Check if the PBO feature is activated for the Ryzen 9 7900 CPU to reach its advertised speed.
Check if the PBO feature is activated for the Ryzen 9 7900 CPU to reach its advertised speed.
Hello! I recently assembled a PC equipped with a Ryzen 9 7900 processor.
After performing some performance tests, I discovered my scores were significantly lower than average (about 25% below the median).
After updating the BIOS and installing the appropriate chipset drivers, the improvement was modest, reducing the loss to around 10–15%. This remains a considerable gap, and during benchmarks I observed the CPU consistently capped at 4.47 GHz across all cores. In single-core tests, I reached up to 5.29 GHz per core.
I attempted to turn on PBO in the BIOS, which allowed me to achieve up to ~5.05 GHz during benchmarking, and the scores improved noticeably, bringing me closer to the median values I’ve seen online.
The question remains: should PBO be enabled to unlock advertised boost frequencies? Enabling it increases CPU temperature and power consumption, which is undesirable given my preference for quiet operation.
I will disable PBO anyway, as it results in a completely silent system during testing. Although temperatures remain stable (never exceeding 72°C), I can still hear the fan running.
Overall, my main goal is to verify whether this behavior is typical.
By the way, regarding the BIOS update on my MSI A620M-E motherboard, my 5600 MHz RAM is now limited to 5200 MHz or less. If it exceeded that, the system wouldn’t start and the EZ Debug LED would flash an error. 4800 MHz still functions, but 5200 MHz causes crashes—only fixing this requires connecting the dedicated "BAT" pins on the motherboard. A-XMP also triggers crashes. Previously, everything worked smoothly with 5600 MHz displayed in Windows Task Manager. I’m unsure if this is connected to the issue.
I reverted to the previous BIOS version, and everything is functioning properly. I can re-enable A-XMP or set the frequency to 5600 MHz without any problems.
Thank you in advance for your assistance.
PBO isn't required for the advertised speed boost, but it's not a given either. With a 65W TDP chip and a three-chiplet design, it's likely the system includes strong throttling features. The A-series motherboards usually aren't built for a Ryzen 9, and the lack of a VRM heat sink will restrict power delivery when performance demands rise. This review of the Ryzen 9 7900 also shows it struggled to reach 5.4GHz: https://www.techpowerup.com/review/amd-r...00/24.html. In this situation, missing 200MHz results in roughly a 4% clock speed drop, which doesn't always equate to a 4% performance hit.
Yeah, I understand your concern. At 4.4 GHz across all cores it does seem quite low, doesn't it? That could indeed affect how the CPU performs, especially with a 65W processor. I didn't expect that much from the motherboard either. TechPowerUP's data suggests a boost of around 5.0 GHz, which is significantly higher than what I receive (and what PBO would provide).
Then you should have obtained the 7900X
😉
For comparison, you'd need to employ the identical benchmark "This test uses a custom-coded application that simulates real-world performance—it isn't a stress test like Prime95", which itself is not a stress test benchmark like this:
"On CCD 0, the frequency settles at 4350 MHz and on CCD 1 at 4250 MHz. In short, six cores at 4350 MHz and the remaining six at 4250 MHz. Thus, CCD 0 is clearly more powerful"
Review: Ryzen R9 7900, R7 7700 and R5 7600 - Page 4 of 10 - Overclocking.com EN
Today marks the official release of the non-X Ryzen 7900, 7700 and 7600 models, which feature a lower TDP than their larger counterparts
en.overclocking.com
EDIT:
Which BIOS version were you using and why did you not reach 5600MHz?
It didn't work on version 1.40 (7E28v14) yet, but it functions on the earlier stable release (the one I received with the motherboard), which was version v.130 (7E28v13).
What caught my attention is the lower scores I see in cinebench compared to what should be expected with more cores, which is quite surprising.
EDIT: I reviewed the article and it does reference clock speeds similar to mine. So I’m unsure what’s causing these low results. The cooler temperature without PBO never goes above 62 °C.
I also didn’t choose the X variant because I prioritize quiet operation and the performance was only slightly below average for the non-X model (around 5 to 10%).