Asrock Z170 OC Formula-DDR4

[rtsurfer]

Getting tired of pimping a board that isnt available sorry but managed some good 32m and wanted to share my subtimings.

 

 

Considering most of the terts are next to worthless, I'd say you've been sharing your timings all along.

 

Awesome run, little bit more.

[xxbassplayerxx]

Splave, what kind of volts on the sticks?

[phobosq]

and what cooling, all air?

[Splave]

air 1.85v in bios 1.88ish real

[GtiJason]

Looks like @GtiJason has been pushing his fastest 4K+ runs with 2T on that board. He currently has the fastest 5G 32M time with that board. Ranking here.

 

 

Yep, usually around ½ second slower than 1T, but I don't look at it that way. I see still 4 seconds faster than E Die

[xxbassplayerxx]

Have you had any luck getting 1T going?

[sabishiihito]

PCH voltage isn't doing anything here for 1T.

 

Sent from my SM-G928T using Tapatalk

[Guest Digg_de]

 

Not stable with CW. It's a 45/45 run.

[Guest Digg_de]

XTU Run with 1978 12-12-28 1T

 

[sabishiihito]

Things magically started working (seriously, I have no idea what changed).

 

Patriot Viper 4 DDR4-3400@3733 12-12-28-1T (I think this is max for this kit)

 

Corsair Vengeance LPX DDR4-3600@3866 12-12-28-1T

no waza:

 

waza:

[GtiJason]

Things magically started working (seriously, I have no idea what changed).

 

Patriot Viper 4 DDR4-3400@3733 12-12-28-1T (I think this is max for this kit)

 

Corsair Vengeance LPX DDR4-3600@3866 12-12-28-1T

no waza:

 

waza:

 

That's crazy man, B die is a strange creature. Doug and myself got our highest 1T frequency's the first time trying them. I think my 3600c16 kit booted Win 10 at 3980c12 1T it's been as bad as unable to boot 3733 for me. I haven't fired up the 170 in a week or so. Been busy playing with some fantastic imc es 3258's that allow me to run lower vdimm, sa volts and io's than any of the 10ish retail HW/DC cpu's i have owned and will run 5g on h20. Still wish z97 ln2 board was even half as easy as AsRock to work with. Seriously a poo ton of work even with saved profiles epecially cold

[Sparks.nl]

I was playing around yesterday and booted with mostly auto settings for timings:

 

http://valid.x86.fr/5duzr7

4200MHz seems to boot as well.

 

Edit: I ordered an LN2 kit for memory. Let's see what it can do...

Edited February 5, 2016 by Sparks.nl

[Sparks.nl]

Question: Since my bios 2.10 update I can't seem to boot 4000MHz+ memory speed (B die) at 1T.

Going back to 1.93 does not change it. Does anybody else have this issue?

I had a working bios profile that does not work any more.

I left the motherboard without any wires connected and battery out for 27 hours, it didn't change anything.

[Splave]

must be missing a setting, check that ME and microcode is the same as before

[Sparks.nl]

I think that I had a microcode update in the new bios. Where can I see this?

 

Edit: Sisoft Sandra reports the microcode as version: MU065E0374

Edit 2: I checked the ME version with the Intel tool. It doesnt't seem to support Z170. I don't have any idea how to read it.

 

I am not sure which version I had before.

Edited February 8, 2016 by Sparks.nl

[Splave]

should be on cpu page 2D

[Guest Digg_de]

On the way to a 5m57....s time.

My best result from last night.

 

3866 12-12-28 1T stable with Win7 64Bit, but no chance with XP32.

 

[Sparks.nl]

should be on cpu page 2D

 

I looked at the devices in Windows.

For Management Engine Interface it tells me that I have version 11.0.0.1172 for the version of the driver assembly.

 

In the CPU part of the Bios it tells me the microcode version is 506E3 74.

At the Microcode Update parameter I can choose 15, 2D, 49 and 74.

Is this the part where I can choose which version I want to use?

[Guest Digg_de]

 

Edited February 10, 2016 by Digg_de

[Guest Digg_de]

In the CPU part of the Bios it tells me the microcode version is 506E3 74.

At the Microcode Update parameter I can choose 15, 2D, 49 and 74.

Is this the part where I can choose which version I want to use?

 

Yes, but don't ask me about the differences between the MCs.

[Sparks.nl]

I did a short test and it seems that some versions of the microcode can not boot with (really) tight timings.

It needs more volts or I need to set trfc "loose".

XTU score's were different as well, altough I did only one run for each test Xtu, pi and cb 11.5).

 

Edit: What is for sure is that it definitly has influence on the memory stability and settings.

If other motherboards can't set it then this is an advatage for Asrock.

 

Question: does this mean that I can undo the lock of the non k cpu's by choosing a different microcode?

Edited February 10, 2016 by Sparks.nl

[Guest Bullant]

 

 

Nice,nice little boost from pulling rtls in correctly and couple other things

 

Keep pushing it

[GtiJason]

On the way to a 5m57....s time.

My best result from last night.

 

3866 12-12-28 1T stable with Win7 64Bit, but no chance with XP32.

 

Stable Win 7 64 bit but not XP 32 bit, So strange beginning to think B die is an evil prank and there is someone out there who controls it's behavior simply to bunny with us. Well played Samsung but just know I'm patiently waiting for you to break, make that first mistake and it's over.

[Guest Digg_de]

Nice,nice little boost from pulling rtls in correctly and couple other things

 

Keep pushing it

 

I never fix the RTL/IOs in the past. Playing with subtimings till the good RTL/IOs coming from itself. But the OCF don't set them everytime to the same stats.. it varies. Sometimes it help to set a lower Mem-Divider and then back to the higher one, or changing tRFC a little bit like 208, 218, 228 instead of 220 like the profile do.

[Splave]

RTL is dynamic and partially calculated each time you retrain