Intel Bartlett Lake CPUs Misreported at 7GHz on Linux: The P-State Bug Explained

Since the launch of Intel's Bartlett Lake P-core-only processors for the embedded market, Linux users have noticed an odd discrepancy: the Intel P-State driver shows a clock speed exceeding 7 GHz. However, the actual maximum turbo frequency for the flagship Core 9 273PE is 5.7 GHz. This guide explores the bug, its implications, and what it means for users. Below are the most common questions surrounding this issue.

1. What exactly is the reported bug with Intel Bartlett Lake CPUs under Linux?

The bug involves the Intel P-State driver—a power management and frequency scaling module used on Linux systems—mistakenly reporting that Bartlett Lake processors can reach clock speeds of 7.0 GHz or higher. In reality, the hardware's maximum turbo frequency is capped at 5.7 GHz. This misreporting is purely a software error; the CPU does not actually run at 7 GHz under any condition. The driver appears to misread or miscalculate the frequency scaling data, leading to inflated numbers in system monitoring tools that rely on P-State information. While it may initially seem like a performance windfall, the reported 7 GHz has no physical basis and should not be taken as a sign of actual speed improvements.

2. Which specific processor exhibits this 7 GHz misreport?

The issue was first identified on the Intel Core 9 273PE, a high-end model within the Bartlett Lake family designed for embedded applications. This processor is part of Intel's P-core-only lineup, which omits the hybrid architecture (E-cores) found in consumer chips. While the Core 9 273PE can boost to 5.7 GHz under optimal conditions, the P-State driver incorrectly labels it as a 7 GHz chip. It is likely that other Bartlett Lake SKUs may also be affected, as the bug seems rooted in the driver's handling of the P-core-only design rather than a single model. Linux users with any Bartlett Lake CPU are advised to verify frequencies independently using tools like turbostat or cpufreq-info rather than trusting the P-State reading.

3. Why does the Linux P-State driver incorrectly report such high frequencies?

The root cause is thought to be a misinterpretation of frequency scaling data by the P-State driver. Bartlett Lake processors use a different internal clocking scheme compared to previous Intel architectures—especially those with hybrid cores. The driver might be reading a raw frequency multiplier or base clock value without applying the correct ratio or offset suitable for these new chips. Additionally, firmware-level tables (ACPI) provided by the motherboard could contain placeholder or ambiguous entries that confuse the driver. Intel's embedded platform also may lack the same level of Linux optimization seen in consumer or enterprise CPUs. Until a patch is released, users can mitigate the issue by switching to the acpi-cpufreq governor or manually capping the maximum frequency via sysfs.

4. Is there any real performance benefit from this erroneous reading?

No—the inflated 7 GHz reading is purely cosmetic and does not reflect actual CPU performance. The processor's physical silicon cannot sustain or even reach that frequency due to thermal and power constraints. In practice, the maximum turbo frequency remains 5.7 GHz, as verified by benchmark tools and hardware monitoring. The bug only affects the reported value in software like cpupower or i7z; it does not alter the CPU's voltage, frequency scaling algorithms, or real-world speed. Therefore, users should not expect any performance gains. On the contrary, relying on the erroneous reading could lead to misguided overclocking attempts or misconfigured cooling systems, potentially causing instability or damage. It's safest to treat the P-State displayed value as a known glitch and rely on third-party frequency monitors or direct register reads for accuracy.

5. How can Linux users verify the true clock speed of their Bartlett Lake CPU?

To confirm the actual operating frequency of a Bartlett Lake processor, use tools that bypass the P-State driver. The turbostat utility (part of the Linux kernel tools package) reads hardware counters directly and reports frequencies per core based on actual cycles. Similarly, cpufreq-info from cpufrequtils shows the current frequency and available governors. For a low-level approach, check the /proc/cpuinfo file—while it often reflects the scaling driver's value, it can still be informative if compared with turbostat. Additionally, running a stress test like stress-ng while monitoring with s-tui or htop will display real-time speeds. If all these tools confirm max frequencies around 5.7 GHz, the P-State reading of 7 GHz is definitely erroneous.

6. Are there any existing patches or workarounds for this P-State bug?

As of the latest kernel releases, no official patch has been committed to fix the Bartlett Lake frequency misreport. However, Linux kernel maintainers are aware of the issue, and a fix is expected in a future update. In the meantime, users can apply a simple workaround by switching the CPU frequency scaling driver from intel_pstate to acpi-cpufreq. This can be done by adding the boot parameter intel_pstate=disable in the GRUB configuration. Restarting the system will then use ACPI-based frequency scaling, which typically reports correct frequencies. Another option is to manually set the maximum frequency via sysfs: echo 5700000 > /sys/devices/system/cpu/intel_pstate/max_perf_pct (if using intel_pstate). While not a permanent solution, these steps eliminate the confusing 7 GHz reading until a proper patch is released.