Ryzen 7000 / Zen 4 Owner's Club

[ENTERPRISE]

19 hours ago, J7SC_Orion said:

 

...here you > go

 

...and the latest Zen timings I use for DDR5 8000

 

What does your latency look like at those speeds/timings? Just wondering as im working on a kit now, granted at 6400 MT/s but looking at tight timings to better the latency. Then after that may push for higher clocks, at or close to 7000 MT/s and try to tighten timings again.

 

Looking at your UCLK you have it at DIV/2 ?

 

 

[neurotix]

2 hours ago, ENTERPRISE said:

What does your latency look like at those speeds/timings? Just wondering as im working on a kit now, granted at 6400 MT/s but looking at tight timings to better the latency. Then after that may push for higher clocks, at or close to 7000 MT/s and try to tighten timings again.

 

Looking at your UCLK you have it at DIV/2 ?

 

 

I'm pretty sure if you go back a few pages, J7SC_Orion posted screens of DDR5-8000 with AIDA64 and his latency was like 62ns.

 

Also, the Uclk/Memclk div 2 is because once you pass a certain speed on RAM, the Memclk will no longer be half the RAM clock.

 

I've been told that to outperform say, 6200MHz with tight timings, you need to be running 7600MHz to match it, and then if you get your kit to 8000MHz it would outperform ~6000 with tight timings.

 

This is my understanding of it from chatting with the guys here. I have G.skill Flare X B-Die and was warned that going over 6600 on it is really hard or impossible. Ideally for very high RAM speeds on Zen 4, you want Hynix A-Die. Check your kit with Thaiphoon Burner if you haven't.

 

Hope this helps.

Edited September 22, 2023 by neurotix

[J7SC_Orion]

6 hours ago, ENTERPRISE said:

What does your latency look like at those speeds/timings? Just wondering as im working on a kit now, granted at 6400 MT/s but looking at tight timings to better the latency. Then after that may push for higher clocks, at or close to 7000 MT/s and try to tighten timings again.

 

Looking at your UCLK you have it at DIV/2 ?

 

 

 

...yeah, I run mem DIV/2 at DDR5 8000. According to Buildzoid and a few others, DDR5 7600 speed is were DIV/2 becomes more or less worth it over DIV/1, at least in many more modern apps (think also resizable_BAR). As to latencies, I have now arrived at / dropped to the high 50 ns range in Aida bench, after much additional tuning. Things such as disabling PowerDown Mode and GearDown Mode give you an extra push in at least Aida cache and memory benchmark. That said, while I have disabled GearDown because I dropped a few primaries to 45, I re-enabled PowerDown again - sometimes, disabling it can lead to intermittent and hard-to-track oddities. 

 

Given the most recent AGESA updates which opened up a whole new world for Ryzen RAM clocking, a bone-stock 7200 kit at DIV/2 starts in the low to mid-70s ns latency - but the stock XMP and/or EXPO settings are soooo generous for the vendor as to not throwing any surprises (and RMAs). A well-tuned but 24/7 '''sane''' multi-purpose work-play setup should get to 59 ns - 64 ns range, depending on various factors such as your mobo's topology, your specific RAM kit and such...and I have seen low 50ies (51.x) ns on the 7950X/3D some folks showed w/ their Aida, but only with very high voltages.

 

I have three rules for my 7950X3D build re. voltages (and related temps):

1.) V_SOC has to stay in the 1.25X V range (up to 1.30 V is considered mostly 'safe' even after the bubble-burns on the socket / CPU underside earlier in the year), and
2.) VDD DDR has to stay in the 1.42X V range. My RAM's stock voltage is 1.35 V, with many newer DDR5 high speed kits even at 1.4 V or 1.45 V
3.) Cool DDR5 is happy DDR5 - I have some extra Arctic P12 fans right above the DDR5 RAM - this kind of RAM gets hot (including PMIC) and you want keep it in the low 40s at most when under stress. Many folks resort to water-cooling as DDR5 has some fundamental differences (and heat sensitivity) compared to DDR4 Samsung-B, for example

 

Final advice: Open your favourite beverage before diving into the RAM-tuning-and-stress testing rabbit holes (there are many of them)...it's going to take a while and a lot of testing...

 

 

 

 

Edited September 23, 2023 by J7SC_Orion

[ENTERPRISE]

9 hours ago, neurotix said:

I'm pretty sure if you go back a few pages, J7SC_Orion posted screens of DDR5-8000 with AIDA64 and his latency was like 62ns.

 

Also, the Uclk/Memclk div 2 is because once you pass a certain speed on RAM, the Memclk will no longer be half the RAM clock.

 

I've been told that to outperform say, 6200MHz with tight timings, you need to be running 7600MHz to match it, and then if you get your kit to 8000MHz it would outperform ~6000 with tight timings.

 

This is my understanding of it from chatting with the guys here. I have G.skill Flare X B-Die and was warned that going over 6600 on it is really hard or impossible. Ideally for very high RAM speeds on Zen 4, you want Hynix A-Die. Check your kit with Thaiphoon Burner if you haven't.

 

Hope this helps.

 

5 hours ago, J7SC_Orion said:

 

...yeah, I run mem DIV/2 at DDR5 8000. According to Buildzoid and a few others, DDR5 7600 speed is were DIV/2 becomes more or less worth it over DIV/1, at least in many more modern apps (think also resizable_BAR). As to latencies, I have now arrived at / dropped to the high 50 ns range in Aida bench, after much additional tuning. Things such as disabling PowerDown Mode and GearDown Mode give you an extra push in at least Aida cache and memory benchmark. That said, while I have disabled GearDown because I dropped a few primaries to 45, I re-enabled PowerDown again - sometimes, disabling it can lead to intermittent and hard-to-track oddities. 

 

Given the most recent AGESA updates which opened up a whole new world for Ryzen RAM clocking, a bone-stock 7200 kit at DIV/2 starts in the low to mid-70s ns latency - but the stock XMP and/or EXPO settings are soooo generous for the vendor as to not throwing any surprises (and RMAs). A well-tuned but 24/7 '''sane''' multi-purpose work-play setup should get to 59 ns - 64 ns range, depending on various factors such as your mobo's topology, your specific RAM kit and such...and I have seen low 50ies (51.x) ns on the 7950X/3D some folks showed w/ their Aida, but only with very high voltages.

 

I have three rules for my 7950X3D build re. voltages (and related temps):

1.) V_SOC has to stay in the 1.25X V range (up to 1.30 V is considered mostly 'safe' even after the bubble-burns on the socket / CPU underside earlier in the year), and
2.) VDD DDR has to stay in the 1.42X V range. My RAM's stock voltage is 1.35 V, with many newer DDR5 high speed kits even at 1.4 V or 1.45 V
3.) Cool DDR5 is happy DDR5 - I have some extra Arctic P12 fans right above the DDR5 RAM - this kind of RAM gets hot (including PMIC) and you want keep it in the low 40s at most when under stress. Many folks resort to water-cooling as DDR5 has some fundamental differences (and heat sensitivity) compared to DDR4 Samsung-B, for example

 

Final advice: Open your favourite beverage before diving into the RAM-tuning-and-stress testing rabbit holes (there are many of them)...it's going to take a while and a lot of testing...

 

 

 

 

 

Thanks guys for the very helpful insights. 

 

My RAM kit is Hynix A-Die. My timing and AIDA benches are below. 

 

 

 

What puzzles me is I have seen people get to around 50ns latency with similar RAM kits and timings, but I am struggling to achieve better than 64 ish.  Ignore that I am running 1.5v on the RAM, this was for a quick sanity check as I saw some people running CL28 required more voltage to get there. When benching the RAM gets to about 50c, so I may rig up...AKA cable tie another fan over the RAM. 

 

Any advice on improving my latency would be greatly appreciated, and don't worry. I know just how long it takes to perfect RAM settings, any time I have had to swap out a kit I have tuned...it hurts me lol.

[kaliz]

Try to set and check latency in Aida again!  GDM disabled, Powerdown disabled, TSME disabled, round trip latency enabled,

Nitro settings: enabled, enabled  1 2 0 8x 8x

[ENTERPRISE]

1 hour ago, kaliz said:

Try to set and check latency in Aida again!  GDM disabled, Powerdown disabled, TSME disabled, round trip latency enabled,

Nitro settings: enabled, enabled  1 2 0 8x 8x

 

Thanks for the input. 

 

Gear Down Mode - I cannot find this in my MSI BIOS, will see if I am able to find it

Powerdown Mode - I already had this disabled 

TSME - I have disabled this now 

Round Trip Latency - I cannot find this in my MSI BIOS, will see if I am able to find it.

 

The Nitro settings, if I apply what you recommend, I am unable to POST, perhaps a different combination, but I know little about the Nitro settings.

[J7SC_Orion]

12 hours ago, ENTERPRISE said:

 

 

Thanks guys for the very helpful insights. 

 

My RAM kit is Hynix A-Die. My timing and AIDA benches are below. 

 

 

 

What puzzles me is I have seen people get to around 50ns latency with similar RAM kits and timings, but I am struggling to achieve better than 64 ish.  Ignore that I am running 1.5v on the RAM, this was for a quick sanity check as I saw some people running CL28 required more voltage to get there. When benching the RAM gets to about 50c, so I may rig up...AKA cable tie another fan over the RAM. 

 

Any advice on improving my latency would be greatly appreciated, and don't worry. I know just how long it takes to perfect RAM settings, any time I have had to swap out a kit I have tuned...it hurts me lol.

  

...in addition to what @kaliz mentioned, a higher speed on the InFin fabric will also cut latencies though 2100 FCLK you showed in your post is nothing to sneeze at. I can run 2200 but typically have it at 2175 (2167 + Aorus 'bonus boost') 

[ENTERPRISE]

6 minutes ago, J7SC_Orion said:

  

...in addition to what @kaliz mentioned, a higher speed on the InFin fabric will also cut latencies though 2100 FCLK you showed in your post is nothing to sneeze at. I can run 2200 but typically have it at 2175 (2167 + Aorus 'bonus boost') 

Thanks for the feedback. Looks like MSI in their infinite wisdom must hide GDM and Round Trip Latency settings as I cannot find them anywhere. 

 

Also, have decided to start from semi scratch. Oddly my RAM throughput speeds dropped, even when restoring back to standard EXPO...they are lower than they were. So flashed BIOS again...only as restoring defaults through BIOS didn't restore expo speeds and just running memtest as a sanity check. 

 

Will fit a 120mm fan faced at the ram as well to aid in temp control. 

[J7SC_Orion]

22 minutes ago, ENTERPRISE said:

Thanks for the feedback. Looks like MSI in their infinite wisdom must hide GDM and Round Trip Latency settings as I cannot find them anywhere. 

 

Also, have decided to start from semi scratch. Oddly my RAM throughput speeds dropped, even when restoring back to standard EXPO...they are lower than they were. So flashed BIOS again...only as restoring defaults through BIOS didn't restore expo speeds and just running memtest as a sanity check. 

 

Will fit a 120mm fan faced at the ram as well to aid in temp control. 

 

...DDR5 setups are a bit trickier than DDR4 IMO - DDR5 has 'some' error correction code (re. single bit memory errors) though it is not the full ECC with an additional chip on the DIMM (per Corsair's web site). Because of this 'ECC Light', finding the edge of the canyon takes more testing. On my DDR4 systems I have GSkill GTZR Samsung-B for years and years, ezee-peezee to set up. DDR5 will take some more time to master.

 

Apart from lower primary timings and maxed InFin, tRFC in secondaries also help with higher bandwidth and lower latencies, but it looks like you are already doing well on tRFC. I do wonder about your tFAW '15' in your Zen Timings screenie above, btw. It might be that your systems is just ignoring it - 15 is super low for DDR5 

 

...btw, Unigine's Superposition 8K does give your system memory a bit of a workout as well once you have run the usual RAM stress tests which can be a bit boring. At least Superposition 8K is s.th. to look at and listen to while waiting for it to do its thing. 

[ENTERPRISE]

7 minutes ago, J7SC_Orion said:

 

...DDR5 setups are a bit trickier than DDR4 IMO - DDR5 has 'some' error correction code (re. single bit memory errors) though it is not the full ECC with an additional chip on the DIMM (per Corsair's web site). Because of this 'ECC Light', finding the edge of the canyon takes more testing. On my DDR4 systems I have GSkill GTZR Samsung-B for years and years, ezee-peezee to set up. DDR5 will take some more time to master.

 

Apart from lower primary timings and maxed InFin, tRFC in secondaries also help with higher bandwidth and lower latencies, but it looks like you are already doing well on tRFC. I do wonder about your tFAW '15' in your Zen Timings screenie above, btw. It might be that your systems is just ignoring it - 15 is super low for DDR5 

 

...btw, Unigine's Superposition 8K does give your system memory a bit of a workout as well once you have run the usual RAM stress tests which can be a bit boring. At least Superposition 8K is s.th. to look at and listen to while waiting for it to do its thing. 

Yeah DDR5 is definitely proving to be a little bit more of a challenge to get right that is for sure. It is a fair point regarding the ECC. Would it make sense to simply disable it ? I note in the BIOS it can be disabled but I would imagine its better to leave on. Fair note on TFAW, will keep that in mind to check to make sure that isn't causing issues.

 

Fair point regarding Superposition 8K would be a good workout and visually more pleasing than memtest lol. 

 

I might contact MSI support and ask why they hide GDM and RTL as I think that is poor on a higher end board.

[neurotix]

1 hour ago, ENTERPRISE said:

Yeah DDR5 is definitely proving to be a little bit more of a challenge to get right that is for sure. It is a fair point regarding the ECC. Would it make sense to simply disable it ? I note in the BIOS it can be disabled but I would imagine its better to leave on. Fair note on TFAW, will keep that in mind to check to make sure that isn't causing issues.

 

Fair point regarding Superposition 8K would be a good workout and visually more pleasing than memtest lol. 

 

I might contact MSI support and ask why they hide GDM and RTL as I think that is poor on a higher end board.

 

You can disable ECC in the bios and you may get slightly better performance or results in overclocking.

 

Don't worry too much about tFAW 15, I would suggest you set it to 16 though as it seems to like multiples of 4. I run my RAM at 6200MHz with 16 tFAW which is very close to what you're running. It's really tight but assuming you have no errors and the system is stable it should be ok. I suggest using GSAT for a minimum of 4 hours.

 

Is there anyway you can post your AIDA64 results? I missed them if you posted them. You have an issue with high latency right?

Edited September 23, 2023 by neurotix

[J7SC_Orion]

2 hours ago, ENTERPRISE said:

Yeah DDR5 is definitely proving to be a little bit more of a challenge to get right that is for sure. It is a fair point regarding the ECC. Would it make sense to simply disable it ? I note in the BIOS it can be disabled but I would imagine its better to leave on. Fair note on TFAW, will keep that in mind to check to make sure that isn't causing issues.

 

Fair point regarding Superposition 8K would be a good workout and visually more pleasing than memtest lol. 

 

I might contact MSI support and ask why they hide GDM and RTL as I think that is poor on a higher end board.

 

 ...re.  tFAW, here is a time-stamped segment by Buildzoid running DDR5 8000. He explains the relationship between tFAW and other parameters (ie. tRRD_S). At lower RAM clocks both can be reduced but their relationship has to stay intact

 

 

[ENTERPRISE]

19 hours ago, neurotix said:

 

You can disable ECC in the bios and you may get slightly better performance or results in overclocking.

 

Don't worry too much about tFAW 15, I would suggest you set it to 16 though as it seems to like multiples of 4. I run my RAM at 6200MHz with 16 tFAW which is very close to what you're running. It's really tight but assuming you have no errors and the system is stable it should be ok. I suggest using GSAT for a minimum of 4 hours.

 

Is there anyway you can post your AIDA64 results? I missed them if you posted them. You have an issue with high latency right?

Thanks for the insight. My latency was at around 65ns

18 hours ago, J7SC_Orion said:

 

 ...re.  tFAW, here is a time-stamped segment by Buildzoid running DDR5 8000. He explains the relationship between tFAW and other parameters (ie. tRRD_S). At lower RAM clocks both can be reduced but their relationship has to stay intact

 

 

Thanks for the video, very informative with respects to what each timing does and what other timings they are related to

 

Upon further testing, I have abandoned 6400MT/s OC and going to focus on 6600MT/s as I decided I would just push for the maximum as opposed to a semi OC with tight timings. It seems I cannot go any higher than 6600 at this point in time, I tried a wide range of different voltages and options after lots of research and I cannot post. So for now I will assume the wall is 6600MT/s and will now fly with that. 

 

Now testing 6600MT/s with Karhu at DIV/2 @ 2000Mhz FCLK. I will see if I can push 2067/2100 FCLK later on if all tests go well. Once I have found my maximum stable FCLK I will crack on at minimizing the timings. I believe that is a sensible order to do things:

 

1. Find max RAM Clock (All timings Auto in BIOS)

2. Find max FCLK 

3. Tighen timings 

4. Lower voltages, though not 100% necessary.

 

RAM sees 42c under load now that I have the fans installed, so I am please to see that.

 

Looks like Buildzoid hates AIDA, and he makes some fair points. Is there another good program for testing memory latency ? Just so I can use it to validate/see there differences against AIDA.

[J7SC_Orion]

4 hours ago, ENTERPRISE said:

Thanks for the insight. My latency was at around 65ns

Thanks for the video, very informative with respects to what each timing does and what other timings they are related to

 

Upon further testing, I have abandoned 6400MT/s OC and going to focus on 6600MT/s as I decided I would just push for the maximum as opposed to a semi OC with tight timings. It seems I cannot go any higher than 6600 at this point in time, I tried a wide range of different voltages and options after lots of research and I cannot post. So for now I will assume the wall is 6600MT/s and will now fly with that. 

 

Now testing 6600MT/s with Karhu at DIV/2 @ 2000Mhz FCLK. I will see if I can push 2067/2100 FCLK later on if all tests go well. Once I have found my maximum stable FCLK I will crack on at minimizing the timings. I believe that is a sensible order to do things:

 

1. Find max RAM Clock (All timings Auto in BIOS)

2. Find max FCLK 

3. Tighen timings 

4. Lower voltages, though not 100% necessary.

 

RAM sees 42c under load now that I have the fans installed, so I am please to see that.

 

Looks like Buildzoid hates AIDA, and he makes some fair points. Is there another good program for testing memory latency ? Just so I can use it to validate/see there differences against AIDA.

 

...I do use Aida but get Buildzoid's point(s). For general memory testing, I use OOCT / RAM and also highly recommend "memtest64"  from TechPowerUp. It doesn't show you the actual latency but is a good test to run for final stability (ie. 10 loops min). I believe Hydra 1.3Pro also has some RAM test w/ latency but I'm just using it now to get per-core CO values. So far (ie below), I've just been using all-core trial+error...below is a fully stable eclk 104 screenie with all-core - these things can clock ! 

 

 

 

 

 

[The Pook]

4 hours ago, ENTERPRISE said:

Now testing 6600MT/s with Karhu at DIV/2 @ 2000Mhz FCLK.

 

GSAT is faster than Karhu's RAM Test if you want to save some time. I used to use RAM Test exclusively but I switched to GSAT when I had a intermittent issues that I couldn't track down. RAM Test would run overnight without issues but GSAT would catch it within a couple hours. 

 

and it's free

[J7SC_Orion]

I finished with all the memory tuning and finally got around to do per-core CO. That said, the Aorus board also has a 90L5 'all-core' setting on top which I did leave in place when doing the per-core CO tests.  Below is for stock 100 bclk / DDR5 8000 and using the balanced Power Plan in Win 11 Pro . Translating that into my fav eclk 104 setting is going to be tricky since I am running a negative boost cap on that to compensate for the 4% increase in CPU speed - the 7950X3D (fortunately) does not allow me to set a positive v-core offset on CCD 0 at least. All this means that I probably have to run positive CO values for eclk 104, rather than the negative CO for the stock bclk. FYI, I among other tools used Hydra Pro to validate my CO for bclk though for now, I am staying below the 'max CO' by -4 or each core.

 

[ENTERPRISE]

18 hours ago, J7SC_Orion said:

 

...I do use Aida but get Buildzoid's point(s). For general memory testing, I use OOCT / RAM and also highly recommend "memtest64"  from TechPowerUp. It doesn't show you the actual latency but is a good test to run for final stability (ie. 10 loops min). I believe Hydra 1.3Pro also has some RAM test w/ latency but I'm just using it now to get per-core CO values. So far (ie below), I've just been using all-core trial+error...below is a fully stable eclk 104 screenie with all-core - these things can clock ! 

 

 

 

 

 

Yeah I will also run memtest after Karhu etc just to validate again Looking good on the CPU clocking, that is something I will get to once I have finished all the RAM bits and bobs!

17 hours ago, The Pook said:

 

GSAT is faster than Karhu's RAM Test if you want to save some time. I used to use RAM Test exclusively but I switched to GSAT when I had a intermittent issues that I couldn't track down. RAM Test would run overnight without issues but GSAT would catch it within a couple hours. 

 

and it's free

Thanks for that, I completely forgot about GSAT, will get that now and add it to the tests!

[ENTERPRISE]

13 hours ago, J7SC_Orion said:

I finished with all the memory tuning and finally got around to do per-core CO. That said, the Aorus board also has a 90L5 'all-core' setting on top which I did leave in place when doing the per-core CO tests.  Below is for stock 100 bclk / DDR5 8000 and using the balanced Power Plan in Win 11 Pro . Translating that into my fav eclk 104 setting is going to be tricky since I am running a negative boost cap on that to compensate for the 4% increase in CPU speed - the 7950X3D (fortunately) does not allow me to set a positive v-core offset on CCD 0 at least. All this means that I probably have to run positive CO values for eclk 104, rather than the negative CO for the stock bclk. FYI, I among other tools used Hydra Pro to validate my CO for bclk though for now, I am staying below the 'max CO' by -4 or each core.

 

Oh and another thing, looking to try and get my IF stable at 2100. I did some reading and come across Reddit

 

Quote

VDDG_CCD, VDDG_IOD, and SOC voltages are the main ones impact the IF.

VDDG_CCD should stay below 1v, VDDG_IOD below 1.05V and SOC below 1.15V (for Zen2 and 3) for long term use. Every CPU has different sweet spots for each voltage, sometimes it's better to lower a voltage.

 

Does this ring true with you ? What are you currently using voltage wise ?

[J7SC_Orion]

3 hours ago, ENTERPRISE said:

Oh and another thing, looking to try and get my IF stable at 2100. I did some reading and come across Reddit

 

 

Does this ring true with you ? What are you currently using voltage wise ?

 

...this seems to be geared towards Zen3...below on the left are Zen timings for my 5950X and on the right is the 7950X3D. They are different animals to some extent. For V_SOC on the 7950X3D, 1.3 V is supposed to be the safe max but I stay well below that, even at the speeds I am running. Some earlier bios would automatically push that to 1.4 V and above when enabling EXPO (or XMP) and that led to several such CPUs frying themselves.

 

The MEM VDD is the main DDR5 voltage and I have seen folks push 1.7+ V with extra RAM coolers. My M-die comes in the 1.35 V, 1.4 V (and possibly 1.45 V) as nominal values from the factory so the 1.425 V I set it at is not a worry. FYI, I have lowered MEM VDDQ below a bit to 1.325 V since that screenie was taken.

 

[ENTERPRISE]

8 minutes ago, J7SC_Orion said:

 

...this seems to be geared towards Zen3...below on the left are Zen timings for my 5950X and on the right is the 7950X3D. They are different animals to some extent. For V_SOC on the 7950X3D, 1.3 V is supposed to be the safe max but I stay well below that, even at the speeds I am running. Some earlier bios would automatically push that to 1.4 V and above when enabling EXPO (or XMP) and that led to several such CPUs frying themselves.

 

The MEM VDD is the main DDR5 voltage and I have seen folks push 1.7+ V with extra RAM coolers. My M-die comes in the 1.35 V, 1.4 V (and possibly 1.45 V) as nominal values from the factory so the 1.425 V I set it at is not a worry. FYI, I have lowered MEM VDDQ below a bit to 1.325 V since that screenie was taken.

 

Thanks for the insight. Im struggling to get FCLK stable at 2100. Tested at 1.29 VSOC originally but was unstable, I believe I improved stability a little by dropping the VSOC down to 1.2 and I also dropped MEM VDDQ and VDDP (I believe)to 1.35 for good measure but I based that mainly due to Buildzoids comments in the video you mentioned.

 

I need to do some more research as I would like to see if I can push to get 2100 FCLK stable. I can use Windows and run Memtest for a couple of hours but eventually the system will reboot. Any suggestions? 

 

I think Buildzoid did another video based on pushing the FCLK, might see tomorrow if that gives any more insight.

[J7SC_Orion]

13 minutes ago, ENTERPRISE said:

Thanks for the insight. Im struggling to get FCLK stable at 2100. Tested at 1.29 VSOC originally but was unstable, I believe I improved stability a little by dropping the VSOC down to 1.2 and I also dropped MEM VDDQ and VDDP (I believe)to 1.35 for good measure but I based that mainly due to Buildzoids comments in the video you mentioned.

 

I need to do some more research as I would like to see if I can push to get 2100 FCLK stable. I can use Windows and run Memtest for a couple of hours but eventually the system will reboot. Any suggestions? 

 

I think Buildzoid did another video based on pushing the FCLK, might see tomorrow if that gives any more insight.

 

V_SOC is a bit of a two-headed monster which fights with itself; high RAM clocks want it one way, but high InFin speeds the other way. It took me a while to find the common ground between the  at 1.255 V_SOC for InFin at 2167 MHz and DDR5 at 8000. I don't know much about your specific CPU and mobo/bios, but if you have the latest bios and thus AGESA loaded, I would start with V_SOC at 1.25 V and the same for DDR_VDDQ and CPU_VDDIO_MEM. BTW, you likely will have to do two separate profiles - one for 1:1 (ie 6400) and one for 1:2 (ie. 8000). Best to lock one of the two in first (I settled on 1:2 8000 quite early on due to the specific type of new M-die 2x 24 GB).

 

Once you get to 1:2 dividers and around 8000 MHz DDR5, you can test out V_SOC again at 1.25 V_SOC to 1.26 V_SOC but raise DDR_VDDQ and CPU_VDDIO_MEM to 1.3 V. Folks usually start out having these two the same but each mobo / chip combo can be different. 

 

Apart from TPU's Memtest64, I also use y-cruncher by itself, and the latest version of OCCT. I have not used Hydra Pro 1.x (1usmus) for RAM yet but I think it has a RAM stress / stability section for it as well .

 

FYI, I had someone suggest the other day to run Karhu at (3x 30,000=) 90,000 % . You run these things long enough and error will creep up which is why servers have the full-meal EEC deal onboard. 

 

 

[ENTERPRISE]

I have to say. Not enjoying this RAM OC experience this time round. Really struggling to get it stable. I have for now given up on anything beyond 2000 FCLK. 

 

Will stick with 2000 FCLK @ 6600MT/s. I had this stable with 1.45 Volts on the RAM, including VDDP and VDDIO with 1.29 SOC but intermittently if I reboot the PC it will hang at error A6, which looking up the error codes is something to do with SCSI and nothing obviously related to the OC, other than SCSI must be choking for some reason. Now testing at 1.5 volts and will see what happens. 

 

Really pissing me off now lol. Either I have a really bad Silicon Lottery or im badly missing something or this boards BIOS sucks. Note I have the latest BIOS for the MEG Ace X670E, so its not like im using an old AGESA.

[J7SC_Orion]

6 hours ago, ENTERPRISE said:

I have to say. Not enjoying this RAM OC experience this time round. Really struggling to get it stable. I have for now given up on anything beyond 2000 FCLK. 

 

Will stick with 2000 FCLK @ 6600MT/s. I had this stable with 1.45 Volts on the RAM, including VDDP and VDDIO with 1.29 SOC but intermittently if I reboot the PC it will hang at error A6, which looking up the error codes is something to do with SCSI and nothing obviously related to the OC, other than SCSI must be choking for some reason. Now testing at 1.5 volts and will see what happens. 

 

Really pissing me off now lol. Either I have a really bad Silicon Lottery or im badly missing something or this boards BIOS sucks. Note I have the latest BIOS for the MEG Ace X670E, so its not like im using an old AGESA.

 

...reminds of an older system (an Intel 8C/16t 5960X ROG Rampage LGA 2011) that sort of drove me crazy. It had a name: Nemesis. Time and time again, the relatively sane oc setting on Nemesis would pass every single short, medium and long stress test but once every 5 weeks or so, it would just throw (always different) errors and reboot int he middle of something. After resetting and testing everything, it would be fine again, until the next time...eventually, I gave that whole system away (set to stock speeds) since I still had a 5960X ES left over from another project.

 

More specific to your current sinuation, I would browse Buildzoid's channel for settings of other RAM types (rather than the specific 24 GB / stick new M-die I had linked). I know he did some earlier ones for A-die, and also the initial-release 'M-die'

[kaliz]

Do you have a zentimings screenshot Enterprise? And how did you set your CPU? Any PBO or Curve Optimizer?

[neurotix]

@ENTERPRISE

 

I would try 1.25v VDD_SoC and 1.35v Misc voltage for 2100MHz, and leave VDD_Mem on Auto. Give the RAM 1.435v and you will probably (if it's anything like mine) see the Mem controller voltage be at like 1.45v+.

 

That's how mine is set currently anyway and it's rock solid stable. Also I would suggest lowering the RAM speed to 6400 or 6200 if trying for 2100 fclk which is what I run. (6200MHz, 2100 fclk).

 

Hope this helps.