Originally Posted by wisepds
ohh!!!! a bird or a CH6 bios? jejeje please @elmor, tell me that is new bios, coold bug is history and my system is now stable...
These BIOSes should work fine with your DRAM settings when booting from a fully powered off state.
Originally Posted by CeltPC
Can you elaborate on what Geardown is?
Command Rate sets how many clock edges to wait after setting up addressing (how long to wait before reading or writing data), while Geardown is lowering the frequency of the commands sent to the DRAM (data speed is still the same).
Originally Posted by xyzkill3r
Fresh instal of the new beta and everything at stock, It increased my BCLK (now reporting as 100.6 in bios). I don't think my old hd7950 likes any touching of the BCLK (fans on graphics card only spin 100%), even manually setting it to 100 on bios 1201 caused the same behavior.
So far, have only been able to boot the new beta by clearing cmos and booting directly into the drive via the boot window. Subsequent boots/retries spin up graphics card and QCODE 55. Though that's supposed to mean issues with memory.
Can you try changing AI Overclock Tuner to Default? Did you flash the 0003 BIOS?
Originally Posted by virpz
The first board I had was the Prime X370, my second was the Taichi and now the C6H.
When I noticed the +12V droop on software with the Prime I had it confirmed with the DMM. I then went to check my lane distribution and since everything was correct I assumed my PSU was going bananas. But then I got the Taichi, everything there was rock solid as it used to be, confirmed the +12V under load with a DDM. Now I got the C6H and that +12V behavior droop is going on again but then my system is nowhere close to extrapolate my PSU lane ratings, had solid +12V regulation with the Taichi. I must assume it is something with the asus boards or asus boards + the maxrevo.
I would appreciated to have a word from the [email protected]
on this behavior.
Originally Posted by virpz
Yes, many times now.
I have had three X370 boards here, all run with the exact same GPU, CPU, memory, cabling, rail distribution, PSU etc.
First Prime - Second Taichi - Now C6H. With all the asus and the asus only I have experienced heavy +12Vdroop. Now, I want to know wth is going on as I don't want to have my +12V line acting up like that.
I have now run the OCCT Linpack with only the 8P to cpu connect. The +12V behaved the exact same way as with it having the 8P and 4P to CPU.
Software reading or DMM? Software readings are flaky.
Originally Posted by bluej511
So tempted to try out this new BIOS but system is stable already, id love to know if it fixes the random f9>boot loop issue. I may wait a while to see what it actually does, if its just bug fixes, ill pass haha.
It should fix problems like this, yes.
Originally Posted by Bart
I just put my 1700 under water, and under a mild OC I'm locked at 47C under load (3.7Ghz @ 1.35v). Memory is at 3200 14-14-14-14-34 1T. I flashed 0003, then 1107. So far being in a proper case with proper airflow under a good CPU block is doing my 1700 wonders.
Elmor: any details on those 2 99xx bioses? Are they 2T / 1T variations?
Command Rate and Geardown are adjustable now. The difference now is different AMD firmware which are better for specific cases.
You're only seeing 16GB of memory, I think you need to do some debugging and check which DRAM sticks and slots are working.
Originally Posted by SlayerEru
Guess Cold Boot is still intact, prob wont be when the final bios is release (since after all this is a Beta).
I'm on 9943 using XMP profile at 3200mhz from G.Skill Flare X 3200.
Actually the Cold boot is abit worse on these 2 beta ones than the 1201. I usually just get one Cold boot if I take the power off from the wall, but this one I get about 3 Boot loops until the system starts booting up.
Originally Posted by FloppyDrive
Seem to have frequency display problems in Windows at 3600Mhz (BIOS 9945)...
Booting from a fully powered off state should be fixed in the new versions and not cause DRAM training to fail. Some software needs updating to be able to read the added ratios.
Originally Posted by finalheaven
Well I couldn't yet find a stable setting for 3466mhz...
Went back to 3200 @ 14-14-14-34-1T 32gb (4x8gb Samsung B-Die) which seems to be perfectly stable (HCI tested)
Strangely 3200 works so easily SOC 0.95v and DDR 1.35v
Will try again tomorrow again for 3466... may need to step timings back all the way to 16-18-18-36 though
Will probably have to use SOC 1.05v and DDR 1.4v or so...
You can try changing the CLDO_VDDP setting (in AMD CBS, don't remember exactly which submenu), it's the replacement for VDDP found under Tweaker's paradise. Our setting is not affecting anything anymore, AMD changed the internal LDO setting. It can help stabalize things and even help with the memory frequency hole. Recommended values are around 900-1000mV, default is 950mV.
Originally Posted by 4rcherz
tried the 2 new BIOS and can't boot with rated speeds
F9 code whatever i try.
Can run 3200 14-14-14-34 but not the rated 3600 17-18-18-18-38 any suggestions?
See the above recommendation, also try different values for ProcODT. Far from all systems can run 3600 MHz, it depends on the CPU/MB/DRAM.
Originally Posted by Mandarb
Anybody with 2x16GB Samsung B-die RAM able to break free of 2666MHz on one the new BIOS? If so, which one did you use and what voltages/settings procODT?
Currently on 9943 and trying but so far no luck. (2x 16GB G.Skill Trident Z 3200 CL14 here)
We've tested it internally to 3200 MHz.
Originally Posted by Seth-01
At first my rig :
- Ryzen 1700 with Noctua NH-D14
- Asus Crosshair VI Hero flashed with 9943 beta bios
- 2x8 GB GSkill Trident Z F4-3200C14D-16GTZ
Since the beginning no way to have my RAM work or boot at 3200 MHz. 2933 Work fine and stable. I tried all bios and settings to have my RAM boot at 3200 MHz and my hope was high with this new AGESA bios update.
Absolutely no way …. 3200 Mhz is just impossible. Always the mobo boot with 1 bip and then reboot again to find my memory frequency selected at 2133 MHz.
OK then rules with 2933 MHz but …. an idea …. I tried the new 3066 ratio and it’s boot !! Under windows it’s ok, CPUZ can’t see the RAM frequency but HWInfo can. Nice it works !! I reboot, and under RAM ratio i select the 3333 MHz ratio. It works also !!! Wow ….
Why not give anotgher try. What about the 3466 MHz ratio. Selected, reboot and …. I work again. Under windows HWInfo confirm the frequency at 1733 MHz. What the hell and *** !!! Quick test with Cinebench 15 and Firestrike and its stable at the moment. Very impressive but cant understand why 3200 MHz is impossible but 3466 MHz work perfectly. Goal reached for me with at least 3200 MHz for RAM. I can follow testing the thing !
Great bios for me ! THX to ASUS and AMD making this possible. No regrets buying this wonderful motherboard.
(sorry for bad english)
That would be the memory frequency hole, you can also try CLDO_VDDP as above. But why do 3200 when you can get 3466.
Originally Posted by district9prawn
On the new bios, I'm seeing strange numbers in the Aida64 ram and cache benchmark with ram at 3600.
At 3466 and below, memory bandwidth values look fine.
Try setting tRDRD_sc = 1. Is only AIDA affected or do you see a performance drop in any real world application?
Originally Posted by Targonis
I still can't get to 3200, but I can get to 2993 with both of the new BIOS versions. Hynix M-die. G.skill F4-3200C16D-32GVR, 2x16. This is a lot better than needing to struggle to hit 2667, but not all the way there yet. I haven't played with the ProcODT yet to see if I can find one that will get 3200 stable.
3200 MHz can still be tricky on some systems which has worse margin (CPU and MB). 3066 probably works at least?