16GB RAM modules for Crosshair VII + 2700X

[gorpo]

Which modules work and at what speeds/timings? The QVL doesn't have much in terms of 16GB modules. I'm considering the F4-3200C14D-32GTZ.

[lightknightrr]

Wh00p, just checked the voltage: I set it for 1.345, not that other number (which would fry things).

[gorpo]

Ok. Completed 26+ hours of mprime blend test with no errors. CPB, PBO enabled, PE set to level 3 OC, RAM at 3200 DOCP.

Is there any way to know what level 3 changes/overrides? I think the same config as above but with PE on level 2 had failed mprime.

Happy with the performance. 4.35GHz on up to 4 cores, 4.11GHz all cores. Temps are a bit high. Worst part is temps spike easily, meaning that browsing on media rich pages can spike above 55-65C momentarily, spinning up the fans hard.

y-cruncher failed on its 3rd iteration of tests, specifically on N64, when it somehow spiked to 96C. It was so brief it didn't even register on the temp plot on psensor, but the max temp counter shows 96C..

Next I might either noting the DOCP timings and trying the Ryzen Calc timings, or start reducing voltages/temps.

[gupsterg]

@gorpo

P95/mprime would have loaded all cores (unless yu setup testing for single thread), so you may not have tested when single core boosts occur what happens.

Generally I have see Y-Cruncher boost higher than P95 even on all cores. Y-Cruncher when it changes test set, CPU may boost single cores substantially higher than P95, it could be the test failed due to that aspect.

On OCN, in C7H OC thread it has been highlighted how to test single core with P95 by The Stilt and myself recently. This is a must IMO when keeping PB/XFR active and not doing an ACB OC. I will be adding this info today to ROG version of thread in OP.

[gorpo]

gupsterg wrote:
@gorpo

P95/mprime would have loaded all cores (unless yu setup testing for single thread), so you may not have tested when single core boosts occur what happens.

Generally I have see Y-Cruncher boost higher than P95 even on all cores. Y-Cruncher when it changes test set, CPU may boost single cores substantially higher than P95, it could be the test failed due to that aspect.

On OCN, in C7H OC thread it has been highlighted how to test single core with P95 by The Stilt and myself recently. This is a must IMO when keeping PB/XFR active and not doing an ACB OC. I will be adding this info today to ROG version of thread in OP.

That sounds exactly right. When changing tests, for some tests the temp would spike momentarily then settle down lower.

I did do 1, 2, 4 core testing with mprime, but not extensively. Just a few minutes to see the frequency and where the temperature spiked and settled. y-cruncher really does seem to hit it harder.

Looking forward to your updated C7H post. I guess I should be reading OCN too.

[balalahmad]

I have a CH7 + 2700x
The ram I'm using is the G.SKILL TridentZ RGB Series 16GB (2 x 8GB) 3200: Memory Model F4-3200C14D-16GTZR
I've only tried setting it to DOCP in Bios - running so far fine at 14-14-14-34.
Ran a the HCI Memtest, no errors overnight.

Any suggestions on where to go from here? Really new to this, my first build since my last AMD build from 2009.

Thanks.

[gorpo]

balalahmad wrote:
I have a CH7 + 2700x
The ram I'm using is the G.SKILL TridentZ RGB Series 16GB (2 x 8GB) 3200: Memory Model F4-3200C14D-16GTZR
I've only tried setting it to DOCP in Bios - running so far fine at 14-14-14-34.
Ran a the HCI Memtest, no errors overnight.

Any suggestions on where to go from here? Really new to this, my first build since my last AMD build from 2009.

Thanks.

That kit with 8GB modules is really well supported. This thread is about 16GB modules. The motherboard has a number of preset RAM configs you could try.

[srsbsns]

error