As per title, I am seeing my CPU being limited to 5.5Ghz boost after upgrading to the latest stable BIOS for my board, which is version 0705.
I am also having issues with my second CCD being about 200Mhz slower than the first and not hitting 95c as has been well advertised for being “by design” for the 7000 series CPUs.
Can anyone point me in the direction of the AGESA information that might explain what changed in this BIOS to cause this (which is ComboAM5PI 1.0.0.3 Patch A) and is anyone able to offer up a suggestion for what i can do to set my CPU free? Earlier version of the BIOS were problematic for me as i was getting random crashes under gaming loads which turned out to be memory related, which for this system is 32Gb Dominator Platinum 6000Mhz non-EXPO.
Also, can anyone confirm whether the behaviour i am seeing with the second CCD not just being slower but also not hitting the “by design” thermal target is expected, or whether something might be amiss? I am using the Corsair H170i 420mm AIO to cool the CPU and have reseated it twice to ensure an even mount and spread of thermal paste, which is Kryonaut Extreme.
Take a look a your CPU stats using ryzen master (or ryzen monitor on linux, if it works on AM5). There are motherboard limits that you may be hitting like power.
If your issue is hitting max voltage or power, then using a PBO undervolt may give your CPU more room to boost higher. Though there is such a thing as the silicon lottery. Also worth noting that 5.7 GHz is the max single core boost, getting 5.5 GHz all core sounds like it’s well within expectations.
Sorry, what i didn’t mention in my original post is that on the first BIOS the board came with and the subsequent 06xx revisions, i was regularly hitting 5.775 in low thread workloads like gaming, single core test of CB23 or even just sitting in windows browsing or the likes. However, I was coming up against some stability issues which I believe were memory compatibility issues (during gaming the PC would just reset, no BSOD, no error messages, just a reset like someone hit a switch).
The 5.5Ghz limit only appeared once i went onto revision 0703, which at the time was a beta and 0705 which is the most recent stable version.
I have PBO limits set to Motherboard and have been able to set a -30 value on all cores in curve optimiser with no stability issues, yet the 5.5Ghz limit is still there. I have yet to do an extended run of Corecycler to fully verify this.
I have monitored the power limits in HWinfo64 and Ryzen Master (not at the same time) during an all core test of CB23, although it comes close, the system never hits the limits for PPT, TDC or EDC. I am open to setting this limits higher, not sure what benefit would come of that if i am not already on the limit?
I was seeing some random reboots, with no errors logged, while doing some extended stress testing on my X670E Hero with HWINFO64 running. After doing a bunch of testing I am pretty sure the resets were caused by HWINFO64 - specifically having “EC Support” enabled (under the Safety tab). I have not seen any reboots without HWINFO64 running or with “EC Support” disabled. To be fair HWINFO64 does warn you that “EC Support” can cause stability issues on some hardware.
check bois/eufi for h-pet. (high precision event timer)
enable it if its disabled. it will help with system stability…
it often ships in a disabled state and as a result causes no end of issues for the end user, as its use is never described in the manual …
(its a hardware clock windows 7/10/11/ and some distros of linux use to keep system events timed across the system in a way that isnt affected by the cpu overclocking/boosting)
I see the new x3d chips have been announced and they are still flogging them as having a max boost of 5.7Ghz despite the fact they are limited so i wonder if the new chips will be unshackled?
…