ROG STRIX B350-F GAMING issues

[theprob]

Hello,

Anyone experienced the following with the mentioned board:

After 6/10 system starts my system don't see the LAN adapter or my mouse at all. I need to restart the system or replug my mouse to solve the issue. I've experienced the mouse connection issue within UEFI too. Furthermore I experiencing further annoying problems, such as random video signal loose (especially when I start a fan calibration within AISuite 3). After that, I can only revive my system with a hardreset.

I tried to reinstall windows/nvidia driver/chipset driver, etc. And of course, I have the latest BIOS version.

Now, I sent back the board to the vendor, because ASUS Support advised that. They will revise it. I hope soo, that they won't evalute that it's just a SW related issue...

[SuperRichioROG]

Hello All,

Bios 5220 Agesa 1.0.0.3 ABBA is live

https://www.asus.com/uk/Motherboards/ROG-STRIX-B350-F-GAMING/HelpDesk_Download/

[kggrm]

SuperRichioROG wrote:
Hello All,

Bios 5220 Agesa 1.0.0.3 ABBA is live

https://www.asus.com/uk/Motherboards/ROG-STRIX-B350-F-GAMING/HelpDesk_Download/

Have you flashed your board with it yet?

What's been your experience with it?

[SuperRichioROG]

kggrm wrote:
Have you flashed your board with it yet?

What's been your experience with it?

not yet, but I'm sure my patience will run out and I'll update soon.

I've been stable with every BIOS so far so I'm not very concerned. If I update some time over the couple of days/weekend etc I'll post my impressions.

Update: My Patience ran out (kind of in record time for me)

First impressions are it's certainly a bios...My system is still stable...the Ryzen balanced power plan (with it's 99% min power state) is still a bit too aggressive in my IMO so I'll probably dial that back just as I have in previous BIOS versions.

Boost behaviour definitely seems to have improved since the last bios. The previous BIOS 5216 I had a slight drop in performance and would never maintain 4.2ghz for long in any games really. I've just briefly tested Redout (Vishnu circuit) with BIOS 5220 and my 3600 maintained 4.2ghz almost the entire race which is something It has never done with a previous BIOS.

Further testing so far F1 2019 also seen the R5 3600 maintain 4.2ghz much longer than it had previously and even control which is much more demanding seen the CPU maintain higher speeds for longer still occasionally reaching max boost.

Idle behaviour has pretty much been the same with the CPU being a bit boost happy with the default ryzen balanced power plan, reducing minimum CPU power state to 5% sorted that out for me with no obvious/noticeable performance penalty.

Another nice win is my system is booting faster, which is nice.

Pretty good bios update so far.

I'll continue testing but early signs seem good.

[Equinox187]

Updated to the latest bios and no bugs yet and it does boost better but i am just not sure it was worth it for that 50MHz, ok so multi core in CB20 went from 3531 to 3601 and single core seen a similar performance boost and that is good i guess but what is not good is that it is throwing more voltage at CPU to get that speed and that leads to heat even when the voltage is not needed, before the update it would idle at 27c and now after the update it idles at 32c and before the update it would game at around 41c but after update it is 56c and this is with a custom water loop with the same fan profile however voltage is where the issue is under gaming voltage now consistently hits and stays at 1.475v under an all core load that puts all but one of the cores to 4GHz and the last one to 4.2GHz but lets not forget you can get a stable all core OC to 4.1GHz with just 1.3v and an all core 4.2GHz with just 1.35v

[SuperRichioROG]

Equinox187 wrote:
Updated to the latest bios and no bugs yet and it does boost better but i am just not sure it was worth it for that 50MHz, ok so multi core in CB20 went from 3531 to 3601 and single core seen a similar performance boost and that is good i guess but what is not good is that it is throwing more voltage at CPU to get that speed and that leads to heat even when the voltage is not needed, before the update it would idle at 27c and now after the update it idles at 32c and before the update it would game at around 41c but after update it is 56c and this is with a custom water loop with the same fan profile however voltage is where the issue is under gaming voltage now consistently hits and stays at 1.475v under an all core load that puts all but one of the cores to 4GHz and the last one to 4.2GHz but lets not forget you can get a stable all core OC to 4.1GHz with just 1.3v and an all core 4.2GHz with just 1.35v

Personally I'd still be very happy with those temps, as for voltage behaviour I'll take a look next time I have a chance to have a gaming session. AMD don't seem to be able to listen when certain things have been raised, such as their ryzen balanced power plan min cpu power state has always been a problem since they set it at 99%. An easy enough fix but annoying all the same. I've never been too concerned with voltage as long as I hit under 1v on each of the cores when idle (which I do as long as I dial back that power plan) and temps are under control.

Update:
Did a run on cinebench r15, all cores ran at 4ghz (3967.8 to be precise) with a voltage of 1.319v, the highest voltage on one of the cores during the run was 1.406 so all within spec as far as I'm aware. I'm running everything stock on BIOS 5220.

[Equinox187]

SuperRichioROG wrote:
Personally I'd still be very happy with those temps, as for voltage behaviour I'll take a look next time I have a chance to have a gaming session. AMD don't seem to be able to listen when certain things have been raised, such as their ryzen balanced power plan min cpu power state has always been a problem since they set it at 99%. An easy enough fix but annoying all the same. I've never been too concerned with voltage as long as I hit under 1v on each of the cores when idle (which I do as long as I dial back that power plan) and temps are under control.

Update:
Did a run on cinebench r15, all cores ran at 4ghz (3967.8 to be precise) with a voltage of 1.319v, the highest voltage on one of the cores during the run was 1.406 so all within spec as far as I'm aware. I'm running everything stock on BIOS 5220.

I am using the stock ryzen balanced plan with bios settings at default aside from docp 3000, what did you change the 99% to ? As for you voltage it could be CB r15 as it is pretty soft on the newer CPUs, its basically been retired by them and replaced with CB r20 as it is designed for the new generation of CPUs and so hammers them harder, might be interesting to see if your voltage increases like mine with CB r20, and yeah I am ok with the temps it just bothers me that they pump more voltage than is needed.

[SuperRichioROG]

Equinox187 wrote:
I am using the stock ryzen balanced plan with bios settings at default aside from docp 3000, what did you change the 99% to ? As for you voltage it could be CB r15 as it is pretty soft on the newer CPUs, its basically been retired by them and replaced with CB r20 as it is designed for the new generation of CPUs and so hammers them harder, might be interesting to see if your voltage increases like mine with CB r20, and yeah I am ok with the temps it just bothers me that they pump more voltage than is needed.

On previous BIOS versions I found that reducing the min CPU power state in the Ryzen balanced power plan from 99% down to 5% or even 0% (as it was previously set in an older version of the power plan) it reigned in some of the over trigger happy boost behaviour and idle temps on desktop and allowed the CPU to reduce its clock speed as necessary. It's less of a requirement in this BIOS version for me as a lot of the crazy boost behaviour (when using a monitoring tool or opening a web browser etc) has been fixed. I've still done it however as it allows my 3600 to idle at around 0.915 - 0.927v with temps around 35-36.5c on a wraith spire, running everything else stock except 3200 docp. Worth noting I've seen no drop in performance at all from this either in artificial benchmarks or whilst gaming.

Taking note of what you said I just completed a couple of runs in CB r20 and performance numbers were very similar to r15 with the same clock speeds and voltages levelling off at around 1.313v-1.319 with occasional brief spikes between 1.325-an absolute max of 1.444v which only showed on 1 core and only once.

hope this info helps in some way!

[Corecom]

I'm still using the old version 5216 of Bios. I have G.Skill F4-3200C16-8GVK 2x8GB RAM dual rank with Samsung E-Die chips.
With my old 1600X I used the ram at XMP profile, 3066Mhz and 1,365 volts and it was stable. I swithed Bios versions, I used the calculator for 3200Mhz but I never achieved it stable.
Now, with 3700X, I run the ram at 3400Mhz with the states of the XMP profile with 1,37 volts. It's stable at games and tests.

[Corecom]

I'M sorry for my incorrect information at my previous post!
I have Samsung D-Die RAM dual rank modules, not E-Die, and my RAM is'nt stable at 3200Mhz! at all I figured it out after some days testing.
My RAM modules are stable now at 3333Mhz 16-16-16-35 CR1 at 1.365V. I tested a few day ago and there's no crashes/halts at these settings at games/programs.
I overclocked the RAM modules to 3466Mhz to 16-18-16-38 CR1 and it's stable too at 1.37Volts but there's no such a difference at PC productivity in games/Cinebench 15/20 so I run my system with my RAM settings at 3333Mhz for 24/7 16-16-16-35 CR1 at 1,365 Volts.

[gdp77]

Anyone tried Asus B350-F Gaming with 2x16GB or 4x8GB memory modules? What memory speeds can you achieve?
I have a 2x8GB (2933MHz) setup (Ryzen 1600) and I am thinking of upgrading to 32GB. Do I add 2 more 8GB (single rank) modules or should I buy 2x16GB (all of them are dual rank)?
I would like to keep 2933MHz setting as well.

What do you think?