, there should be no reduction in performance, or even potentially an increase as the reduced power draw may allow more boosting.

I did have a think about potentially affected processors. Vast majority of reports do seem to be for the '900 parts, which hit the highest speeds, and presumably need the highest voltages to do so. '700 tier and below are far less frequent. While all 14th gen is potentially at risk, lower 13th gen (13600 non-k and lower) were rebadged Alder Lake so might be immune. 13900/14900 variations make up about 1.8% of Steam Hardware Survey, and 14700 makes up 0.8%, all out of Intel's 66% share. These were more easily identifiable due to their relatively unique core counts. Assuming HEDT/WS/Server parts of the same core count are insignificant. That's why I can't break out 13700 as 16 cores share with many AMD CPUs too for example. It is a small but not insignificant proportion that could be affected.

My gut feeling is that lower tiers, '600 and lower, have the lowest risk exposure and they're likely to be the most popular. So when the new update drops then in most cases I'd guess any further significant degradation will be halted and they'll be fine for the rest of their normal life. The biggest pain will be for the '900 and maybe '700 chips.

 

[Viper69]

Intel has found the problem their fix will be coming mid-august in the form of microcode.

July 2024 Update on Instability Reports on Intel Core 13th and 14th Gen Desktop Processors

*Update 7/29 regarding customer support process guidance (included below as well): https://community.intel.com/t5/Processors/Clarification-Update-on-Intel-Core-13th-14th-Gen-Desktop/m-p/1618462 Based on extensive analysis of Intel Core 13th/14th Gen desktop processors returned to us due to...

community.intel.com

Intel Addresses Desktop Raptor Lake Instability Issues: Faults Excessive Voltage from Microcode, Fix Coming in August

www.anandtech.com

Intel finally announces a solution for CPU crashing and instability problems — claims elevated voltages are the root cause; patch coming by mid-August [Updated]

Patch is on the way.

www.tomshardware.com

I've read through all the posts here to bring me up to speed. To be clear Intel thinks OR KNOWS the root cause? 2. They KNOW they have a fix coming out in August, or they THINK it will fix the issue?

 

[dfonda]

I've read through all the posts here to bring me up to speed. To be clear Intel thinks OR KNOWS the root cause? 2. They KNOW they have a fix coming out in August, or they THINK it will fix the issue?

No they be speculating.

 

[]-[itman]

I've read through all the posts here to bring me up to speed. To be clear Intel thinks OR KNOWS the root cause? 2. They KNOW they have a fix coming out in August, or they THINK it will fix the issue?

I would say, thinks, or at least it's unconfirmed. They were very slick with their wording by saying that RPL had a problem with excess voltage and that the update will address the root cause of the excess voltage problem. Note that they didn't say that the excess voltage problem is the root cause of the instability. So, until the update actually is available and we get some very thorough, independent testing results, we just have to be in wait mode.

 

[Viper69]

I would say, thinks, or at least it's unconfirmed. They were very slick with their wording by saying that RPL had a problem with excess voltage and that the update will address the root cause of the excess voltage problem. Note that they didn't say that the excess voltage problem is the root cause of the instability. So, until the update actually is available and we get some very thorough, independent testing results, we just have to be in wait mode.

So I haven’t had issues yet, should I update when available regardless of the what the independent testing tells us? I remember when updates eons ago did help. Now many times updates do more harm than good with electronics it seems.

 

[Woomack]

Update to the latest BIOS, as it fixes the issue in some way (not fully, but on some brands, it sets slightly different power limits), and wait for the new version with an updated microcode.

On the other hand, this issue has been with us for two years, and somehow, people have been pushing these CPUs to the limits without significant problems (except for overheating). Until earlier this year, barely anyone noticed any problems. This just suggests that the real issue doesn't affect many CPUs or is not visible during work. If not for the noise on the web, I bet 99.5% of users would be happy, not knowing there is any issue.
There is also the fact that during full load, most 13/14 K series CPUs throttle during high load = run at lower (read safe) voltage to keep the low enough temps. So, as it sounds funny, if you have a bad cooler, the CPU runs slower but shouldn't show instability.

 

[Flamethrower1972]

He is my problem. I already bought the Gskill RAM and its speed is 7000MHz. I have thought about buying the i5-14600k but the fastest ram speed that goes up to is 5000MHz. So I think I should buy the i7-14700k. Now keep in mind that I am not buying the CPU until sometime in 2025 and the i5 is $100 cheaper. What should I do? And please don't tell me to go AMD cuz I do not use AMD products.

 

[Woomack]

14600K and 14700K have the same memory controllers. I haven't seen even one CPU that couldn't make 7200+.
Also, in about 2-3 months, new Intel CPUs and motherboards will be available (assuming there won't be any delay).
You can also get AMD

 

[Flamethrower1972]

14600K and 14700K have the same memory controllers. I haven't seen even one CPU that couldn't make 7200+.
Also, in about 2-3 months, new Intel CPUs and motherboards will be available (assuming there won't be any delay).
You can also get AMD

So save $100 and get the i5-14600k and with XMP I can get to 7000MHz???

 

[Woomack]

I see you are the one who doesn't care about the Intel CPU issues mentioned in this and some other threads

My old 13600 non-K ES could run with RAM at 7600 stable. The worst chip I had was actually one of the 14900 Ks that could work at 7400 max, but I'm not sure if it was stable, as I returned it when I noticed it couldn't even post at 7600+.
7000-7200 seems guaranteed to work stable, but the above is very random. This is where AMD comes as every Ryzen 7000 that I tested could make 8000

 

[Flamethrower1972]

I see you are the one who doesn't care about the Intel CPU issues mentioned in this and some other threads

My old 13600 non-K ES could run with RAM at 7600 stable. The worst chip I had was actually one of the 14900 Ks that could work at 7400 max, but I'm not sure if it was stable, as I returned it when I noticed it couldn't even post at 7600+.
7000-7200 seems guaranteed to work stable, but the above is very random. This is where AMD comes as every Ryzen 7000 that I tested could make 8000

I just want to get to the RAMS speed of 7000MHz that's all and I bet I can reach that in the XMP in the BIOS.

 

[EarthDog]

Update to the latest BIOS, as it fixes the issue in some way (not fully, but on some brands, it sets slightly different power limits), and wait for the new version with an updated microcode.

On the other hand, this issue has been with us for two years, and somehow, people have been pushing these CPUs to the limits without significant problems (except for overheating). Until earlier this year, barely anyone noticed any problems. This just suggests that the real issue doesn't affect many CPUs or is not visible during work. If not for the noise on the web, I bet 99.5% of users would be happy, not knowing there is any issue.
There is also the fact that during full load, most 13/14 K series CPUs throttle during high load = run at lower (read safe) voltage to keep the low enough temps. So, as it sounds funny, if you have a bad cooler, the CPU runs slower but shouldn't show instability.

QFT!! Mid August was the last I heard.




That said, let's stick to the subject. This isn't about ram speed, but problems with Intel processors.