©ASUSTEK COMPUTER INC. ALL RIGHTS RESERVED.

Shamino · ‎05-21-2020

Braegnok · ‎12-31-2020

So far everything is stable, running 16-16-16-36 CR2. Going to play around with tRFC, run some more tests.

I bumped VCCIO/SA up to 1.25V, running Prime95 now for a few hours,..

n1kobg · ‎01-01-2021

Braegnok wrote:
So far everything is stable, running 16-16-16-36 CR2. Going to play around with tRFC, run some more tests.

I bumped VCCIO/SA up to 1.25V, running Prime95 now for a few hours,..

What voltage is the RAM? Can you go lower on the D1s? there is a formula u can calculate this section.
For 4300mhz you can increase the throughput & latency further. You can set VCCs upto 1.35 but thats the absolute limit if u dont have active cooling (1.3 is fine for good board). On that voltage it will go close to 1.4v in reality. Lower tcke to 6, twrrd_sg/dg by 3-4, twrrd_dr/dd & twrwr_dr/dd to 7 or 8 depending which have better performance. tRFC can be 344, even less if u have stability, like 280. Check if increasing twr to 16 will give u more read/write speed.

morph_ · ‎12-31-2020

Tune secondaries down it will increase testing speed and performance same with tertiary groups

DD can also be 0 afaik as you are just running 2 sticks.*

You should probably ramp up your uncore speed to 4.7-4.8.*

I wonder if 4300/4400 is achievable at 16 flat with bit more dram voltage 1.49 etc...*

How are you testing stability btw?*

Personally I’d go EKWB gear for water.*

Braegnok · ‎01-01-2021

morph. wrote:
Tune secondaries down it will increase testing speed and performance same with tertiary groups

DD can also be 0 afaik as you are just running 2 sticks.*

You should probably ramp up your uncore speed to 4.7-4.8.*

I wonder if 4300/4400 is achievable at 16 flat with bit more dram voltage 1.49 etc...*

How are you testing stability btw?*

Personally Iâ€™d go EKWB gear for water.*

The only timings I have adjusted/set in above post are the Primary's, tREFI, and I'm working on tRFC, will be between 378 - 294 when I'm finished testing.

All other timings Secondary, and Tertiary are still in Auto at this point, working my way down my list, testing as I go.

For testing first is MEMTEST 86 bootable flash drive, than after booting into Windows I use Karhu RAM Test, Prime95 large FFTs.

I have already committed to using Optimus blocks on this system, I have Optimus Signature V2 CPU block, Absolute D5 Reservoir 306mm system in hand and have Pre-Ordered Optimus Absolute GPU block on September 14, 2020. Production has been scheduled on the 3090 FE block for third week of January 2021.

I have EKWB gear on my Z390 system, CoolStream XE radiators,.. vary nice hardware. EKWB has not yet released a block for the 3090 FE, coming soon perhaps.

morph_ · ‎01-01-2021

Braegnok wrote:
The only timings I have adjusted/set in above post are the Primary's, tREFI, and I'm working on tREC, will be between 378 - 294 when I'm finished testing.

All other timings Secondary, and Tertiary are still in Auto at this point, working my way down my list, testing as I go.

For testing first is MEMTEST 86 bootable flash drive, than after booting into Windows I use Karhu RAM Test, Prime95 large FFTs.

I have already committed to using Optimus blocks on this system, I have Optimus Signature V2 CPU block, Absolute D5 Reservoir 306mm system in hand and have Pre-Ordered Optimus Absolute GPU block on September 14, 2020. Production has been scheduled on the 3090 FE block for third week of January 2021.

I have EKWB gear on my Z390 system, CoolStream XE radiators,.. vary nice hardware. EKWB has not yet released a block for the 3090 FE, coming soon perhaps.

tRFC & Trefi Â*should be the last values to tweak.

Braegnok · ‎01-01-2021

morph. wrote:
tRFC & Trefi Ã‚*should be the last values to tweak.

I have been following the Guide you posted 12-16-2020 11:23 PM post #422 on this thread.

The Guide lists tRFC & tREFI as third step, with Primary Timings being second step ?

I am getting conflicting information,.. lol

Going to stick with the Guide, and keep plugging along.

Not all motherboards react the same, as with memory modules also different results are common.

Braegnok · ‎01-01-2021

@n1kobg, my memory voltage is 1.45V, is shown as DRAM on HWiNFO64 Sensor Status above CPU VCCIO.

Thanks for the advice.

This 4200MHz profile is going to be my 24/7 settings, going for the lowest voltages possible with 100% stability.

I will be working on other higher MHz profiles later,.. which voltages are of no concern, only performance #s and stable enough to post a screen shot.

martine-dee · ‎01-01-2021

Following a lamp that sort of pinches the aesthetics, I just found out that M12E has a BIOS switch. I think I will save that old BIOS purely for curiosity, even though switching now requires each BIOS to restore their OLED firmware. It won't be used much, that's for sure.

For the record, that untouched BIOS is 0403 from March 23, 2020.

Keep exploring, keep innovating, keep creating

morph_ · ‎01-03-2021

tRFC & tREFI generate a fair bit of heat, especially from a long run high-temperature stability stand point.

Apothysis · ‎01-04-2021

morph. wrote:
tRFC & tREFI generate a fair bit of heat, especially from a long run high-temperature stability stand point.

tRFC and tREFI are related to refreshes, tREFI being the average interval between refreshes and tRFC being the time until next command after a refresh. A refresh hits all banks, rows and columns with voltage (and thus heat), so increasing the time interval between refreshes leads to less power generated.

They are sensitive to heat (data integrity degrades faster at high heat) but they do not increase heat generation, if anything increased tREFI reduces heat.

Asus Z490 stuff