p2.0 bios broken taichi x370
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=4930
Printed Date: 22 Dec 2024 at 1:57am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: p2.0 bios broken taichi x370
Posted By: Optik
Subject: p2.0 bios broken taichi x370
Date Posted: 23 Apr 2017 at 5:55pm
|
hi i want tell u some very important information about the p2.0 bios dont install it. p2.0 bios not stable smt disable dont work cpu voltage controll dont work boost my cpu at 2,7V, at p1.60 bios 1.168V https://valid.x86.fr/k0ccc0" rel="nofollow - https://valid.x86.fr/k0ccc0 ?p2.0 bios https://valid.x86.fr/1xvsvu" rel="nofollow - https://valid.x86.fr/1xvsvu p1.60 qvl ram dont start with ddr4 3200, soc voltage 1.25V dont help to fix it . i will try now how p.16 is working with OC/UV i hope it works fine. bios p1.60 work very fine ddr4 2933 with cl 14 timings work. |
Replies:
Posted By: datonyb
Date Posted: 23 Apr 2017 at 8:39pm
|
this is the second person to say this so i went and checked mine (allthough my cpuz dosnt report such high volts hw info does http://smg.photobucket.com/user/datonyb/media/vcore%20voltage%20bios%20v%202_zpshree9vvk.png.html" rel="nofollow">  ------------- [url=https://valid.x86.fr/jpg250]  [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: TooQik
Date Posted: 23 Apr 2017 at 9:06pm
|
I'm noticed that CPU-Z often shows my CPU Vcore jumping over 2 volts also but HWInfo64 has never shown the Vcore as more than 1.4 volts, so I'm inclined to believe that CPU-Z isn't reading the voltage correctly. Here's a screenshot of CPU-Z:  |
Posted By: datonyb
Date Posted: 23 Apr 2017 at 9:18pm
which is the opposite of what im getting  cpuz stays very fixed at 1.296-1.312 in bios i have fixed volts at 1.30 with LLC level 2 hwinfo shows min 1.296 but current at 2.5-2.6 ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
TooQik wrote:Posted By: TooQik
Date Posted: 23 Apr 2017 at 9:27pm
Are you loading HWInfo on startup or CPU-Z? I ask because I have HWInfo setup as my main monitoring tool and it starts with Windows. I expect that any other monitoring tool that is opened second will be the one that shows errant data values. |
Posted By: datonyb
Date Posted: 23 Apr 2017 at 9:38pm
|
neither run on start up ive just jumped back into bios and loaded setup defaults (eg pure and simple exactly as asrock would have any user run this board) now everything should tickover at minimal volts but it still peaks over 2 volts when i click cpuz bench just uploading the screen shot now i will edit it on shortly here the screen shot at 100% safe bios setup................... 'safe'  http://smg.photobucket.com/user/datonyb/media/running%20task%202%20volts_zps5wdh4egl.png.html" rel="nofollow">  ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: datonyb
Date Posted: 23 Apr 2017 at 9:49pm
|
" rel="nofollow - so the question is is bios version 2 for the taichi actually shoving 2.5+ volts into our ryzen cpu which is recommended to be 1.45 v max ? on a side note due to another post about the graphics cards running at x8 i had a look at the version 2 bios concerning that setting on the tachi and nowits been changed to auto/gen1/gen2 and gpuz reports mine running at 1.1 instead of 3 (ok update here it shows 1.1 unless i load the card with graphics task then switches to 3) im feeling at the moment that asrock can keep this bios version2 to themselves until a stage at which they smooth out any bugs (be it false reports or actual dangerous settings) ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: datonyb
Date Posted: 23 Apr 2017 at 9:58pm
|
my cpuz now reports between 0.5 volts and 1.1 volts depending on idle and running a quick cpuz bench while hardwareinfo has a max of 2.4 volts this issue does need priority responce i feel EDIT ok ignore anything above this thread for recordings of cpuz i was using version 1.78 and noticed you have later version 1.79 with ryzen support now im much less happy cpuz reports volts between 1.6 and 2.6  http://smg.photobucket.com/user/datonyb/media/2000%20volts_zps1jbmwvjt.png.html" rel="nofollow">  ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: datonyb
Date Posted: 23 Apr 2017 at 10:23pm
|
i have just posted this link to this discussion to bryan at techcity.tv who seems to know these baords very well and is indeed the guy we keep linking to with an overclock guide for the taichi hopefully he can help with some insight and a view ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: parsec
Date Posted: 23 Apr 2017 at 11:50pm
|
Thanks for reporting this guys. I hope the insanely high CPU-Z VCore reading are wrong. I can't believe that they are right. Actually, I don't. Just to let you know, with my X370 Killer SLI/ac board, using a 1700X, I see CPU-Z showing a VCore of over 2V at times, and values like 1.7V, 1.8V too. Running HWiNFO at the same time, I don't see a VCore over 1.280V, and the VID readings are 1.219V. For those that use HWiNFO64, in the Sensor Status display, please find the VRM readings. My board shows two sets of VRM information, both whose title shows IRF IR35204 and the name of the board. One is for the CPU's VRM, the other for the SOC's VRM. It shows temperatures, VR VROUT and VRIN, etc. I imagine you'll get this with the X370 Taichi, and possibly more. You can tell which of the two sets of VRM readings is for the CPU and SOC by checking the VR VOUT reading. The SOC VR VROUT will be ~0.900V or a bit less. The CPU VRM VROUT will be the VCore, which stays constant. Here is an example of CPU-Z 1.79 and HWiNFO64 v5.51-3141 running at the same time. Please check the VRM VROUT readings compared to what I captured at the same time from CPU-Z:  I've modified the title/heading of the VRM data readings in HWiNFO, so it won't exactly match the default titles. I have the VCore set to 1.28125V in the UEFI, a preset value. The CPU VRM IRF IR35204 VR VOUT reading of 1.279V closely matches my VCore setting of 1.28V in the UEFI. Above this I have the VCore reading in bold, whose maximum is 1.280V. I have CnQ and C6 enabled in the UEFI, which I assume accounts for the low Minimum and Current VCore values you see. The CPU-Z VCore value was not static at 2.560V, but I caught it at that value with Windows Snipping Tool. I wanted to show you that I get that with my board too. If you take the VR VOUT value of 1.279V and multiply by 2, you get 2.558V. Multiplying the Maximum VCore value I have of 1.280 by two, is exactly 2.560. CPU-Z 1.79 which is pretty new, is the first one to support Ryzen. CPU-Z 1.78 would not even start on my Ryzen PC, it would lockup at 65%, and could not be killed with Task Manager. HWiNFO64 has had the ASRock Ryzen boards supported for weeks, with only minor updates to support specific boards from various manufactures. Personally, I trust HWiNFO64, since so much of its Sensor display matches what I know should be correct. Sensor readings are not always direct readings, they are digital data and must be configured/calculated to get the correct reading to display. A reading being wrong by a factor of 2 or 1/2 is a common mistake when preparing a data reading for display. For example, it took CPU-Z several attempts/releases to get the Intel Haswell processor VCore reading (with that processor's on-die voltage regulators) right. IMO that is what we are seeing here again, a mistake in CPU-Z. Do you really think your Ryzen CPU could survive having 2.5V+ applied to it for even a short period of time over and over again? I don't. You can of course verify my readings and few calculations for yourself. I'm just reporting what I see in HWiNFO64 vs CPU-Z. Believe me, I am relieved that, IMO, the CPU-Z VCore reading is wrong. So go ahead and tell me why I'm wrong...  ------------- http://valid.x86.fr/48rujh" rel="nofollow"> 
|
Posted By: Optik
Date Posted: 24 Apr 2017 at 12:39am
|
Only a-tuning tool says 1,35v with p2.o bios other software show me 2,7. Smt disable don't work, dbl vcore settings. P1.60 working best Asrock pls fix it. And don't make bad option like in p2.0. Only what u must do put agesa code into P1.60. |
Posted By: JohnV
Date Posted: 24 Apr 2017 at 12:42am
|
Hey guys, just going to throw my 2-cents into this thread and say that this is the same with the professional gaming 2.0 bios. And that the CPU-Z / HWmonitor numbers for vcore don't line up with either HWiNFO64 or http://www.ocbase.com/" rel="nofollow - OCCT v4.5.0 . While running a 1700 @stock the voltages run 0.36v min to 1.28v max values on the high performance plan, as expected of a properly functioning R7. Those are the readings I get with HWinfo and OCCT. I also wanted to add this tidbit of information from over on the amd forums about Ryzen Master also not being entirely correct and an http://community.amd.com/thread/214762" rel="nofollow - alternate explanation ... or shall we say, alternate facts  Lastly, Parsecs theory on what may be happening is probably correct, and I'm sure we'll eventually see some more software updates, as well as bios updates, which will give us a clearer picture of what is actually going on inside of our new processors. Thanks. |
Posted By: datonyb
Date Posted: 24 Apr 2017 at 1:29am
|
ok so it seems panic over (to be fair 2.5 volts should have caused some user noticable error like a dead pc ) but its nice to know exactly what is what heres the screen shot of hwinfo 64 http://smg.photobucket.com/user/datonyb/media/hwinfo64_zps42b80emc.png.html" rel="nofollow">  ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: datonyb
Date Posted: 24 Apr 2017 at 8:06pm
|
just as an update i have had a reply from asrock vendor rep on the support forum here in the uk quote 'We can assure you that 2.6V is way beyond what the board's overvoltage protection will even allow. Simply a software measurement error, no cause for worry.' unquote i have asked if this issue with bios 2 and software from cpuz (eg cpuz AND hwmonitor from cpuid) can be looked into ,so as to prevent over reporting volts ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: TooQik
Date Posted: 24 Apr 2017 at 10:20pm
|
Yes, I wasn't worried by the high voltage as I figured I'd either be smelling something burning or the PC would have shut down by now if it was actually the case. Here's another screenshot of a voltage HWiNFO reports incorrectly for my PC:  That's one hell of a 12 volt power supply.  |
Posted By: Optik
Date Posted: 24 Apr 2017 at 10:50pm
|
Install p1.60 and u don't have problems. P2.0 Smt disable don't work Voltage reading don't work correctly 2 settings 4 voltage settings the Mainboard calculate both an say 2,7v not 1,35v. Pls fix it Asrock. |
Posted By: datonyb
Date Posted: 25 Apr 2017 at 12:19am
while i agree with you saying please fix it asrock i cant agree to blindly just saying install v1.6 with that you lose agesa 1004 which carries significant postives from my point of view (which may well differ from your for our own reasons) my better option i have decided is to stick with v2 (i dont need to turn off smt,i have no issues with my ram working ) so im MY case im better off with v2 other people may well be suited to version 1.6 or the beta they were using 1.9X ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: Optik
Date Posted: 25 Apr 2017 at 2:27am
My system running fine at the moment 3,8ghz 1,35v soc voltage 1,0v with ddr4 2933 cl 14 timings 1,36v. Aots running very fine and smooth |
Posted By: datonyb
Date Posted: 25 Apr 2017 at 6:35am
|
well thats good then you may find you can lower the cpu vcore a little more i am at 3.9 and 1.30 volts it is very stable .and runs cold ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: Optik
Date Posted: 25 Apr 2017 at 8:33am
defaul vcore 1.168 allcoreturbo 3,5 ghz, max cpu vcore 1,55 if u run prime to test stabilty its useless try boinc asteroids@home and i have g.skill flare x 3200 and dont running with that setting what da hell qvl list u kidding me.
|
Posted By: TooQik
Date Posted: 25 Apr 2017 at 8:51am
I don't gain anything by going back to 1.60. The only "problem" I currently have is I can't get my RAM to run at speeds higher than 2400. This is the same for both the 1.60 and 2.00 BIOSes so I'm better off sticking with 2.00 until the next AGESA update in May. Patience is a virtue, although these days it's more and more a lost virtue. Too many people are part of the NOW Generation. |
Posted By: Optik
Date Posted: 25 Apr 2017 at 9:05am
" rel="nofollow -
QVL RAM? Ryzen dont like 13 15 17 timings and dude dont worry all other boards have the same problems with ram
------------- Taichi X370 AMD R7 1700X Arctic Liquid FreeZer 240 G.Skill Flare X 3200 cl 14@2933 PowerColer RX480 Phanteks EnTHoo Luxe BeQuiet Dark Power Pro 11 550W |
Posted By: TooQik
Date Posted: 25 Apr 2017 at 9:19am
" rel="nofollow -
Until the RAM issues are sorted out the QVL is a guide more than a verified list of what works. I'm aware that all vendors are having the same issue, that's why I'm not worried and will patiently wait for the next AGESA release before updating my BIOS. |
Posted By: parsec
Date Posted: 25 Apr 2017 at 9:37am
I had that +12v (over 22V) reading too with that version of HWiNFO64 with my Ryzen system. I guess you did not see the +12V reading in my HWiNFO64 screenshot in my previous post on page 1 of this thread. That's with the v5.51-3141 Beta version, which fixes that bug. I like to use the Beta versions since they have the latest features, and rarely have any issues. You can also see sensor readings that are meaningless, like the 7° C Auxiliary reading that never changes. That +12V display error is a great example of how the outputs of the sensor/monitoring chip must be processed to provide the correct value. All boards don't use the same monitoring chips, and good programs like HWiNFO are coded to determine which monitoring chip and board is being used. From there the appropriate processing is applied to the monitoring chip's output. Yes, that's a lot of work and support, which is why some programs, even famous ones like CPU-Z, have bugs in them occasionally. I've casually worked with the programmer of HWiNFO to fix a bug in it occasionally, when new boards I have are released. You can read in the HWiNFO News section (basically release notes) how enhancements are added for various boards in different versions. This is scary stuff really, we depend on these programs and assume they are always correct. Their programmers do a great job, but no one is always perfect, and since there a few standards about how this all works, it's not easy getting everything right all the time. ------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: TooQik
Date Posted: 25 Apr 2017 at 10:13am
|
Thanks for the heads-up on the beta version having fixed this bug parsec. I can appreciate the amount of work involved from all sides when new products are released. That's why I try to get information from a number of sources before looking for a solution to a problem I don't really have. When I first noticed the high 12 V reading in HWiNFO I rebooted, went into BIOS and checked the HW Monitoring values, these were all normal so I figured HWiNFO wasn't reading the voltage correctly. The hardware manufacturer usually knows how to work with their hardware better than a third party.  |
Posted By: parsec
Date Posted: 25 Apr 2017 at 10:48am
|
Yes, the hardware manufacture ought to know how there own product works... they better. You've brought up another situation that the third party programmer must deal with, as the HWiNFO programmer has mentioned in his forum. He has asked mobo manufactures for information about their sensor/monitoring designs, and is refused. He must figure it out himself, by using his experience and when possible the documentation for the monitoring chip from its manufacture. That still does not give them every detail, such as which sensor inputs or outputs are not being used for example. That's where owners of various boards can help them get things right. HWiNFO64 always has at least a few sensor readings that are meaningless, but that depends on the board manufacture, type of system (AMD or Intel), chipset, the board itself (high end or entry level), etc. Plus how can they have one of every mother board made by every manufacture to work with? Recently ASRock began using a new monitoring chip on a few of their boards, and most of the existing monitoring programs could display only a few readings correctly, until the programs were updated (if they ever were updated.) ------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: Bill H
Date Posted: 25 Apr 2017 at 1:15pm
|
I am liking this thread. Let me say, before I go any further that I have a Professional Gamer. which is pretty much the same as the Taichi(missing 5Ghz ethernet). I did notice right away that BIOS version 2.0 does have voltage fluctuation problems.
I do like that I can save BIOS tweaks tho. I have been seeing voltages way above the normal range. CPUID HWMonitor and HWiNFO64 as well as CPU-Z are indicating a huge voltage fluctuation in CPU VCORE. From 1.392v to 2.8v. I went back to v1.60 and the problem went away. With v1.60 I get a great overclock to 4.1GHz and mem speeds of 3200MHz 14-15-15-15-36-1T. Using OFFset mode for the CPU and SOC voltages works well along with LLC I will just have to wait for ASRock to figure it out. As for now, the only thing I am missing is being able to save my configurations. ------------- X370 Pro Gamer - 1800x@4000 - EVGA GTX 1080 Hydro Copper SLI - G.SKILL F4-3000C14D-16GTZ@3200MHz - Single liquid loop |
Posted By: datonyb
Date Posted: 25 Apr 2017 at 8:09pm
|
what is good is real people with real setups posting what they can do and their settings it helps educate us all p.s. i like your ram choice (it does very very well on intel systems hopefully it will eventually show same promise on the am4) ------------- [url=https://valid.x86.fr/jpg250] [/url]3800X, powercolor reddevil vega64, gskill tridentz3866, taichix370, evga750watt gold |
Posted By: Optik
Date Posted: 25 Apr 2017 at 9:38pm
|
At weekend I will try 3,6ghz with undervolting. 3,6 GHz 3,7ghz is ryzen sweetspot ------------- Taichi X370 AMD R7 1700X Arctic Liquid FreeZer 240 G.Skill Flare X 3200 cl 14@2933 PowerColer RX480 Phanteks EnTHoo Luxe BeQuiet Dark Power Pro 11 550W |
Posted By: Lisanderus
Date Posted: 30 Apr 2017 at 11:58am
| Hwinfo64 5.51-3141 (beta) work fine for me, previous version and latest CPU-Z (1.79) showed a doubled voltage. The only thing I noticed is that it does not always show a drop in frequency in idle time (to ~1500mhz, always p0 - 3700), voltage drops. |