Ryzen 7000 / Zen 4 Owner's Club

[J7SC_Orion]

3 hours ago, ENTERPRISE said:

Possible its only an Intel thing then. 

 

The Aorus bios has so many nooks and crannies to begin with, but some of their parameter's names and shortforms are different. Still the values around tRTP in the screenshot above are close, IMO.

 

In other news,  after just trying DDR5 8200 for the fun of it, I used DDR5 8100 more extensively tonight; no crashes or WHEA errors so far. The only change was to bump the DDR voltage to 1.445 V

 

 

[Hurricane28]

9 minutes ago, J7SC_Orion said:

 

The Aorus bios has so many nooks and crannies to begin with, but some of their parameter's names and shortforms are different. Still the values around tRTP in the screenshot above are close, IMO.

 

In other news,  after just trying DDR5 8200 for the fun of it, I used DDR5 8100 more extensively tonight; no crashes or WHEA errors so far. The only change was to bump the DDR voltage to 1.445 V

 

 

 

Looking good chief. 

I am looking for new RAM atm, so hard to choose what set to get man.. 

 

I have 200 to spend. Any ideas? Also the kits you can get in the US or outside of europe, are too expensive or not for sale here sadly so im kinda bound by one reseller. 

[ENTERPRISE]

On 13/10/2023 at 13:46, kaliz said:

@kaliz is this from an Intel BIOS ?

[neurotix]

Mine has like half the options in that screenshot in the same section.

[kaliz]

i have to apologize i think Round trip latency is a Intel thing and AMD does not name it like that! i have found it on my B760M bios but not in B650E nor B650 boards , only training options in Nitro settings menu. You can try 4 taps for a similar effect and tighten down those nitro settings to 1 or 0

[ENTERPRISE]

5 hours ago, kaliz said:

i have to apologize i think Round trip latency is a Intel thing and AMD does not name it like that! i have found it on my B760M bios but not in B650E nor B650 boards , only training options in Nitro settings menu. You can try 4 taps for a similar effect and tighten down those nitro settings to 1 or 0

No problems bud, thanks for confirming

[J7SC_Orion]

44 minutes ago, ENTERPRISE said:

No problems bud, thanks for confirming

 

I dug this old post out (from Linustechtips.com ) which may or may not hold true...

 

 

At the end of the day, RTL reminds me a bit of "tRFC (ns)" in Zen, which is a derived parameter, not set. In any case, at least this whole exercise got me to look much closer at all the nooks and crannies of my 670E Aorus Master bios; I took over 20 pics but instead, > here is the link to the pdf manual w/ bios pages  

 

[Hurricane28]

woot woot, 

 

Got my present early today! 

 

Idk if this is placebo but my PC does feel little snappier. 

[Hurricane28]

I tried to boot at 2100 Fclk but that went south real quick lol. 

 

Back to xmp for now. 

 

What should and could i change and why on this set? 

[ENTERPRISE]

13 minutes ago, Hurricane38 said:

I tried to boot at 2100 Fclk but that went south real quick lol. 

 

Back to xmp for now. 

 

What should and could i change and why on this set? 

What Die is your new kit ?

[Hurricane28]

14 minutes ago, ENTERPRISE said:

What Die is your new kit ?

Its M die. 

 

I cannot run typhoon burner anymore for some reason but on the stick says M so its M die. 

[J7SC_Orion]

4 hours ago, Hurricane38 said:

I tried to boot at 2100 Fclk but that went south real quick lol. 

 

Back to xmp for now. 

 

What should and could i change and why on this set? 

 

4 hours ago, Hurricane38 said:

Its M die. 

 

I cannot run typhoon burner anymore for some reason but on the stick says M so its M die. 

 

From what I can gather, Hynix switched over completely to the 'new' M-die along with A-die, with an exclusive A-die period in-between separating that from the (original) M-die models.  SKUs didn't really seem to change in the this transition period, so it is a guessing game unless someone has access to serial-number-production-date-input data keys. Since I only have one DDR5 kit - 2x 24 GB of new M-die 7200 which can reach > 8200 - I can't really draw too many comparative conclusions. 

[ENTERPRISE]

I just wanted to say thanks for all the help with the RAM OC, this actually helps with my Dominator Titanium review. @J7SC_Orion I will pick your brain regarding Curve Optimizer in a few days lol.

[neurotix]

Happy to help.

[Hurricane28]

1 hour ago, J7SC_Orion said:

 

 

From what I can gather, Hynix switched over completely to the 'new' M-die along with A-die, with an exclusive A-die period in-between separating that from the (original) M-die models.  SKUs didn't really seem to change in the this transition period, so it is a guessing game unless someone has access to serial-number-production-date-input data keys. Since I only have one DDR5 kit - 2x 24 GB of new M-die 7200 which can reach > 8200 - I can't really draw too many comparative conclusions. 

 

So it appears indeed. 

 

Later M die seem to be just as good as A die or is merged like you said and rebranded as M die.. Its so weird how they do things lately tho... 

 

Anyway, after a hassle with BIOS and settings etc. etc. i managed to get my kit working at 6400Mhz xmp profile. The first 

 

F10 Bios didn't like the RAM that much because it was an older AGESA, hence i talked about this before, and now i have the proof that newer AGESA really works and can improve significantly. 

 

On F10 i could boot, run some browser related stuff but crash in Warframe and other thins, it simply wasnt stable at all. 

 

After resetting the BIOS and restart all the way to Windows 11 than restart and reload the xmp, still not stable enough. Completely shutting the PC off and do cold boot, still not stable. 

 

Than i had enough and flashed the newer F11A BIOS with the AGESA 1007B with checksum A666, that's the devil for ya which sometimes is quite accurate with DDR5, and it booted straight up no issues and stable for 1 hour of memtest64 in Windows 11. I did a reboot and tested again, still one hour stable. Cold boot, still stable. 

 

Temps are a tad high tho so i have to think about how to cool it and still look cool without a fan in front of the ram lol. 

 

Here are the tests so far: 

 

 

[kaliz]

those 6400C3239G16G kits are M-die nowadays, for a brief period they put the best Hynix A-die in them from Nov 22 until Mar 23... yours might be newer.   And barcode says it all if last letter is M. 

 

Try this:

 

6400 : 2133 : 6400

 

 

[J7SC_Orion]

11 hours ago, ENTERPRISE said:

I just wanted to say thanks for all the help with the RAM OC, this actually helps with my Dominator Titanium review. @J7SC_Orion I will pick your brain regarding Curve Optimizer in a few days lol.

 

...looking forward to help you torture yourself with CO 

[kaliz]

to find the best Curve for your CPU this program lets you use Curve Optimizer (PBO) in Windows : SMUdebugTool  

Release v1.3.2: add PBO tab for testing purposes, limited to 16 cores · irusanov/SMUDebugTool

GITHUB.COM

---

add PBO tab for testing purposes, limited to 16 cores and family 19h update core dll to v 1.6.3

 

 

If you run a hammering benchmark, like Ycruncher, you can find your limits pretty easily. Also other bencmarks like 7zip are very usefull to find the max curve and optimal RAM frequency/fclk frequency .

 

[Hurricane28]

15 hours ago, kaliz said:

those 6400C3239G16G kits are M-die nowadays, for a brief period they put the best Hynix A-die in them from Nov 22 until Mar 23... yours might be newer.   And barcode says it all if last letter is M. 

 

Try this:

 

6400 : 2133 : 6400

 

 

 

 

Tried your timings and voltage and here is the result: 

 

 

Edited October 19, 2023 by Hurricane38
Aida64 scores are all over the place with reads and writes tho, that might be indication of instability or is it an Aida64 thing?

[neurotix]

1 hour ago, Hurricane38 said:

 

 

Tried your timings and voltage and here is the result: 

 

 

Yeah your AIDA score is a lie. No way you have over 100GB/sec copy.

[Hurricane28]

15 minutes ago, neurotix said:

Yeah your AIDA score is a lie. No way you have over 100GB/sec copy.

That was a one time happening tho, its at 65K now for read and copy lol. Write is stable 95k every test. I ran it 10x to be sure, scores are consistent after this one. 

 

Previous bios i couldn't even boot at 6400 let alone be stable at 2133 FcLK. Amazing what they can do with newer bios. 

[kaliz]

really nice, now you can tighten up the primary timings to 30-38-38-50-88     4-8-16-4-16-(120ns)-384   you can get around up to 70000 read with this CPU and RAM combo!! and about 60ns latency!

[kaliz]

just ran 6600 C26-36-36-26-58-390 for the competition

 

 

/img/logo.png

kaliz`s y-cruncher - Pi-1b score: 42sec 958ms with a Ryzen 7 7800X3D

HWBOT.ORG

---

The Ryzen 7 7800X3D @ 5545.8MHzscores getScoreFormatted in the y-cruncher - Pi-1b benchmark. kalizranks #null worldwide and #null in the hardware class. Find out more at HWBOT.

 

[neurotix]

Agreed with Kaliz suggested timings for you. Ideally you'd want to be at cas 30.

 

Also, nice score in y-cruncher kaliz. I didn't realize when I benched yesterday that you can submit them for the Country Cup. I may have to do that.

[kaliz]

its for 4 cores 4 threads only so you have to disabled threads!