X370 Tai Chi - Issues
Printed From: ASRock.com
Category: Technical Support
Forum Name: AMD Motherboards
Forum Description: Question about ASRock AMD motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=5751
Printed Date: 22 Dec 2024 at 1:12am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: X370 Tai Chi - Issues
Posted By: Haaviko
Subject: X370 Tai Chi - Issues
Date Posted: 30 Jul 2017 at 6:34pm
|
" rel="nofollow - Hello, I've recently purchased the X370 TaiChi motherboard along with the G Skill 3200Mhz Flare X Ryzen Ram & the Ryzen 7 1700 CPU. My BIOS settings are as followed: OC Tweaker CPU Freqency(Mhz): 3800 CPU Voltage(V): 1.35v - LLC2 XMP Settings Profile for 3200Mhz C14 SoC Voltage: 1.1v - LLC2 Advanced > CPU Configuration Cool n Quiet: Disabled C6 Mode: Disabled Advanced > AMD CBS > Zen Common Options Core Performance Boost: Disabled Global C State: Disabled Since I'm using 1DPC (2 modules total) single rank modules (GSkill Flare X 3200Mhz 16GB Kit) my settings for optimal performance is: BankGroupSwap: Disabled BankGroupSwap Alt: Enabled GearDownMode: Disabled 1T So using these settings I've managed to get into Windows, do some benchmarks.. gaming and even Prime95 with no WHEA errors. But as soon as I shut down the computer or restart, I usually get a BIOS error code with black screen. Something like 68 or 0d etc. Can somebody help me out? Kind Regards, Haaviko. |
Replies:
Posted By: datonyb
Date Posted: 30 Jul 2017 at 9:25pm
|
as strange as it sounds try bankgroup auto /enabled geardown auto/enabled failing that set the ram timings to 16/14/14/14/ i also have a taichi and have been playing with my gskill trident z for quite some time now what i can get from the ram in windows and on warmboot/restart is much tighter and faster than any cold boot (mine will always do the boot loop thing and then reset defualts) yet runs all day without issue when set and rebooted warm ------------- [url=https://valid.x86.fr/jpg250]  [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: Haaviko
Date Posted: 30 Jul 2017 at 9:57pm
| Yeah I can easily set the BGS Auto and GDM Auto and don't have the issues, but the truth is.. I've noticed better performance using GDM Off, BGS Off, BGS Alt On for my setup. |
Posted By: prodev42
Date Posted: 30 Jul 2017 at 9:59pm
|
no reboot issues here. my 1700 @3.9ghz, used offset+pstate way so did not set oc mhz and cpu volt in oc tweaker directly...offset volt 0.1625(1.35v - 0.1875v), cpu llc 1, soc 1.1v, soc llc auto, DRAM volt 1.4v, pstate0 fid (the hex for 3900) .. ran CPU:OCCT for 10 min was stable, ran cinbench without crashing, before ran aida64 1 hour for same setting @3.875 no crash...should be ok for my use for now, only play the game pubg and don't need to do 100% cpu stuff like video encoding...will test more tomorrow.From reading everyone's post on reddit and overclock.net for a week i consider my chip average in da lottery. I think raising DRAM voltage, raising SOC voltage, set CPU LLC to 1 but SOC LLC to auto was the key for me..without changing these 3 places i would never get to 3.9 stablely using 1.35v. Oh and my Flare X Cas 14 ram ran at 3200 using XMP profile, cooler is Noctua D15 AM4 edition. Please note I did not change anything else in the bios such as Cool n Quiet, C6 Mode, Global C-state etc, let me know if I should to make it even better.
|
Posted By: Haaviko
Date Posted: 30 Jul 2017 at 11:56pm
| Looking for support from Asrock staff please! |
Posted By: datonyb
Date Posted: 31 Jul 2017 at 1:18am
|
prodev my taichi and r7 1700 gets 3.900 at 1.3 volts and has from day one i set llc 2 and turn off all the power saving stuff you mentioned pstate settings 9c 8 27 set voltage to fixed on first page and llc level 2 hope this helps ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: prodev42
Date Posted: 31 Jul 2017 at 2:38am
" rel="nofollow -
did u change ram voltage ,soc voltage and soc LLC? ..i think you got a great chip and mine will never be able to pull this off. by power saving stuff you meant the following 4 items? Advanced > CPU Configuration Cool n Quiet: Disabled C6 Mode: Disabled Advanced > AMD CBS > Zen Common Options Core Performance Boost: Disabled Global C State: Disabled
|
datonyb wrote:Posted By: Haaviko
Date Posted: 31 Jul 2017 at 9:49am
| " rel="nofollow - Hey do staff ever check these forums or do I need to call Asrock directly? |
Posted By: chainsaw
Date Posted: 31 Jul 2017 at 10:50am
|
you can try to call them. i have not gotten anyone to answer yet. have left a message for a callback and nothing. The Email support is fast but did not offer any help. ------------- ASRock Fatal1ty X370 Gaming K4 P3.40 AMD RYZEN 7 1700 @ 4Ghz 1.39v Noctua NH-D15 "blocks top 1x pci-e" 16GB GSkill Ripjaw V F4-3200C16D-16GVRB @ 2933 14-15-15-15 T1 1.365v GTX 1080 x2 Sli |
Posted By: parsec
Date Posted: 31 Jul 2017 at 10:52am
So you get "something like 68 or 0d etc." POST codes are rather specific, and your statement is rather unspecific. How are we supposed to know what the issue might be if we don't know what POST code is shown on the Dr Debug display? You gave us two codes, and etc? What it etc? Then you said this usually happens when you shutdown or restart. So it is also intermittent? Doesn't happen all the time? Intermittent issues are the most difficult to diagnose. So sometimes a shutdown fails, and the PC remains on with a black screen? POST does not run on a shutdown, but you see a code displayed? A POST code on a failed cold boot or restart, when POST runs is when the codes are displayed. If POST fails, a single code is displayed and stays displayed. Is that not happening in your situation? I ask because of the "68 or 0d etc." thing. POST codes that are shown and then disappear means that test passed. Going with 68, which is Chipset Initialization error, that is a difficult one, could be many things. The standard cure for a chipset initialization error is simply clearing the UEFI/CMOS or pressing the Reset button. No idea if you tried that, and if so, does it work even temporarily? What UEFI/BIOS version are you using? Do you get this behavior with say stock CPU and memory speeds? Your memory at 3200 is right at the edge of stability with a Ryzen system. My FlareX 3200 is not stable at 3200 with a different Ryzen board, I must use it at 2933. Have you tried a lower memory speed, say 2933 or 3033? Have you tried setting the three memory options you listed at their default settings? GearDownMode, per AMD rules, is enabled by default above a memory speed of 2667 (any memory OC), so Auto = Enabled. Setting it to Disabled is pushing it. Your SOC voltage at 1.200V might help. No information about the rest of your hardware, or Windows version being used, or AMD chipset driver version installed, and Windows Power Plan. Chipset problems can be related to hardware connected to the chipset, such as the drives in the system. If you were hoping for a well known, simple fix, I'm sorry but I don't have one for you. Your issue is rather unique, not something we often or at all see in Ryzen related threads. ------------- http://valid.x86.fr/48rujh" rel="nofollow"> 
|
Posted By: Haaviko
Date Posted: 31 Jul 2017 at 1:12pm
|
Well as I said the two ones I've seen are 68 and 0d, but yes so using all the settings I listed above.. GDM OFF, BGS Off with BGS Alt On and 1T I can boot into Windows and play, benchmark etc. with my 3.8Ghz OC and 3200Mhz XMP profile.. But If I turn off the computer using "Shutdown" or "Restart" in windows.. then once I either turn the computer back on using power button OR the system has been rebooted using Restart I see a black screen with 68 and fan revving. I'm using those settings because of this:  |
Posted By: Haaviko
Date Posted: 31 Jul 2017 at 1:14pm
| As you can see 1T CR, GDM=Off is a performance increase over GDM=On 2T |
Posted By: wardog
Date Posted: 31 Jul 2017 at 2:06pm
|
Set SoC to 1.2 and your mem to 3.75, checking that VTT right below it goes to 0.690v Post back please. |
Posted By: Haaviko
Date Posted: 31 Jul 2017 at 2:38pm
| Tried with your settings wardog.. now no BIOS or post just 06 on mobo with revving fan till I reset CMOS. |
Posted By: wardog
Date Posted: 31 Jul 2017 at 3:59pm
OK. SoC on Auto, SoC LLC 3, mem still at 3.75 and VTT 0.690 I'm off to bed so I'll be back in 6-7hrs |
Posted By: Haaviko
Date Posted: 31 Jul 2017 at 6:02pm
|
" rel="nofollow - Other information: Windows 10 64-bit with latest update, All drivers for the X370 Taichi from Asrock website including the AMD all-in-1 (Chipset Drivers), I did not install the AMD 17.10 Chipset from AMD themselves only the AMD all-in-1 directly from Asrock MOBO website page. The error was happening during posting not when loading windows, so you'd press the power on switch.. it would be a black screen with nothing on it.. look at motherboard and a bunch of codes cycle through the finally it would land on 68. EDIT/Update: I just downloaded latest AMD chipset drivers 17.10 which is newer than the Asrock Taichi All-In-1 and it seems to have no issue now? |
Posted By: Haaviko
Date Posted: 31 Jul 2017 at 9:00pm
|
" rel="nofollow - Nevermind.. after testing for hours (playing games etc.) even did some shutdown/restarts all was good now getting 0d :( You think going back to 2.40 BIOS might improve stability going for the GearDownMode + 1T?
|
Posted By: parsec
Date Posted: 31 Jul 2017 at 10:57pm
|
It's impossible to be certain if the 2.40 UEFI will fix your issue or not, but is worth trying. We've seen both positive and negative reports from users with the 3.00/AGESA 1.0.0.6a UEFI update. Suggestions for the 0d code, which is related to memory, the GPU or other devices, is to reseat the memory and GPU, and try removing extra USB and PCIe devices, if you are using any. ------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: wardog
Date Posted: 31 Jul 2017 at 11:33pm
" rel="nofollow -
Are you at a point with this build that beginning from scratch wouldn't be a PITA? |
Posted By: ket
Date Posted: 01 Aug 2017 at 1:15am
|
" rel="nofollow - POST code 0d is related to memory, VGA card, and USB devices. While it also covers misc PCI-E devices, its very unlikely any such devices would have a compatibility issue resulting in a POST code of 0d. I'd suggest you disconnect all the USB devices you can, test your memory with HCI memtest for at least 6 hours, then let us know the results. Experience tells me the problem is coming from something memory related so I'd start there. Is your memory dual sided (DS) or single sided? (SS). If you have the memory in DIMM banks A1+B1, or A2+B2, switch them to the unused banks and see if that helps. |
Posted By: datonyb
Date Posted: 01 Aug 2017 at 2:57am
i run the ram volts at 1.35 soc at 1.0 and soc llc at level 2 as well i have used these settings since bios version 1.6 cnq dis c6 dis core boost is still on (dont think it matters myself, i or indeed works when ive got it clocked) global c state dis i literally followed the guide from youtube search 'techcity' his ryzen overclocking video circa early april has the guide using a taichi bios system i dont feel i have a outstanding chip to be honest and for sure the taichi i have and the ram i have allthough do what they are supposed to do ,they dont have any extra headroom like a lot of guys are finding now with other ram sticks (even in fact another local guy here in london with identical age, brand, and part number ram and taichi board as myself who can get faster ram overclock) ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: Haaviko
Date Posted: 01 Aug 2017 at 9:32am
| Ran the HCI memtest and it said 0 errors on all 16 of them, I've already tried reseating the ram. |
Posted By: Haaviko
Date Posted: 01 Aug 2017 at 9:39am
|
I don't think its USB related either, because it runs 100% perfectly If I leave BankGroupSwap, BankGroupSwap Alt and GearDownMode all on Auto, this only happens when trying for.. GearDownMode: Off BankGroupSwap: Off BankGroupSwap Alt: On
1T |
Posted By: Haaviko
Date Posted: 01 Aug 2017 at 10:51am
| Switched to 2.40 BIOS and so far way less stability issues using my settings compared to 3.00 the only thing so far is I got an post error when rebooting.. but this time was a new one: 0E |
Posted By: Haaviko
Date Posted: 01 Aug 2017 at 9:13pm
|
So now back to 3.00 BIOS and yeah post code errors are happening either way.. turning off BGS or leaving it Auto, I think I've figured it out.. If I overclock my system in any way shape or form, that causes the errors while posting.. because I just did UEFI Defaults with ONLY XMP Enabled and BAM no issues all night restart, cold boot etc. I tried fixed voltage, pstate and offset and same errors eventually happen.. 68, 78 and sometimes 0d its random. |
Posted By: parsec
Date Posted: 01 Aug 2017 at 11:10pm
Sorry to say, that is why I suggested trying the Auto settings for those three settings to solve the problem. I understand why you want to use those settings. Have you tried one or two of those options at a time, set to those values? To see if you can have GearDownMode set to Off while the others are on Auto, for example. Another memory tweak according to AMD is actually slightly reducing the VDDP voltage. I assume you have you seen this AMD blog page about Ryzen and memory: https://community.amd.com/community/gaming/blog/2017/05/25/community-update-4-lets-talk-dram" rel="nofollow - https://community.amd.com/community/gaming/blog/2017/05/25/community-update-4-lets-talk-dram ------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: Haaviko
Date Posted: 02 Aug 2017 at 12:16am
Another issue I've been having and not sure what it could be.. I just did a reformat, fresh OS, Asrock x370 taichi drivers, latest non-BETA NVIDIA driver.. and while playing any video game I get this: (Only happens while gaming) |
Posted By: fouf
Date Posted: 14 Oct 2017 at 8:32am
|
Hey, I'm having the same issue. Any sort of overclock and I get error 68, 78, sometimes 0d and 0E.... When I can finally get into windows it runs rock solid. Did you end up fixing this.. would love to know.
|