Crosshair VIII 2501's for testing

[Baio73]

You're welcome And is it now more stable for You?

Still testing, but seems ok... obviously as this RAM is rated 17-17-17 @4000Mhz.
Why BIOS 2701 (beta) and not 2702 (final)?

Baio

[Redline9k]

Okay, so after some testing I can indeed confirm what darkmessias94 is saying in regards to Fmax Enhancer:

Disabling Fmax enhancer will take my single-core turbo speeds from the ~4.6Ghz or so with Ryzen Master under CB20 single-core test to ~4.9Ghz as reported by Ryzen Master.
Additionally, single-core workload voltage is no-longer in the high 1.38v area, but is now properly at 1.47-1.49v.
Fmax enhancer on ASUS boards, that "ASUS exclusive feature" is indeed broken. At least for me on my C8H WiFi and 5950X. It's doing the opposite (hampering single-core turbo) rather than what it's supposed to do (expanding it).
I wonder if low single-core workload voltage is what's causing the system resets? I've not played around with the curve optimiser yet.
People can test this themselves by closing all software except CBR20 and Ryzen Master and observing single-core test voltages. Proper single-core boost voltage is up-to 1.5v falling according the more cores are loaded.
Single-core CBR20 score went up too. Later today I'm going to re-update to the latest BIOS version and keep Fmax disabled and see if my system resets return.
If they don't, I think we know the cause.
CBR20 multi-core score has also gone up to ~11100 points. All-core load shows 4.5Ghz sustained across all cores on Ryzen Master, and voltage drops as is expected. This is a small boost over the stock CBR20 multi-core scores of ~10500 or so.

Does indeed seem like FMax Enhancers' much lower single-core voltages could be what's causing the resets. Would explain why it only happens in light or single-threaded workloads.
Hopefully shamino can use this information for future testing versions.

[Jackalito]

Okay, so after some testing I can indeed confirm what darkmessias94 is saying in regards to Fmax Enhancer:

Disabling Fmax enhancer will take my single-core turbo speeds from the ~4.6Ghz or so with Ryzen Master under CB20 single-core test to ~4.9Ghz as reported by Ryzen Master.
Additionally, single-core workload voltage is no-longer in the high 1.38v area, but is now properly at 1.47-1.49v.
Fmax enhancer on ASUS boards, that "ASUS exclusive feature" is indeed broken. At least for me on my C8H WiFi and 5950X. It's doing the opposite (hampering single-core turbo) rather than what it's supposed to do (expanding it).
I wonder if low single-core workload voltage is what's causing the system resets? I've not played around with the curve optimiser yet.
People can test this themselves by closing all software except CBR20 and Ryzen Master and observing single-core test voltages. Proper single-core boost voltage is up-to 1.5v falling according the more cores are loaded.
Single-core CBR20 score went up too. Later today I'm going to re-update to the latest BIOS version and keep Fmax disabled and see if my system resets return.
If they don't, I think we know the cause.
CBR20 multi-core score has also gone up to ~11100 points. All-core load shows 4.5Ghz sustained across all cores on Ryzen Master, and voltage drops as is expected. This is a small boost over the stock CBR20 multi-core scores of ~10500 or so.

Does indeed seem like FMax Enhancers' much lower single-core voltages could be what's causing the resets. Would explain why it only happens in light or single-threaded workloads.
Hopefully shamino can use this information for future testing versions.

Thank you for that, as I was experiencing those instability issues myself. Will check it out again over the weekend.

[MultiDoc]

I’m a bit puzzled this thread used to be 18 pages long until yesterday and now suddenly I can only see 5 pages ??? Am I missing something ?

[RedSector73]

I’m a bit puzzled this thread used to be 18 pages long until yesterday and now suddenly I can only see 5 pages ??? Am I missing something ?

I still see 18....awkward!

[Baio73]

An still no solution around!

Baio

[Jackalito]

I’m a bit puzzled this thread used to be 18 pages long until yesterday and now suddenly I can only see 5 pages ??? Am I missing something ?

I'm still seeing the same 18 pages of this thread.

[uplink]

Still testing, but seems ok... obviously as this RAM is rated 17-17-17 @4000Mhz.
Why BIOS 2701 (beta) and not 2702 (final)?

Baio

I'm livin' by a rule, if it works, don't try to fix it . And from my current experience with AMD platform, "stable" ain't much better than "beta" and vice versa. Now I'm hyper stable with "a okay" performance [not review pages, but I'm good with it] on my system, and I can uninterruptedly work and game. When I'll see some newer and higher numbers I'll try them, though, there's not many reasons to go from 2701 to 2702. From what I've read on plenty forums, they didn't fix the RAM problem, so the only chance is that Shamino or whoever makes those BiOS-es nicked something. I'm waitin' for new AGESA version release and till than I'm sitting tight

[chris.skaling]

My 2 cents...
This BETA BIOS gives me the same exact problems as the previous BETA 2402 and the 'official' 2311.
The last version that worked good for me is 2206.
My rig is 3700X - CrossHair VIII Formula - 2x8Gb G.Skill F4-4000C17-8GTZR and 2x8Gb G.Skill F4-4133C19-8GTZR.
CPU is @stock, my struggle is to set the RAM to 3600 CAS14 (as they worked with BIOS 2206).
The system is rock solid both under Windows (Kharu) and under DOS (PassMark MemTest), can pass a +12h of stress test, but as soon as I reboot, it hangs on a variety of errors like "TEST CMOS", "CHECK CPU" or "LOAD VGA BIOS". Have to press the case reset button to make it boot again.
Sometimes I have BSOD loading Windows.

Here are my settings:





Nothing else set in BIOS, just DRAM voltage to 1.37v instead of 1.36v and disabled SATA (I have only NvME drives), WiFi and Bluetooth.
In the Power Supply Section, where there are 2 values suggested, I always set the highest one (140% instead of 130% and so on).

Any suggestion before reverting to BIOS 2206 is very much appreciated, thanks.

Baio

EDIT:
Testing under Windows with 1.075v VSOC instead of 1.1v, seems stable:

Is anyone else getting these issues? Right now i am getting most freqently the VGA BIOS issues during post, eventually it boots but won't output to the screen. Have to CMOS reset. Everything else seems fine. I just downloaded the latest posted official ROG CROSSHAIR VIII FORMULA BIOS 2702 posted on asus support, but same issue. This is happening with stock settings, just optimized defaults. Any ideas?

[RedSector73]

Any ideas?

An still no solution around!
Baio

On the CH8, I've read the Cstate has been said to be too aggressive that some CPUs either blue screen or reboot due to the CPU core voltage being lowered too much when the CPU is not under load.

It should be under Advanced, AMD CBS, NBIO Common Options, SMU Common Options, DF Cstates. Could set this to Disabled.
Source: https://www.overclock.net/threads/as...hread.1728796/