GamersNexus analyzes potential silicon “oxidation” issues, no thermal failures and more

By /

Instability issues with Intel’s 14th and 13th generation processors have existed for over a year, but while Intel has yet to come up with solid reasons for the issues, others have come up with potential causes of silicon degradation and crash issues with these chips.

Intel’s instability issues escalate from ‘software limited’ to possible silicon defects, official response is ‘WAIT’

Crashes, instability, and performance issues with Intel’s 14th and 13th generation processors have plagued several consumers to the point of excruciating, and the community is now determined to switch to alternatives such as AMD’s offerings.

Here’s a timeline of Intel’s 14th and 13th generation instability issues so far:

  • [Dec 2022] The first cases of “not enough video memory” problems on 13th generation Core i9 CPUs were reported as early as 2022. (Source: Reddit). This is a few months after the release of Intel’s 13th generation processors and RTX 40 GPUs.
  • [Throughout 2023] Several users have started reporting game crashes and issues in high-profile AAA titles such as Callisto Protocol and Hogwarts Legacy (among many others). Both tiles are based on the Unreal Engine (released in 2023) and include a shader compilation process at the beginning. There are currently thousands of such issues reported on the Steam and Reddit community forums.
  • [February 2024] A post by Sebastian Castellanos brings the issue into focus, highlighting a “disturbing trend” of stability issues affecting 14th and 13th generation CPUs.
  • [February 2024] Our editor along with various other technical outlets will start reporting the issue and bringing the matter to the mainstream media. It’s quick to point out that underclocking and undervolting the chips can lead to better stability.
  • [April 2024] NVIDIA has issued a formal statement that the stability issues are not related to its GPUs, but rather to Intel processors.
  • [April 2024] Intel announces investigation into reports of instability issues affecting 13th and 14th generation CPUs.
  • [April 2024] Motherboard manufacturers are quickly introducing “basic” power profiles in the new BIOS, which lower the power limits of high-end 14th and 13th generation processors, but also lead to significant performance drops.
  • [May 2024] Intel issues a statement advising motherboard manufacturers to ship “Intel Defaults” in the BIOS by default, rather than custom settings.
  • [June 2024] Intel discovered a bug in its eTVB microcode while investigating the root cause of the 14th and 13th generation instability issues. It won’t fix the instability issues. Board manufacturers are quick to roll out BIOS with a new microcode patch.
  • [July 2024] Intel denies RMA to HardwareTimes, which has already RMAed at least two CPUs that caused instability issues.
  • [July 2024] Wendell of Level1Techs and Steve Burke of GamersNexus are the first major tech media on YouTube to further highlight the instability issues and delve into more details such as OEM frustration over the matter and possible recalls.
  • [July 2024] GamersNexus is alerting to a potential “oxidation” flaw in certain batches of Intel’s 14th and 13th generation processors that can lead to instability issues.
  • [July 2024] Intel says the 14th and 13th generation mobile CPUs are not affected by the same instability issues as desktop CPUs.
  • [July 2024] Intel issues a statement confirming that it has found the root cause of the instability issues, which happens to be increased operating voltage, and will issue a new microcode fix to its partners by mid-August\
  • To be continued…

We’ve seen game studios like Alderon Games and Epic Games highlight the issue on their platform, along with tech content creators like Wendell of Level1Techs bringing awareness to the issue to their viewers.

Despite the problem spreading to the mainstream media, the Blue team failed to address the root cause. The company has been working with AIB and board partners to mitigate the issues while discovering a few others such as the eTVB bug but apart from that there has been no proper communication from the blue team which tells us two things, either the company is worried about overheating. resistance from their customer base (clients, partners, OEMS) or they want to drag the issue out as much as possible until something new comes out and people simply forget about it.

Now, GamersNexus has gathered statistics on how this issue affects 14th and 13th generation Intel consumers, and according to one “unnamed” Intel customer, they’ve seen instability issues between 600,000 and 2 million CPUs. This only includes 13th generation units and information regarding 14th generation SKUs is currently unavailable. Interestingly, one Intel customer disclosed that the affected units have production dates from March 2023 to April 2024, which includes more than 12 months of retail SKUs spinning in the markets experiencing the issue.

From what we’ve heard, 1/3 of all Intel Raptor Lake processors that have shipped are Core i9-13900K or 14900K units, so that’s roughly 40-60 million units (Mike Bruzzone’s estimate). If that’s the case, then Intel could be facing a huge recall that would end up being a major disaster for the company and could be a potential reason why it’s taking its time to respond to the community.

Our editor Hassan reported in a post on X a while ago that he started facing these issues in early 2023, just a few months after the release of 14th gen desktop CPUs. Even though the BIOS mitigation made things a bit stable for it, it’s easy to say that applying the current “performance cap” patch will reduce the performance of your chip from what you originally had.

For the more interesting bits, GamersNexus compiled possible reasons for the instability issues based on internal Intel documents and information from customers. Newly surfaced information claims that Team Blue may have faced a “manufacturing” issue with the affected 13th and 14th generation chips, where insufficient “anti-oxidation” was applied to the SKUs, causing the processors’ electrical connections to break.

Credit: GamersNexus

This reasoning makes some sense considering that capping performance levels didn’t solve the problem at all, despite Intel releasing the relevant microcodes. While we won’t go into how the oxidation affected the functionality of the CPUs (see the GamersNexus video below for details), to sum it up, it could have affected individual layers, so the solution does not lie in any kind of mitigation at the software level.

It is important that the “performance limiter” issue does not drive this story. It’s not a “board performance limitation” issue. It’s a chip problem and always has been. The performance throttling issue was resolved using microcode. We have no idea if it will hit Meteor Lake or not yet. The current possible affected processors are [about] 8 million shipped that we know of.

If you disable Turbo Boost, you can get ‘stability’ until corrosion/contamination causes the CPU to fail. We have reports of some CPUs not even booting without blue shielding because the contamination/corrosion is so bad.

– Big Intel customer for GN

Well, what’s next? According to GamersNexus, vendors are looking for a middle ground, with some moving towards limiting the clock speed to 5.3-5.5GHz at the OEM level, while others are waiting for Intel to come up with a solution. Intel has begun working with suppliers and has indicated they will provide replacements for affected CPUs, and there are also rumors of a potential “major” recall, but nothing is certain yet.

Also below is the failure rate segregated by Intel SKU. interestingly, there are no thermal failures, suggesting that the instability problem is probably something more complex.

Meanwhile, leaker @Jaykihn also states that reports of oxidation being the cause of problems with Intel’s 13th and 14th generation CPUs seem unlikely. It also states that the report does not fit because it has statistics of Intel 7 chips (process node) tested as of June 2024.

Ian Cutress for More Than Moore also considers possible reasons for these problems:

We have issued extensive instructions on how to resolve the issue based on Intel’s guidance along with suggestions from third party sources; therefore, you can check them if you haven’t implemented the solutions mentioned. Additionally, we talked to various board partners about this and they said that they are being very cautious and spending more time testing and evaluating clock and power behavior, not only for existing chips, but also for upcoming CPUs like Arrow Lake.

For now, we have to wait and see how Team Blue handles the situation. With Arrow Lake-S desktop CPUs just around the corner, the whole fiasco is getting interesting and unfortunate at the same time.

Intel needs to say something: Oxidation requirements, new microcode and benchmarking challenges

Intel has a pretty big problem

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top