“I’ll never buy an Intel CPU at launch again.”
That’s what one Reddit user told me after I spoke with them about instability issues with their Core i9-13900K. The user said they went through four Core i9-13900Ks over several months, leaving them without a functioning gaming PC for three months due to repairs. Another told me they slowly replaced every component in their system expect for the CPU, saying they’ve “NEVER had an issue with Intel products.” This time, though, the Core i9-14900K was the problem.
Intel needs to address the instability on 13th-gen and 14th-gen CPUs. We’ve seen BIOS updates, comments from PR managers, and documentation shared with motherboard vendors. But we’re reaching a point where Intel needs to publicly address the problem, as a growing number of users — both personal and business — are speaking out about the instability, with some going as far as to call Intel’s high-end chips “defective.”
The problem is getting worse
Let’s back up. The first reports of instability on the Core i9-13900K and Core i9-14900K made the rounds late last year, but it wasn’t until a report out of Korea that the issue picked up steam. Apparently gamers were returning these CPUs in droves due to crashes in Tekken 8 and other Unreal Engine games. That prompted Intel to work with motherboard vendors to release the Intel Baseline Profile, which changed a load of BIOS settings for stability. In testing, some outlets found this new profile could lead to up to a 9% drop in performance.
The discussion died down, but it’s picking up once again. YouTube channel Level1Techs released a video investigating the source of crashes and found a lot of uncomfortable data. One developer the YouTuber spoke to said that they might lose over $100,000 in lost players due to multiplayer server crashes on servers running the Core i9-13900K or Core i9-14900K. Sources from Dell, Lenovo, and HP also told the YouTuber that they believe around 10% to 25% of these chips have issues, while Level1Techs speculated that number might be closer to 50%.
The crash rate is what really stands out. Over a 90-day period, Level1Techs found 1,584 crashes due to a decompression error in an unnamed game developer’s crash database. Out of that, 1,431 of the crashes were from the Core i9-13900K and Core i9-14900K. Keep in mind this is for all crashes in this 90-day period. For context, AMD’s CPUs in totality represented only four errors in the database over this period. Before you think that Intel is just overrepresented, Level1Techs found that AMD CPUs make up about 30% of CPUs in the data, despite representing only 0.25% of crashes. Just the Core i9-13900K and Core i9-14900K alone — ignoring all other Intel CPUs — make up over 90% of the crashes.
It seems the problem is only getting worse, too. “For the players that play consistently, the number of errors they’re encountering with their system are definitely increasing over time for [the Core i9-13900K and Core i9-14900K], I can say that,” Wendall from Level1Techs said.
“…we discovered a pattern: almost all [crashes] were coming from systems with 13th and 14th generation Intel processors.”
Wendall also pointed out that this rate of crashes suffers from some survivorship bias. It doesn’t account for things like hard system crashes when you get a Blue Screen of Death (BSOD). This data is just for one error, too. We’ve seen numerous other errors, such as the infamous error of a GPU being out of memory, that’s hard to attribute to the CPU in a crash database.
The story up to this point with instability has been focused around power, and specifically how motherboard vendors are feeding too much power to the Core i9-13900K and Core i9-14900K. Wendall’s investigation disagrees with that assertion, however. In speaking with game developers who had these CPUs deployed in multiplayer servers, Wendall saw that 50% of the servers using either one of these CPUs have suffered from instability. That’s also when using the conservative workstation-class motherboard chipset, which doesn’t feed unnecessary power to the CPU.
This is the straw that’s breaking the camel’s back. A data center provider told Wendall that “support incidents have been unusually high for that configuration,” speaking on the Core i9-13900K and Core i9-14900K. Because of that, this provider charges over $1,000 more for support on Intel systems compared to AMD. Game developers are now speaking out on their own with issues, too.
As mentioned, Alderon Games went as far as to call Intel’s CPUs “defective” and said it would be transitioning all of its multiplayer servers to AMD. Alderon Games makes Path of Titans, which isn’t the most popular game. Maybe it’s just a bad batch? That doesn’t look like the case. Warframe is in the top 50 most played games on Steam — No. 29 at the time of writing — and regularly has over 50,000 concurrent players just on Steam alone (the game is available on other PC storefronts, as well as consoles).
“After aggregating hundreds of reports from helpful players we discovered a pattern: almost all [crashes] were coming from systems with 13th and 14th generation Intel processors,” one Warframe developer wrote in a forum post.
It’s officially widespread
We’re beyond a small group of users who have unstable overclocks on their systems. This is officially a widespread problem that affects users who are doing everything by the book. I mean, consider that Epic Games has a dedicated support page for crashes in Fortnite related to these two CPUs. You don’t see that very often.
Intel has failed to get in front of this problem, and it’s growing to be too big to ignore. The biggest guidance we’ve seen so far is a forum post made last month by Intel’s Thomas Hannaford — the PR manager for Intel, if that says anything. That’s despite Intel promising it would make a statement in May when these issues started popping up earlier this year. And the guidance here is for users to update their BIOS, use the default profile, and potentially suffer a performance loss.
But even then, the issue isn’t resolved. Intel itself has made it clear that the investigation is ongoing, and Wendall says that, although the BIOS updates help, they haven’t fixed the problem entirely. Every user I spoke to said that they weren’t overclocking their system, and even after BIOS updates, problems returned.
“This was one of the most frustrating issues to troubleshoot that I’ve ever experienced. There were so many variables to consider, including dozens of BIOS configurations. It does give me some pause about the next time I upgrade my PC, and I’ve been an avid Intel user for over a decade,” one user told me.
It’s not unreasonable to expect a fix from Intel after so many months. This is a complex, inconsistent issue, however. You can grant Intel some grace for not having a concrete fix given how many variables are at play. It’s hard for me to extend grace with the lack of communication, however, especially given how prevalent these problems still are.
A communication failure
Last year, AMD was caught up in a problem where some of its Ryzen 7000 CPUs could cause physical damage to the motherboard socket with improper power settings. The problem was resolved in two days. More importantly, AMD came out with clear guidance almost immediately and informed its support team to prioritize cases with users affected by the problem. Intel may have done the same internally, but it hasn’t communicated that.
Instead, the biggest revelations we’ve seen up to this point have come from leaked internal communications between Intel and its motherboard partners, independent investigations into performance issues, and sporadic guidance from everyone from Epic Games to the Oodle decompression tool. I can find a dozen independent troubleshooting guides online before finding any sort of official statement from Intel.
We’re reaching a breaking point. Intel isn’t, and it shouldn’t, rush to conclusions just to provide some reassurance to affected users. That could make the problem worse when it hasn’t identified the root cause. That doesn’t change the fact that Intel doesn’t seem to be providing much direct support.
We have little in the way of official statements, and according to Wendall’s investigation, even system providers are unclear about how to address the problem. There’s a serious communication issue here.
Here’s what I’d like to see from Intel: a dedicated page for instability on 13th-gen and 14th-gen CPUs. After months of secondhand sources and reports, users need a dedicated spot where they can go to see what the latest developments are and get support.
I’d also like to see an apology from Intel, not only for this issue in the first place but also the desperate lack of communication over the past several months. Intel’s forum post detailing these issues has nearly 50,000 views at the time of writing — even featured posts on Intel’s forums only pick up a few thousand — so this is clearly something that users are seeking out. Intel needs to make this information easier to find.
Ideally, this should happen immediately. After the video posted by Level1Techs, it seems that we’re getting more and more reports of fed-up customers — both for personal and commercial use — adding fuel to an already blazing fire.
Update 7/22/24: Since publishing this article, Intel has released a statement that pins blame on higher operating voltages due to faulty microcode. It says a microcode update will arrive in mid-August.
