What steps should be taken to guarantee X3D+EXPO+PBO is executed securely and accurately?
What steps should be taken to guarantee X3D+EXPO+PBO is executed securely and accurately?
Check your assumptions before enabling PBO and EXPO on this setup:
CPU: AMD Ryzen 9 9950X3D
OPN number: 100-000000719
Batch number: CF 2536PGY
RAM: Kingston Fury Beast Black 64GB 6000Mhz CL36-44-44 (KF560C36BBE2K2)
Mobo: ASUS ROG Strix X870-E Gaming Wifi
PSU: Be Quiet! Dark Power 14 1200W Titanium (ATX 3.1)
AIO: Be Quiet! Silent Loop 3 420mm (top mount)
GPU: ASUS Astral ROG RTX 5080 OC
Case: Fractal Define 7 XL (2x140mm front, 1x140mm back)
OS: Debian Linux
I need to verify my understanding is correct. Before proceeding, let me clarify a few points:
1) The motherboard appears to have 22 VRM phases—18 for Vcore and 4 others. In the image you shared, it seems there are 20 identical-looking MOSFETs. Does this indicate 18 genuine phases and two fake doubler (2x2) phases?
2) The specifications mention 18x110A for Vcore, 2x110A for SoC, and 2x80A for other components. Based on this, I assume the SoC receives doubled or 'fake' phase supply.
3) According to the specs, the rig should comfortably support PBO and EXPO without excessive voltage spikes.
4) The PSU is of high quality, and the VRM configuration seems adequate, so stable power delivery should be assured.
5) The SoC voltage max is 1.3V, and it should operate fine at 1.2V or lower.
6) VDDIO/memory controller voltage should stay under 1.4V; running it safely below that is feasible.
7) Only the CPU core voltage needs adjustment if PBO is enabled.
8) The build was updated via USB flashback port before installing CPU/RAM, and everything started without issues. I disabled PBO and overclocking initially, then re-enabled EXPO after troubleshooting.
9) After a few weeks of use, temperatures stayed normal, and performance remained stable during gaming—GPU fans only slightly increased.
10) All 32 logical cores ran smoothly under heavy text processing; no crashes or warnings.
11) I noticed ASUS has had past issues with default motherboard settings, so this cautious approach makes sense given the recent concerns.
12) I’m still learning, but based on the info and testing, it seems safe to proceed cautiously.
Please confirm if you need any further clarification.
If I were in your position, I would have switched to a better brand like MSI or Gigabyte. The problem started with ASRock motherboards and now appears on Asus ones, which is ironic since ASRock is actually a subsidiary of Asus, even though reports say they're becoming separate companies.
I've thought about it, but it was tough to locate something with solid specifications from a reliable seller.
From what I understood, the situation appears to be that the boards came with voltage settings turned on at a level that was excessively high. The X3D chips are sensitive to this, leaving minimal margin for error, which led to some CPUs failing. Mostly on ASRock boards, which I avoided, but I now realize the problem also affected ASUS models somewhat.
It's really frustrating to find clear information. AMD, Asrock, and ASUS are all giving vague responses and minimizing the issue. Asus boards are quite popular and might be overrepresented in discussions. Research suggests it could have occurred with any board under certain conditions. The issue seems tied to AMD CPUs, though the exact cause isn't fully clear—just strongly linked to X3D. Meh.
It's obvious the core of the problem lies in voltage settings.
It's evident that enabling PBO and EXPO plays a key role in triggering this issue.
It's clear this is specific to Ryzen X3D chips.
It's strongly suggested that both the X3D chips and the motherboard voltages together contribute, though it's not entirely clear how responsibility is assigned.
Some people claim it's exaggerated, but updates and new stock should resolve it, and popular boards are getting worse reviews due to higher numbers.
I don't know what to trust. It's annoying having to discard the entire build just to pay more for a board, but maybe I'll try it anyway.
Kept everything very clean and stable during the build process, reducing the time between opening the socket and inserting components, which eases my concerns. After reviewing updated reports, it's clear the main issue is with the CPUs, which have failed across multiple boards—not just from asrock or ASUS. Most likely, AMD made a mistake here.
It’s frustrating having invested in all this equipment only to miss out on its full potential. Honestly, it’s risky to use faulty hardware where the problem isn’t known, hoping that no-pbo/no-expo will protect it and prevent future complications.
I’ll just make sure I have warranty or CPA coverage, study the settings and variables thoroughly, and proceed. At least then any issues will be visible before the warranty expires. I don’t want to sit on something uncertain for years.
The issue is that most boards are overclocked by default. This setting is meant to improve performance in benchmark tests. Many users apply excessive voltage to ensure stability across multiple chips. I would begin with Core Optimizer to reduce voltages at higher boost speeds. After that, turn on PBO to restore voltages, though not as high as before.
It's challenging to pinpoint the exact reason, but publicly available information suggests the issue might stem from motherboards being pushed beyond their default configurations and X3D chips being affected. Even after updates were released to address this, the problem persisted due to many influencing factors.
I chose to give it a shot and left the specifics here for others to review. If I encounter any issues, I’ll make sure to return and update this post (whether it’s necro or not, at least the information is saved).
I spoke with the seller, who confirmed that by keeping UEFI/BIOS defaults and avoiding manual overclocking or voltage adjustments, I can activate PBO and EXPO, which remains valid for five years. That provides sufficient assurance, so I prefer testing it to see if the system can manage itself.
CPU OPN number: 100-000000719
CPU batch number: CF 2536PGY
It’s worth noting that the OPN number refers to the tray, while my CPU is housed in a separate box. I’m not entirely clear on that distinction.
Once I received the motherboard, I immediately upgraded to UEFI/BIOS version 1804 using a USB flashback, without connecting any components. I disabled PBO and left EXPO off. After a few days of stable operation, I briefly tried PBO and EXPO again. It functioned correctly, though I wasn’t fully confident and turned it off afterward.
I continued using the CPU at 4.3Ghz for 3–4 weeks with EXPO disabled. Later, I updated to UEFI 2004 via USB flashback.
Edit: After installing updates, I didn’t clear the CMOS, as all profiles and settings were already reset. I booted without any changes and confirmed everything worked.
I enabled PBO, kept everything at default, and it passed a quick test—Space Marine 2 AOK.
Switched PBO to AMD ECO mode; the same test passed.
EXPO Profile I was set to 6000Mhz. memtest86+ from grub completed with 100% success, no errors. The RAM module is listed on the motherboard’s QVL, which surprised me since I had assumed otherwise. The only oddity was that fans remained stuck at a very high setting after training until the next reboot (temperatures stayed within normal range).
I ran performance and stress tests on the CPU. All 32 cores operated at full speed for about five minutes under normal 20°C conditions. The highest temperature reached was 75°C with brief spikes to 76°C.
So far, everything is functioning well. No issues, no overheating, no kernel or dmesg warnings, and no unusual odors.
The performance improvement is clear. I’ve played games with identical settings, pinned to the X3D cores, before and after. Some titles that are more CPU-intensive showed increases of 40 and 60+ FPS.