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

ROG Crosshair VI overclocking thread - Page 1031

post #10301 of 17580
Quote:
Originally Posted by Raja@ASUS View Post

According to what Elmor said, yes it is. Once AMD iron out their problems, the Vboot and DRAM voltage should track when on Auto (like it does on other platforms).

Will you be implementing this fix on the X370 Prime Pro, too? This board doesn't even have an option for Vboot voltage, so it's always at 1.2V with no way to change it. This severely limits which memory speeds you can use with this board. For example, I can't use 3200 @ C14 with my B-die RAM, due to cold boot problems. It's perfectly stable once you manage to get it to actually boot and have the 1.35V take effect but the 1.2V Vboot is making cold boots a pain.

If not, will you at least consider adding manual Vboot adjustments to the Prime UEFI? This board probably has the worst memory support of all X370 boards due to this bug (boards like the Taichi also have options for Vboot adjustment)
Edited by JimmiG - 4/20/17 at 5:38am
post #10302 of 17580
Quote:
Originally Posted by BoMbY View Post

Is there any chance you could build an auto-tuning for the ProcODT, and other settings, or maybe enable more ProcODT options (because the steps currently are so large, and different)?


These values should auto-sense or have a behind scenes table; the platform is just a bit raw right now. And no, you can't add values to this yourself. The values are limited by the divider network that AMD built into the CPU, and that's based on platform line impedance and DRAM specifications. In the ideal sense, for tweakers, adding some fine control into the architecture would be nice, but there are considerations that engineers take into account - often, the crazy tweakers are not their focus. Look at how difficult it is just getting people to use what we have...
Edited by Raja@ASUS - 4/20/17 at 5:43am
post #10303 of 17580
Quote:
Originally Posted by JimmiG View Post

Will you be implementing this fix on the X370 Prime Pro, too? This board doesn't even have an option for Vboot voltage, so it's always at 1.2V with no way to change it. This severely limits which memory speeds you can use with this board. For example, I can't use 3200 @ C14 with my B-die RAM, due to cold boot problems. It's perfectly stable once you manage to get it to actually boot and have the 1.35V take effect but the 1.2V Vboot is making cold boots a pain.

If not, will you at least consider adding manual Vboot adjustments to the Prime UEFI? This board probably has the worst memory support of all X370 boards due to this bug (boards like the Taichi also have options for Vboot adjustment)

I will know more about what the UEFI for that board does over the next few weeks, when I'm at HQ.
post #10304 of 17580
At least for me Vboot does not seem to help when booting completely cold, had it up to 1.45V. But if it fails it is only a little inconvenient to take a quick detour through the BIOS.

Following Rajas advice I did a small memtest session, seems fine:



53.3 also seems to be the optimum for my 8GB b-die 1DPC config.

I then tried to fine tune SoC voltage a bit and found a rather distinctive instability when it was too low (for me that was at 0.95V for the 3200 memory): no code 8 but just a soft reset like behaviour. Might be useful to determine the cause of an instability.
main
(10 items)
 
  
Reply
main
(10 items)
 
  
Reply
post #10305 of 17580
Quote:
Originally Posted by littlestereo View Post

Absolutely, it's within AMD's XFR voltage range

Please be careful with that kind of advice.

XFR is a very temporary boost to 1 core only so cannot be compared to running all cores at 1.45v.

The power involved will be drastically different at a given voltage.
post #10306 of 17580
Quote:
Originally Posted by malitze View Post

At least for me Vboot does not seem to help when booting completely cold, had it up to 1.45V. But if it fails it is only a little inconvenient to take a quick detour through the BIOS.

Following Rajas advice I did a small memtest session, seems fine:



53.3 also seems to be the optimum for my 8GB b-die 1DPC config.

I then tried to fine tune SoC voltage a bit and found a rather distinctive instability when it was too low (for me that was at 0.95V for the 3200 memory): no code 8 but just a soft reset like behaviour. Might be useful to determine the cause of an instability.

Cool. Thanks for sharing.
post #10307 of 17580
Quote:
Originally Posted by Raja@ASUS View Post

I will know more about what the UEFI for that board does over the next few weeks, when I'm at HQ.

Thanks. I thought the lack of Vboot and 1.2V DRAM training was a trick to sell more of the expensive boards, but if it's actually an AMD bug I guess maybe there was no malicious intent smile.gif
post #10308 of 17580
@elmor

Properly the best quick example. To catch it in Bios is hard and TBH i got better thing to do but i feel i must catch it in action so back to Bios
https://1drv.ms/v/s!Ap54bdL_bUvmg2L5nIXUSsff7Zqz

EDIT
Phone is low on juice got a SS thou


Gonna flash a different bios remove battery ect and repeat with fully charged phone

Or is this read out errors i should ignore as well ?
Edited by Ubardog - 4/20/17 at 6:28am
post #10309 of 17580
Quote:
Originally Posted by JimmiG View Post

Thanks. I thought the lack of Vboot and 1.2V DRAM training was a trick to sell more of the expensive boards, but if it's actually an AMD bug I guess maybe there was no malicious intent smile.gif

Not really. The teams are, at times, quite separate in their approach to certain things. For all I know, it tracks on those boards. I won't know until I ask.
post #10310 of 17580
Quote:
Originally Posted by Reikoji View Post

Warning: Spoiler! (Click to show)

Drat. 1 error in the 8hr 20m timeframe. Game over for 18-16-16-38!

Well i knew 18-16-16-38 was the tip of the ice berg. Will try 18-17-17-39 smile.gif 18-18-18-39 would be close to guaranteed I believe, its what I started with before I started trying reducing it.

The error is probably because I didn't close all useless backround processes too redface.gif.

Hello

If testing 64GB of memory GSAT is a much quicker option. 2 to 3 hours should uncover most any issue.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD Motherboards
Overclock.net › Forums › AMD › AMD Motherboards › ROG Crosshair VI overclocking thread