Overclock.net › Forums › AMD › AMD Motherboards › Asrock X370 Taichi Overclocking Thread
New Posts  All Forums:Forum Nav:

Asrock X370 Taichi Overclocking Thread - Page 229

post #2281 of 2553
I'm pushing 1,418 Vcore, LLC3 and 1,45 on DRAM for 4.0&3466C14 on my 1600X. It's been stable so far after 1750% HCI and couple of hours of P95. I'm gonna wait on my new intake fans before longer P95 testing since it's been getting toasty(78C).

I needed to bump tRC to 52 for it to get stable since it wasn't, even with 1.46 on DRAM.
It doesn't even boot with increased SoC (1.1) so I'm sticking with 1.05V.
I put in 1.8V for PLL and 1.03V for Vprom_1.05 as per chew's recommendations since the board's overvolting those a bit, might try 1.75 PLL.
Edited by Brightmist - 10/23/17 at 3:48am
post #2282 of 2553
Quote:
Originally Posted by drkCrix View Post

Since you own this board now I suggest reading up on it
http://www.overclock.net/t/1624603/rog-crosshair-vi-overclocking-thread

If you look under bios flashing instructions and then under bios you will see 1401 between 1201 and 1403. So yes there is a 1401 bios.

Have fun in the Asus thread.

I was going by what's on the offical website: https://www.asus.com/us/Motherboards/ROG-CROSSHAIR-VI-HERO/HelpDesk_BIOS/
post #2283 of 2553
Thanks for the 1usmus Ryzen DRAM calc I was able to achieve stable 3200 on my dualrank F4-3200C14D-32GTZR 2x16gb kit with little to no problems
I also got better timings compared to the stock XMP profile. It seems that I'm all set currently. And it seems that in fact instability I got at the beginning was more due to wrong
cad_bus and rtt settings, not just SoC / VDDP / DRAM voltages. I've read a couple of times that I could try to bruteforce a plenty of VDDP values going up and down but it appears that it was unnecessary as I'm basically on very similar VDDP that was on auto (0.9 instead of 0.92, but changing that alone didn't improved stability when I was trying to figure out the problem)
ProcODT is on 68.8 80.
BIOS freezes is gone (or maybe I got lucky and just didn't had one in a three or four days) Pictures (Click to show)



But I'm curious - will I need to reset the settings and remove one RAM stick as somebody recommended previously to ensure stability before updating to the next bios (when it will appear)?

IMPORTANT NOTE: this have few corrections. Few people asked me for more thorough testing that I ignored (I was tired from long memtest runs and last times just stayed with quick Linpack testing and just making my routine tasks like gaming.
You should be able to run on exact same settings (in case of same hardware), I had no cold boot, reboot issues, BSODS, app or games crashes, etc.
But memtest shows a couple of errors very soon (at 30+ percent)

I fixed it by changing ProcODT to 80 and tRFC to slightly higher value (Ryzen DRAM calc 0.9.6 v8 shows slightly higher tRFC as an alternative)



Voltages are based on the same calc screen (VDDP 0.9, SoC 1.05, CLDO_VDDP 0.425, DRAM 1.4)
Edited by Korrektor - 10/24/17 at 5:56pm
post #2284 of 2553
Quote:
Originally Posted by Korrektor View Post

Thanks for the 1usmus Ryzen DRAM calc I was able to achieve stable 3200 on my dualrank F4-3200C14D-32GTZR 2x16gb kit with little to no problems
I also got better timings compared to the stock XMP profile. It seems that I'm all set currently. And it seems that in fact instability I got at the beginning was more due to wrong
cad_bus and rtt settings, not just SoC / VDDP / DRAM voltages. I've read a couple of times that I could try to bruteforce a plenty of VDDP values going up and down but it appears that it was unnecessary as I'm basically on very similar VDDP that was on auto (0.9 instead of 0.92, but changing that alone didn't improved stability when I was trying to figure out the problem)
ProcODT is on 68.8.
BIOS freezes is gone (or maybe I got lucky and just didn't had one in a three or four days)

But I'm curious - will I need to reset the settings and remove one RAM stick as somebody recommended previously to ensure stability before updating to the next bios (when it will appear)?

Can you run HCI memtest without issues? I have the exact same ram and capacity and even with lower settings I get immediate errors (12 instances, since I use an 1600x). Especially at ProcODT 68,8. Lowering it to 60 in fact increases stability but I got one error with around 300% coverage.

I am now trying CR 2, even though I do not like such a low setting. If you are indeed stable with your memory, can you tell me what exact cad_bus and rtt settings you are using? Also your voltages, etc.?

Receiving stability with your settings would be a dream....
post #2285 of 2553
HI all, first time posting here,i hope you can help. I have a 1700 for 2 months now with taichi and 16gb of 3200c14gfx. Overclocking has been rather easy due to most of you in here (thanks for that) but problems started when i moved countries 2 weeks ago and took my desktop with me.
There were some weird lag issues, when i first set it up after the trip ,even in bios. boot time was taking ages so i decided to reseat everything and do a clean windows 10 installation(1709 build).
And then all HELL broke loose!
Windows, browsers, software , mouse, keyboard are all randomly lagging and freezing. The disturbing thing though was when installing drivers after a restart i saw a '00' error code while everything was at stock confused.gif
My chip is week 22 so i dont know if that plays a part as i remember there was an issue with chips before week 25.

Thanks in advance !
post #2286 of 2553
@Amir007; I am quoting your post in spoiler as it is very long. I am neither defending Asrock nor agreeing with you completely. Just take this another point of view.

I understand your disappointment with Asrock x370 Taichi. I have had problems with my build too. But I am not sure if Asrock is to blame here.

1.) CPU temps got reported at 256°C for an extremely short moment.
2.) I got really high temps on my Ryzen 1700 CPU: vcore 1.25; 3700MHz; LLC2 or LLC3... close to 80°C after several hourse of Prime95 or IntelBurnTest. I tried lots of different things: Reseating the cooler and reapplying the thermal paste, trying different coolers but all of my tries to no avail. But when I increased the voltage with "offset" mode, the highest temperature (the peak) did not increase at all. Very strange.
3.) If I use "fixed voltage" and set my voltage over 1.35v, the computer significant lowers my CPU clock. Most likely the X370 Taichi is to blame here. I found a workaround by researching the internet: Use the offset mode. Despite the fact that I solved it, it is still annoying to have this bug.
4.) I was never able to get 4.0GHz stable with my Ryzen 1700, even at 1.45v. The CPU can be the cause here. Again, just presenting what I experienced.

But all things above considered, the thing is that most X370 motherboards have similar or the same problems. As this is a new chipset, I can understand problems can more often exist but they got solved it over time with BIOS updates. However, there are/is issue(s) that persist(s) and this is disappointing.

As for my third point (#3), there is still no fix from Asrock as of 3.20 BIOS.

As for your "black screen" or 00 thing, it can also be your RAM sticks. I am not sure tough because black screen when playing games etc. usually means the CPU stability is to blame.

Quote:
Originally Posted by Amir007 View Post

Warning: Spoiler! (Click to show)
Hey kmac - Do yourself a favor...(I should've listen to 1 or 2 of the guys on here who told me from the get-go and before my chin hair turned gray) and go exchange your ASSrock Taicrappy mobo for another top of the line motherboard such as Asus CROSSHAIR VI Hero X370. I did a lot of research last night and BOY I'm a such happy camper now. Not sure if you read my recent ASSrock issues I was having in the past few days but If i hadn't done benches via CPU-Z and Cinebech I wouldn't have discovered that my 1800x was running at 4 core performance. This was by far the single worst god damn experience I've ever had with a motherboard. That Taichi + 1800x was a nightmare to get it to work stable in default settings. Why should ANYONE need to play around with the bios to get a stable working environment without the OC attempts? ASSrock is aware of this issue and it seems as if they are not sure what they are doing or if it's fixable.

It is almost as if I somehow booted up using x1800 on an Intel Assrock Taichi mobo, no joke man. I too was getting exactly what you are getting when playing games. My monitor would just turn off and would get a "00" debug code, this means CPU isn't working right due to a flaky motherboard's bios. This would also happen if/when I would run Cinebench, my screen would freeze and black screen and 00 debug code. The funny part is I had everything running at default. My memory was at stock 2400mhz and tested stable in memtest86 for 6 hours so memory was fine and most likely the CPU was fine too. Don't waste your time and return that pathetic buggy motherboard. I kind a feel sorry for people who invested in this mobo and i bet half of owners probably are not even aware their 1800x is running half speed. lol I agree the mobo looks good and looks good on paper and you get a lot for the money but if the soul of the mobo (bios) isn't ripe then it has no stable personality. I felt like i was working with a mobo with down syndrome. Return that POS dude. It's not your 1080 bcuz i have a 1080 too and runs perfectly. I can guarantee you if you swap it out tomorrow all of your troubles will go away. I mean, even chew dude on here proved the mobo needs a lot of fixing, and most likely it'll probably need a new PCB revision as Microcenter had 6 of them for sale as open box and none in stock as new. I asked the agent there and he even said that they are not sure why they haven't gotten any new stock for a over a month now. You know why? ASSrock probably knows they were selling bricks/lemons. They wasted so many hours of my time I feel like suing them for it. God damn Pathetic.

Look at me now:



Notice how I'm not anymore getting the multi bug at 550Mhz like i was before during 100% load. And also my max turbo is also working perfectly at 4.1Ghz as well. This wasn't present on the ASSfhuckachi at all.

Before:





Microcenter has a very nice deal going now - even better than when I got the Taichi. You receive a $50 instant discount if u purchase mobo+cpu. I ended up returning the x1800 as well and got me a full microcenter store credit. I ended up paying only $24 more to get this fully operational motherboard + 1800x. Total was $539.98 + Tax brand new and sealed. You will not need to mess around with the bios to be able to run 100% stable. Remove it from your case, dump it back in the box and divorce it. smile.gif

Here are my BIOS settings for my R7 1700 and X370 Taichi and Trident Z 3200 CL14 2*8GB (16GB):

OC Tweaker settings:
Overclock mode (do not confuse with "voltage mode"): Auto
CPU Frequency and Voltage Change: Manual
CPU Frequency: 3900MHz
CPU Voltage: Something like 1.18; this is the default value when you change "CPU Frequency and Voltage Change" to "Manual"; I will report back when I get the exact number later (changing this value has no effect as far as I can remember.)
Voltage mode: OC Mode
CPU LLC: 2
CPU Vcore voltage: Offset mode
Offset voltage: 0.25v (With this value, various software under windows 10 reports my voltage is 1.425v.
Also DRAM Voltage is set to 1.35 after choosing XMP Profile (which is normal)

CPU Configuration (under "Advanced)":
Everything is disabled here.

Zen Common Options:
Core Performance Boost: Disabled
Global C-state Control: Disabled

RAM Settings under OC Tweaker settings: (I chose the XMP profile but then only changed these two settings)
Gear Down Mode: Disabled
Power Down Mode: Disabled

P-States:
Every p-state except p-state0 is disabled.
Gaming Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD R7 1700@3.9GHz@1.425v Asrock X370 Taichi Evga GTX 1080 Ti SC2 G.SKILL Trident Z DDR4 3200MHz C14 
Hard DriveCoolingOSMonitor
WD Desktop Black 1 TB Nzxt Kraken x62 Windows 10 Asus ROG Swift PG278QR 
KeyboardPowerCaseMouse
Razer Blackwidow 2013 Seasonic FOCUS+ Gold 650W NZXT S340 Elite Razer Mamba 2012 
Mouse PadAudio
Razer Vespula Logitech G930 
  hide details  
Reply
Gaming Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD R7 1700@3.9GHz@1.425v Asrock X370 Taichi Evga GTX 1080 Ti SC2 G.SKILL Trident Z DDR4 3200MHz C14 
Hard DriveCoolingOSMonitor
WD Desktop Black 1 TB Nzxt Kraken x62 Windows 10 Asus ROG Swift PG278QR 
KeyboardPowerCaseMouse
Razer Blackwidow 2013 Seasonic FOCUS+ Gold 650W NZXT S340 Elite Razer Mamba 2012 
Mouse PadAudio
Razer Vespula Logitech G930 
  hide details  
Reply
post #2287 of 2553
@silent12 The only thing I can suggest to hold you over until the smarter people on here reply is to go into your bios and load defaults and then do a clean install of which ever version of Windows you were using before clean installing the Fall Creators Update. FCU has created several problems for a few people so perhaps you are one of the unlucky ones.




edit: the pre week 25 you are referring to is the segfault issue when compiling code on Linux.
Edited by pschorr1123 - 10/24/17 at 10:23am
post #2288 of 2553
@silent12, I suspect RAM or mobo. Can you remove one of your RAM sticks (if you are using two sticks) and try with each of them separately?

It is also likely that both of your sticks are faulty (although probability lowe than %1). So, it would be great if you could try with a different RAMs. I know that most people including me do not have extra RAM/mobo/PSU but can you get them tested by a friend, acquaintance or a computer store offering some sort of service?
Gaming Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD R7 1700@3.9GHz@1.425v Asrock X370 Taichi Evga GTX 1080 Ti SC2 G.SKILL Trident Z DDR4 3200MHz C14 
Hard DriveCoolingOSMonitor
WD Desktop Black 1 TB Nzxt Kraken x62 Windows 10 Asus ROG Swift PG278QR 
KeyboardPowerCaseMouse
Razer Blackwidow 2013 Seasonic FOCUS+ Gold 650W NZXT S340 Elite Razer Mamba 2012 
Mouse PadAudio
Razer Vespula Logitech G930 
  hide details  
Reply
Gaming Rig
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD R7 1700@3.9GHz@1.425v Asrock X370 Taichi Evga GTX 1080 Ti SC2 G.SKILL Trident Z DDR4 3200MHz C14 
Hard DriveCoolingOSMonitor
WD Desktop Black 1 TB Nzxt Kraken x62 Windows 10 Asus ROG Swift PG278QR 
KeyboardPowerCaseMouse
Razer Blackwidow 2013 Seasonic FOCUS+ Gold 650W NZXT S340 Elite Razer Mamba 2012 
Mouse PadAudio
Razer Vespula Logitech G930 
  hide details  
Reply
post #2289 of 2553
I had one Taichi + 1700 running fine at 4.0 GHz since release.

Unfortunally, i sold it last week.

The thing is, i have just bought the same combo again, but new.

But this time...i cant get past 3.9 GHz. The mobo just downclock the chip to 1500 and lock the voltage at 1.2V.

Seems really strange. BIOS version was the 2.2, so this mobo must be a older revision like the one i had, so i dont understand why everything was working fine with my "old" combo, and with this one i am having this stupid bug.

Any help?
Edited by Tasm - 10/24/17 at 10:52am
post #2290 of 2553
Thanks for the answers,

@pschorr1123 I did set everything at default before the clean installation but i will try again with a previous version of windows.

@Rakanoth i will try the ram sticks,yes they are two 8gb SR. unfortunately having moved countries it will be difficult to find any spare DDR4 sticks but i will give it a go.

Having googled though, most people with the same issues said the hard drives was the culprit. I was thinking to upgrade to a 960 evo anyway so that might fix it ?

I currently have a TS256GSSD370 bought like 3 years ago and a 840 evo 82% remaining life according to HWINFO
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD Motherboards
Overclock.net › Forums › AMD › AMD Motherboards › Asrock X370 Taichi Overclocking Thread