Boost frequency update on Ryzen 3000 series, PBO working as intended? WHEA error update #2
We have updates on Ryzen 3000 frequencies on AGESA 1002, PBO working as intended and WHEA errors being recognized. So I will try to make this quick since I’m behind the reviews I am supposed to finish, but well as I mentioned on the last post, AGESA 1002 tested on a Crosshair VII Hero had the same results on the review we did on Ryzen 9 3900X (all frequencies hitting specified turbo).
While TheStilt explained that all cores should not reach Turbo Max (4.6 GHz on 3900X), however, Gupsterg (an enthusiast user on OCN, I do remember him Ryzen 1000 series on OCN) has been able to perform more tests using AGESA 1002 (not used for launch reviews) with the sample he has (Ryzen 5 3600 series).
I will explain it very quick. AGESA 1002 as I mentioned on my review, makes the chip behave with all cores working at max turbo boost (until now, AMD has not replied our e-mails according to boost frequencies).
Gupsterg as mentioned on the last post, with AGESA 1002, hit the same turbo clocks (max, 4.2 GHz) on ALL CORES a few days ago (same behavior we observed with AGESA 1002). Moreover, he has been able to use Precision Boost Overdrive 2 as it was intended by AMD (or at least what everyone understood), by actually increasing the maximum Turbo Clock over the specification. That means, he has actually been able to reach 4275 MHz on all Cores with his Ryzen 5 3600 and he is testing 4325 MHz now.
Gupsterb video on Ryzen 5 3600 with PBO +75 MHz over máx frequency on all cores
Due to reviews, I have not tested AGESA 1002 thoroughly but with this new information on hand, I will try to replicate it and see if I have any success. His video though, actually shows frequencies hitting 4.275 GHz and there is a photo with 4.325 GHz being tested.
Well, as I mentioned on my review, I do not know if this is the normal behavior and how will this affect launch reviews. Hopefully it won’t change results with no performance regressions, although Anandtech had to retest everything on his sample/motherboard combo and there was single core variance.).
PBO is actually working as shown in AGESA 1002, but it is indeed an odd behavior (max boost frequency) and working as AMD marketing has explained on their videos.
There is some few things to note here and it’s the voltages being fed to his sample (Gupsterg), which actually still are under specification (at least through software reading). This might be a behavior exclusively on 3600 chips, since 3900X samples are actually being fed quite some voltage (up to 1.47-1.48v on single core tasks).
Also remember Mr. Robert Hallock from AMD (Technical Marketing), said up to 200 MHz. The gains on Ryzen 9 3900X could be limited, due to voltage/temperatures constraints. PBO does scale with good dissipation but time will tell.
About WHEA errors
NVIDIA finally started doing investigations on what we reported on day 1 of our review coverage on Ryzen 9 3900X. There are currently under investigation, but Tom’s Hardware actually quoted something somewhat interesting:
“Nevertheless, the cause of the BSODs might be deeper. Perhaps it could be chipset related or induced by certain Ryzen 3000-series CPUs. At this time, it’s anyone’s guess until Nvidia sheds some light on the issue.” – Tom’s Hardware
We hope the WHEA errors are related to drivers; since the behavior appears on X470 and X370 motherboards, (other users have reported this). Personally, I have tested on X470 (AORUS Gaming 7 Wifi) reproducing these errors during my week of testing prior to review launch.
I cannot give any conclusion to this odd behavior on both boost behavior, PBO working as intended and the PCIe WHEA errors, so right now the only thing I can do, is report the anomalies in this article.
If you liked this article, you help us a lot by doing like, share. If you have any questions, you can follow us on our social networks: