Thank you for the feedback, will try that out!

maybe you wait until coming weekend though, still quite some bugs in that corecycler script. just read up on it. (don't have a vermeer myself at this time)

Appreciate the feedback.

I too am running at 3800c16 with tighten sub timings, absolutely rock stable so far.

I believe there is one last major AGESA release left for Zen 3, probably slated for summer of this year.

Someone should bring this cooler in.
https://www.techspot.com/news/88865-fan-big...-weighs-44.html

Icegiant Prosiphon Elite review:
https://www.youtube.com/watch?v=U-BWEDfrE9c

Heard it's related to higher overall bandwidth use, hence why it's more common for bonchi to get his errors after swapping to RTX3080.

I heard the same for VR posted by a few people. I know Ampere has driver issues with VR, not sure if the posters saying VR affects it was on 2080 Ti though.

5800X is more common making it easier to buy, presumably because AMD is dumping lots of failed 5950X chips with one CCD in deep sleep. Many 5800X units apparently have a CCD2 sensor reported on HWinfo. Igor's Lab mentioned he found this several weeks ago with the HWinfo dev.

What temps are you seeing without A/C? Thinking of getting the EK AIO as the non-RGB one is rather cheap.

Also, does CTR 2.0 report your chip as a golden sample after it tests your CPU?

IIRC this was built mainly for Threadripper?

For multi chip packages, Zen 2 and 3 falls into that category.

https://www.icegiantcooling.com/

4200 RAM in coupled mode - is this a golden sample for IMC?

https://www.reddit.com/r/overclocking/comme...x_micron_rev_e/

although the trfc is pretty loose for a rev E but not bad... but it's a 16GB single rank kit so I guess that helps too.

attached one other 4200 with DR. not mine but actually since early days there are already some being able to do 4200 synched. I even heard from dr b-dies doing this.


funny thing though is that this is not just a cpu issue apparently. I know ppl testing the same chip on different boards where on one board it maxes out on 1900 or before that and spews errors on >=1933 while just by changing the board n keeping other stuff it'll do 2000.

guess this needs some more testing of users to zero in on it as AMD doesn't seem to bother, at all.

It could be bandwidth related, but as far as I've experienced - doesn't seem to be the only reason for connectivity issues. I have 3 external drives connected all the time too, so coupled that with the DACs I test and when gaming in VR/fighting/racing, the bandwidth load on the I/O should be very high. This is reflected in the ridiculous heat output on the X570 chipset specifically on the CH8, which goes up to 69c for me.

So what Bonchi experienced goes deeper than just I/O bandwidth IMO.

VR has been broken for months with nvidia drivers. 461.09 improved things, but still far from perfect. I'm still getting the odd judder here and there, doesn't make it unplayable like before where there are over 200ms frame spikes every few seconds. The last properly working driver was 446.14, but fucking 3080 doesn't work on that driver now does it? Dumbass nvidia.

This post has been edited by SSJBen: Mar 10 2021, 05:06 PM

It's good to know 5900x demand is getting lower... YES IT listed the processor for RM 2,999 for quite sometime and it's yet to be sold. They have also increased the inventory to 2 units.

Maybe PC enthusiasts like you guys have gotten what you want... People who sitting on the fence still waiting for the price to be normalised and have no urgency to get it yet...

yeap.. it is down to many things ... like chipset drivers, device drivers, windows version, software compatibility, and also RAM compatibility which plays a very BIG role .. and not to mention their crappy thing called AGESA.

This problem i dont think is something new, if you look back.. there's already such feedback since zen2 and Im one of them on a B450 board , as you know I have been having this problem since 2019...... I guess because more people using AMD now so there's more of such issue being reported and it has been over a year yet AMD still havent have a full fix for this. On one AGESA 6 months after I had my system running, they did make my USB hub work but not 100%... And later, took more than a year for them to officially acknowledge this problem and start to collect system configs that are affected, while before that they never entertain or reply any of my feedbacks.

So I dont have confidence that AMD will solve this anytime soon and so I made the leap back to intel.

I think they are stuck in a conundrum. There are more people without the issues than there are with issues.

Now if they proceed and fix it for those with issues (because AGESA's tradition is fix one thing and break 10 other things), the ones without issues might THEN have issues lol. I don't think they know how to deal with that.

This post has been edited by SSJBen: Mar 10 2021, 06:34 PM

I saw a new beta for MSI X570 Tomahawk

ComboAM4PIV2 1.2.0.1

Anyone knows whats the update about compared to AGESA 1.2.0.0 ?

Nothing major, just to fix L3 cache performance issue.

Fix L3 cache bug in AIDA64 benchmark.

no impact to gaming/daily use right?

I have problem with BIOS update ; everytime upgrade ; I cant restore all previous settings.. (even after backup to USB)

it will say the setting not compatible with new bios..

so if no impact to daily use Ill skip for now

No, the L3 cache was read wrongly in AIDA64. It's not read wrongly anywhere else. Don't have to update unless you absolutely love to just play AIDA benchmarks.

BIOS backup to new version is pointless. In fact I suggest you never use your back up from a previous version. There has been many reports when going from one version to the next, even if the backup works the settings may not be identical. This is because AGESA can move certain options around and the backup values are stored as a table.

So if the table is no longer identical, the BIOS no longer knows which value goes where and that's why it blocks it from applying your precious settings. Certain boards (Asus especially) allows the backup to be applied regardless, but there is a big chance your system would not boot up anyways. I know a guy from OCN he used his back up when going from AGESA 1.0.0.3 ABBA to 1.0.0.4 patch B, what he didn't realize was that the voltage options has changed places so his RAM voltage now has been taken place of the IOD. Meaning he was pumping 1.4v into the IOD and that basically has only 1 outcome - fail boot.

OTOH reading some reports that 1.2.0.1 fixed USB drop out issues on Asus x570 boards. Hmm..

Strange, recently my 5950X start to randomly blue screen with WHEA.

But there is nothing at event viewer on which WHEA code (18 nor 19).

Just a bugcheck and a large Memory.DUMP

The random blue screen with WHEA (no error code) seems to happen on idle only.