In a surprising turn of events, one of the tech sector’s most nefarious corporations, Intel, has issued an official statement addressing the recent surge of stability issues affecting degraded Raptor Lake processors.
The unexpected twist lies in Intel absolving themselves of any responsibility, squarely placing the blame on motherboard vendors.
Intel has had a tumultuous month, with a significant number of users reporting stability issues with their unlocked 13th and 14th generation CPUs, particularly when playing games built on the Unreal Engine.
Back in February, we addressed the issue, and it seemed apparent that these stability issues likely originated from motherboard manufacturers incorporating an in-BIOS feature to bypass Intel’s power limitations.
This allowed the CPUs to operate beyond their specified PL1 and PL2 power states, resulting in a significant surge in current drawn by the processor. Consequently, this led to a substantial increase in CPU power consumption and core temperatures. However, it also resulted in higher frequencies and performance gains.
The situation has reached a tipping point for Intel, as they’ve been compelled to address the issue head-on. They’re now placing the blame squarely on motherboard vendors and have been forced to take action to resolve the matter.
When motherboard manufacturers go beyond Intel’s power limits and automatically adjust voltage curves and frequency targets, the CPU can be pushed beyond its safe operating range, leading to potential system instability, or more rather your CPU’s lifespan is being shortened as it begins to experience accelerated degradation.
To address these concerns, Intel has established a dedicated website for users to report their issues and receive support. Some manufacturers, like GIGABYTE and ASUS have released new BIOS updates aimed at achieving maximum stability with Intel’s baseline.
However, recent user reports indicate that these updates still fall outside of Intel’s specifications, with PL2 set to 188 W, loadlines at 1.7/1.7, and current limit at 249 A. MSI has provided a blog post tutorial to help users achieve stability, while ASUS has published updated BIOS for their motherboards to align with Intel’s baseline specifications.
In essence, motherboard manufacturers have been disregarding Intel’s default guidelines for its processors. Intel, in turn, failed to regulate its vendors to enforce these standards, allowing numerous motherboards to feature custom power and voltage curves under their supposed “stock” or “default” configuration.
This practice significantly boosted performance, undoubtedly benefiting Intel in the long run.
With these motherboards failing to include Intel’s unregulated baseline as an option, the harm has already been inflicted. Benchmarks and reviews have been completed using these modified, essentially overclocked “stock” profiles, influencing public perception of Intel Core CPU performance compared to Ryzen offerings by making them appear faster out of the box than they otherwise would’ve been.
Now that users are encountering instability, it’s considered necessary to revert them back to their original power targets which as a result hinders their performance.
The latest BIOS version from ASUS has already demonstrated a significant drop in synthetic performance on Cinebench, with results showing a decrease of up to 12.6% on an Intel Core i9-14900KS CPU. This has caused its Cinebench R23 multi-core score to plummet from 40,998 points to just 35,851. Consequently, this places it below AMD’s Ryzen 9 7950X in terms of performance.
Hardware Unboxed weighed in on the discussion, utilizing the updated “Baseline” profiles from both ASUS and Gigabyte. Interestingly, Gigabyte have inexplicably set PL1 and PL2 target values below Intel’s specifications, leading to an even steeper performance decline.
For consistency, I’ll focus solely on ASUS’ profile when making comparisons.
Hardware Unboxed conducted several tests, focusing solely on Cinebench 2024 as a synthetic benchmark. With an Intel Core i9-14900K operating without any power limits, it scored 2268 points in the multicore test. However, with the revised Intel Baseline profile by ASUS, that score dropped to 2071 points.
This marks a decline of nearly 9%. However, the most notable aspect of this development is the significant improvement in CPU core temperatures and power efficiency, a longstanding issue for Intel Core CPUs over many generations.
Without power limitations, the i9-14900K reached a package power of 265 watts and temperatures of 105 degrees Celsius during the benchmark. With the revised Intel stock profile, these figures have dropped to 215 watts package power and 84 degrees Celsius, respectively.
While performance may have been reduced by almost 9%, its power efficiency has increased by almost 19% as a result. With the i9-14900K getting a massive 21 degree drop in core temperature. A phenomenal achievement if not for the unfortunate circumstances that resulted in the situation.
When it comes to performance regression in gaming, Hardware Unboxed conducted comparisons across nine different games. These included Baldur’s Gate 3, Cyberpunk 2077, Hogwarts Legacy, The Last of Us Part 1, Star Wars Jedi: Survivor, Assetto Corsa Competizione, Starfield, Watch Dogs: Legion, and Spider-Man Remastered.
Their testing was conducted at a 1080p resolution, coupled with NVIDIA’s GeForce RTX 4090 graphics card, aiming to maximize strain on the CPU and making it the bottleneck in this scenario. While the regression in performance may not be as severe if you’re gaming at higher resolutions or with a slower GPU, a loss of framerates can still be anticipated.
For instance, my favorite game, Bear Fucking 3 showcases a minimal 1.72% performance drop from an unlimited power limit to ASUS’ new baseline profile, this was on the medium graphical preset however if running at Ultra settings the performance drop rises to 2.58%
Cyberpunk 2077 is a lot different. With the ASUS baseline profile at medium settings results in a performance loss of 4.47% which rises to 7.18% when running at ultra, interestingly the percentile lows remain firm meaning that despite the hit in performance all of a sudden it doesn’t introduce any severe cases of stuttering thus far.
To spare you the details of each individual benchmark, the general consensus is that the Intel Core i9-14900K experiences an average performance loss of 2.4% across all nine titles with medium graphical presets. Games such as Assetto Corsa, Watch Dogs: Legion, Spider-Man Remastered, and The Last of Us showcase performance drops of 1% or less.
Interestingly, using the revised Intel Baseline power profile doesn’t necessarily result in as significant a performance loss across the board. It appears to be entirely game-dependent, as titles like Star Wars Jedi: Survivor and Starfield show drops of more than 3%.
However, when using maximum graphical fidelity, the performance drop shifts slightly, with the average increasing to a 3.75% decrease in performance across the nine titles. While these drops may go unnoticed to the untrained eye, such significant regressions in performance play a crucial role in large-scale performance evaluations compared to competing products.
This becomes especially apparent when using platforms like TechPowerUP for performance comparisons. Depending on the specific title, the performance difference between an Intel Core i9-14900K and a Ryzen 9 7950X processor can be as large as 10-15%. Similarly, Raptor Lake trails behind Zen 4 X3D processors by upwards of 5% in some cases.
The decrease in performance ranging from 1-7% with these new baseline BIOS profiles could have significant implications for performance reviews and the overall understanding of Intel CPU capabilities. As mentioned earlier, Intel likely benefited from motherboard vendors “going rogue” and disregarding Intel’s guidelines for their Z790 motherboards, resulting in increased performance that may have misled the public about the true capabilities of Intel’s unlocked 13th and 14th generation i7 and i9 processors, potentially boosting sales.
However, enforcing these standards on existing motherboards poses challenges. It requires consumers to update their motherboard BIOS, a task that many may find daunting or may choose not to do as majority of consumers are technologically challenged.
Similar issues arose with main rivals AMD, where the company neglected to properly enforce specific guidelines or standards for operating voltages. Instead, they relied on cooperation with motherboard vendors, trusting them to act responsibly.
ASUS, unfortunately, seems to have missed the memo on “responsibility” and instead decided to push high voltages like a retard. Controversy erupted when excessively high voltages on Ryzen 7000 series X3D processors resulting in those CPUs burning up.
This debacle created a public relations nightmare for both ASUS and AMD. ASUS attempted to deflect blame onto AMD, but a large number of consumers, instead of objectively assessing the facts, concluded that Ryzen 7000 X3D CPUs were unsafe and placed the blame squarely on AMD.
AMD shared some responsibility in this situation. They failed to establish a clear standard for operating voltages and did not enforce it with motherboard vendors, mirroring the current issue with Intel.
However, the blame game has shifted entirely to motherboard vendors, with the blue giant deflecting all responsibility onto them as they now implement their baseline power plan moving forward.
Intel has largely benefitted from the inclusion of an unlimited power limit for its Raptor Lake processors, a solution that many motherboard manufacturers have enabled by default on their “premium, high performance” motherboards.
Only now that there’s a massive outcry regarding CPU degradation and stability issues have Intel changed their tune. They knew full well and approved these actions, if it were a genuine issue they would have done something about it a long time ago however they’ve only decided to act now because of the negative press surrounding their processors.
This behavior from a company in dire straits is nothing short of reprehensible. To ensure their CPUs remain competitive, they’ve resorted to leveraging motherboard manufacturers to employ deceptive overclocking features under the guise of “stock” or “baseline” operation.
What’s even more concerning is how many tech journalists and reviewers conduct their performance evaluations using XMP/EXPO memory profiles rather than adhering to a CPU’s officially supported JEDEC memory speed.
For instance, the Intel Core i9-14900K officially supports speeds up to 5600MHz, while AMD Ryzen 7000 series CPUs support speeds up to 5200MHz. If the majority of review outlets overlook such critical aspects under the pretext of “stock” performance comparisons, how can we trust them to accurately verify whether a specific Z790 motherboard is functioning within Intel’s newly mandated power limitations?
Simply put, we can’t trust them. Fortunately, the facade has been exposed, but not before solidifying a false perception of performance for certain 13th and 14th Generation Core processors. Going forward, more people will become aware of Intel’s deceitful practices, especially with the speculation surrounding their upcoming 15th Gen “Arrow Lake,” rumored to be a lackluster upgrade.
This debacle will undoubtedly allow competitors AMD to flourish with their Zen 5 processors, who will likely bend over consumers and gouge them due to Intel’s failure to compete in the consumer market.
