Overclock.net › Forums › AMD › AMD Motherboards › ROG Crosshair VI overclocking thread
New Posts  All Forums:Forum Nav:

ROG Crosshair VI overclocking thread - Page 814

post #8131 of 17652
Quote:
Originally Posted by Gadfly View Post

That is really high... I personally wouldn't go over 1.15.... As I understand it 1.2 and above is risking your CPU.

Actually we don't even know what to believe anymore. Gigabyte is even saying 1.25v is fine.

http://forum.gigabyte.us/thread/886/am4-beta-bios-thread

"Question: Using F5G and XMP my SOC voltage is set to 1.25! Is this normal/safe?
Answer: Yes, 1.25 should be safe. We increased the default AUTO voltage for stability on certain DIMMS since it seems most users have to set atleast 1.2."
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
post #8132 of 17652
Quote:
Originally Posted by bluej511 View Post

Actually we don't even know what to believe anymore. Gigabyte is even saying 1.25v is fine.

http://forum.gigabyte.us/thread/886/am4-beta-bios-thread

"Question: Using F5G and XMP my SOC voltage is set to 1.25! Is this normal/safe?
Answer: Yes, 1.25 should be safe. We increased the default AUTO voltage for stability on certain DIMMS since it seems most users have to set atleast 1.2."
the issue wasn't the cpu... it was frying the embedded controller on the motherboard with high soc voltages. from what it appears, asus patched it to at least handle up to 1.2v's safely with 1.25v's being so-so.

cpu wise yeah i don't think any official statement has been released about what IS safe 24/7, three+ years soc voltage like AMD did with cpu voltage.
strix
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD Ryzen R7 1800x Asus X370 Crosshair Vi Hero EVGA 1080 FTW G.skill Flare X 3200 2x8 16gb 
Hard DriveHard DriveHard DriveCooling
2x Western Digital Black 2TB 2x 512GB Samsung 850 Pro 1x 512GB Samsung 950 Pro Noctua NH-D15S + 2x Corsair ML120's 
OSMonitorKeyboardPower
Windows 10 Education Dell S2417DG Ducky One White & Grey - Cherry Reds Evga Supernova G3 750 
CaseMouse
Corsair Air 740 Steelseries Rival 700 
  hide details  
Reply
strix
(14 items)
 
  
CPUMotherboardGraphicsRAM
AMD Ryzen R7 1800x Asus X370 Crosshair Vi Hero EVGA 1080 FTW G.skill Flare X 3200 2x8 16gb 
Hard DriveHard DriveHard DriveCooling
2x Western Digital Black 2TB 2x 512GB Samsung 850 Pro 1x 512GB Samsung 950 Pro Noctua NH-D15S + 2x Corsair ML120's 
OSMonitorKeyboardPower
Windows 10 Education Dell S2417DG Ducky One White & Grey - Cherry Reds Evga Supernova G3 750 
CaseMouse
Corsair Air 740 Steelseries Rival 700 
  hide details  
Reply
post #8133 of 17652
Quote:
Originally Posted by bluej511 View Post

"Question: Using F5G and XMP my SOC voltage is set to 1.25! Is this normal/safe?
Answer: Yes, 1.25 should be safe. We increased the default AUTO voltage for stability on certain DIMMS since it seems most users have to set atleast 1.2."
If they have decided to define safe maximum limit voltages basing on user experiences, then we are all' doomed. biggrin.gif
post #8134 of 17652
Quote:
Originally Posted by hotstocks View Post

This bios has about 237 extra settings that no one ever needs to change. Yet every other motherboard bios I have owned in the last 12 years has had T1/T2 dram settings, yet we need two different bioses to choose? ***?

Quote:
Originally Posted by HeliXpc View Post

LOL, I was thinking the same thing, BUT I think this is an AMD limitation because of the architecture

I don't think its too bad. Could change in the future. Remember this is all rushed to market!
post #8135 of 17652
Quote:
Originally Posted by Reikoji View Post


Remember this is all rushed to market!
Damn ******* true.
I was wondering what kind of reviews would have had these CPUs of they were not rushed so much and would have launched in May-June '17.
They would have blown out every single reviewer in every aspect.
I have been using this new platform since launch, all I can say is that after having found the right settings and conditions to working with it is absolutely damn good.
X99 platform is overcomed in every aspect (price/performance firstly).
post #8136 of 17652
Running 1002 today, should 0079 be next for me?
post #8137 of 17652
Quote:
Originally Posted by SirMacke View Post

Running 1002 today, should 0079 be next for me?

0082 or 0083
Edited by Reikoji - 4/9/17 at 11:55pm
post #8138 of 17652
just a short checking earlier, where is bios 1101 is it recalled?
post #8139 of 17652
Thread Starter 
Glad to see most of you having a positive experience with the new bioses. AMD has several updates coming for memory in May timeframe as you guys already might have heard so stay tuned.

I made a small collection of tools for C6H which you can run at startup to change the behavior or solve some issues. I also included updated versions of the tools previously published.

c6h_tinkering_v1.zip 2353k .zip file

c6h_tempsrc changes the SIO temperature between Asus calibrated (default, mode 1) and same as Tctl (mode 2). This version fixes the error when SIO Clock Skew is Disabled.

c6h_qcode enables/disables the Q-code display

c6h_p0vidfix We had some issues with the BIOS bugfix, so I'm releasing a temporary OS fix for getting stuck in P1 state if changing P0 VID.

Quote:
Originally Posted by majestynl View Post

Below my first test-results and feedback with current bios versions:
After several hours testing i do have mixed feeling for now. Need more time to test better!
The new bios version are definitely better if it comes to bug-fixes/compatibility!
But in same time i needed more vcore to past some tests/benches! Need definitely more
time for fine-tuning and testing the new bios features/options.

I'm now sticking with 0081, because this version will be the next release.
Going to fine-tune and play around! Will update more info later.

What i used for basic testing:

  • IBT 10 Passes Standart
  • Realbench stresstest 15-30minutes
  • AIDA64 system-stress-test 15minutes
  • AIDA64 Bench Memory
  • Cinbench R15

Testing Rig:

  • CPU: 1800x
  • Memory: 16GB (2x8) Gskill F4-3200C14D-16GTZR
  • Cooling: Kraken x62
  • PSU: Cooler Master v850
  • VGA: RX480

Before these tests i used Bios version 1002, and was stable & WHEA error free with following settings:

  • OC Type: Pstates
  • Clocks: 4000Mhz
  • Dram: 3200Mhz
  • Vcore: +offset 0.0625v
  • LLC: Level 1
  • Dram: 1.35v
  • Vscoc: 1.15v
  • Timings: 14 14 14 14 34


Bios version: 0079:

- Needed more vcore to pas IBT (+offset 0.6825v/ Not checked WHEA errors)
- Cold boot issue better, not completely fixed without some magic
- Faster Boot time
- ~5% performance boost
- No magic for 3200mhz ram boot
- Core VID HWiNFO Fixed


Bios version: 0081:

- Needed more vcore to pas IBT without WHEA Errors (+offset 0.6825v) + LLC2
- Cold boot issue better, not completely fixed without some magic
- Faster Boot time
- ~5% performance boost
- No magic for 3200mhz ram boot
- Core VID HWiNFO Fixed
- Changed to 1.4v RAM for better results


Bios version: 0082:

- Needed more vcore to pas IBT without WHEA Errors (+offset 0.6825v) + LLC2
- Cold boot issue better, not completely fixed without some magic
- Faster Boot time
- ~5% performance boost
- No magic for 3200mhz ram boot
- Core VID HWiNFO Fixed
- Changed to 1.4v RAM for better results

Bios version: 0083:

- Needed more vcore to pas IBT without WHEA Errors (+offset 0.6825v) + LLC2
- Cold boot issue issue not solved! Re-Plug AC cable gave me post-cycles
- "Overclocking failed error" every time when i reboot with case-button
- Faster Boot time
- ~5% performance boost
- Core VID HWiNFO Fixed
- Needed 2 tries before got 3200mhz booted
- A bit slower in AIDA64 Memory Benchmark


Below you can find screenshots i made during the tests/benches.
All ordered to bios versions:

1002
Warning: Spoiler! (Click to show)





0079
Warning: Spoiler! (Click to show)





0081

Warning: Spoiler! (Click to show)






0082
Warning: Spoiler! (Click to show)





0083
Warning: Spoiler! (Click to show)





Thank you for the extensive report smile.gif

Quote:
Originally Posted by badhairguy View Post

This is after about 2 minutes of prime 95 blend. *edit* disregard the clock settings shown in bios, it never shows the right clocks in that panel. As a side note (bug?), any time i load an overclock profile, I have to manuall retype the bclk value and hit enter or it doesn't actually change the clocks.


The BCLK should apply, it's just the BIOS calculation that isn't refreshed until you retype the value.

Quote:
Originally Posted by tomomosius View Post

is there a way to set it up so like after 5 minutes after booting up in to the system the qcode light turns off?

See the software above smile.gif

Quote:
Originally Posted by S1L3N7D3A7H View Post

@elmor

Bug report for BIOS 0079.

Whenever I have my overclock loaded, if I enable the CB15 performance BIAS, I get code 8 crashes when it starts to load windows. I have reproduced this 3 times to confirm.

The performance bias options rely on setting non-default AMD settings which are disabled by AMD due to instability issues affecting some CPUs. Unfortunately it seems you have such a chip. I believe you can confirm by disabling SMT when using this option to make it work.

Quote:
Originally Posted by Ubardog View Post

Well I thought it was Fixed

29 loops into IBT and a 170v on 1.05V Sb and Instant fail on IBT. All recorded in HW64... The rig didn't crash thou. biggrin.gif

@elmor please if you have time can you have a quick look at my situation and Advise me please


Even if it is just a read out error . It seems to be messing with stability



I have CSV file too but its over 5m

There's no way unreliable readouts are affecting your stability.

Quote:
Originally Posted by gupsterg View Post

@elmor

Thank you for app smile.gif , hope later UEFIs have fix in so no need for app wink.gif .
I been using all same values.

You may recall several times I've posted that my RAM bench in AIDA64 has improved on 0079, even though all same setup. Now SOC is defo needed to be 0.962V in my case, I'll explain wink.gif .

I had been using 0.962V on 1002 recently. You see I only needed 0.950V to pass the lengthy stability testing I did for OC. What I found was every so often on low loads the system would crash with code 8 or reboot spontaneously. I reckon this was due to how when system under load all voltages seem to be a notch higher if you get what I mean. This issue was intermittent and happened after some length, upping SOC to 0.962V on 1002 fixed it for me.

0079 flips out quicker if SOC is lower than 0.962V for me, why I think this is as RAM is running more efficiently/tighter so lack of SOC is more evident.

Well it's not exactly a bug, just a preference. Did it solve the fan ramping issue for you?

Quote:
Originally Posted by Timur Born View Post

When does the averaging algorithm set in then, or based on what condition exactly? Because below 60ish C SIO CPU is just Tctl +5C, only once you cross a threshold that averaging seems to apply.

And how it it weighted and does the weighting change?
Then 110C SIO was a coincidence, because that likely is the temp at which Tctl vs. Socket reaches shutdown values. Is shutdown temp 115C Tctl then?
I have seen (twice) and changed this setting (once), but most of the time it does not show up in my BIOS setup (0081).

Calibration algorithm:

Temp A = Tctl + 5
Temp B = CPU socket thermistor

The reported value will be whichever is lower of Temp A or B.


Thermal shutdown is completely managed by the CPU, so that would be something they have to fix. It sounds as if your CPU is crashing due to the high temperature before it's reaching the critical value (I don't have much details on it). Additionally in this stage the CPU clocks should be off, meaning it's not generating much power (if any). Perhaps @The Stilt has more detailed info?

Quote:
Originally Posted by Wally West View Post

@elmor

edit: I wasnt using the cablemod adapter that switch the green and blue.

I'm currently using the version available on the Asus website. V1.04.26. I also tried the Intel version you link earlier this week (V.1.04.29 or 31 if I remember correctly), but it doesnt change anything except I have more crash/freeze with it.
The iROG versio is (look at the picture below?)
CableMod magnetic rgb, so they are suppose to be aura ready? I have 6 strip and all of them is doing that.
Case: Inwin 303
Thanks you biggrin.gif If you want to see more pictures of the build: http://imgur.com/a/Kl0vh


I also tried with the stock RGB cooler. when I'm choosing a static color, it's working (blue is blue, green is green, etc..), but the rainbow effect isnt synchronize (ninja edit: yes it is sync if I'm calibrating it)
edit: One thing I noticed is when I was calibrating the stock cooler, when it was asking for the color. The order was 1-3-2 (red-blue-green). But now with the cablemod RGB led it's 1-2-3(red-green-blue). What's suppose to be the right order in the calibration setting? red-blue-green or red-green blue in the calibration?

I updated my BIOS and the new windows version (with the gaming mode) at the same time. Yesterday, I tried (in order) 1002 -> 0082 -> 1002 -> 0079 (now I have 0079). I also tried a reset of the PC (but I still have the last windows version)


Thanks for the info, will look into it.

Quote:
Originally Posted by LittleVulpix View Post

@elmor I noticed that the tool to change the cpu temp offsets doesn't work when super IO clock skew is disabled, complaining that something is not available/enabled. I assume this is expected? (but it may be necessary to note this for the tool usage so people don't get confused when it simply quits on them).

Thanks for it anyway. Also yeah it seems like something that causes a bug on the gpu and then the whole board during screen initialization. I'll have to wait for AMD.

Small bug, fixed in the version included above.

Quote:
Originally Posted by f1LL View Post

@elmor:
Two minor (cosmetic?) changes/bugs I could notice immediately when moving from 1002 to 0079:
1. during POST it now says "EVALUATION COPY." on screen.
2. now the POST report always says the speed of the CPU would be 3000MHz, no matter what it actually is.

1. What? During POST or in BIOS? Can you show a picture or video?
2. Low priority fixing this right now.

Quote:
Originally Posted by Huntcraft View Post

I found a super strange bug. If i set ram to 3600 c13 13 13 13 28, it boots at 14 13 13 13 28. If try to set 12 13 13 13 28 it goes to 15 15 15 15 36

If DRAM Ratio is 2933 or above with "1T" BIOSes, you can only set even CAS Latency. It will round up to the closest even integer. It's in the XOC guide. The second case is because your DRAM settings failed and it reverted to defaults.

Quote:
Originally Posted by icebalm View Post

@elmor don't know if you guys know this, but in 1002 and in the new beta firmwares there is a bug that if you make any changes to the CSM options under boot and save them, the next time you reboot and go into setup all the options under Advanced will be screwed up. Making any changes to any options under Advanced will make the same change for all options, so if you disable SATA for instance, it will disable *everything*. frown.gif Only way to fix is to clear CMOS.

Can you give me exact instructions on what to set (starting from CMOS defaults) in order to reproduce?

Quote:
Originally Posted by Karagra View Post

@elmor the bios has a overclock bug I think I figured out how to trigger it but I will have to test more in the morning. (If you fail a overclock in a curtain way, it limits all your overclockss after.) Clearing cmos doesnt help only a bios reinstall.

Do you have any exact instructions and more information?
post #8140 of 17652
Quote:
Originally Posted by elmor View Post

Do you have any exact instructions and more information?
No Sorry I have been overclocking non stop, and for the life of me I cannot fully replicate the issue, It may have to do with I am within a more stable range now so I am not noticing the issue anymore, One thing I noticed which is posted a bit back on the 2933mhz I can overclock between 2933-3600mhz without a failed boot. On the 3200mhz frequency I cannot boot unless i am running 3466+. Also 3200mhz frequency @3600mhz requires 18-18-18-18-42 1.425v to run while 2933mhz @3600mhz requires 16-16-16-16-36 1.37v to run. This was on every bios (did not fully test 79 with this) I am guessing this is just a wait till May type issue that will be resolved.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD Motherboards
Overclock.net › Forums › AMD › AMD Motherboards › ROG Crosshair VI overclocking thread