[Sammelthread] Intel DDR5 RAM OC Thread

Having a chiller is sooo pay 2 win, ahhh :geek:
Good results !
CB15 Extr is hard to run. Its very dynamic load at high VID. It runs because we are not silly and don't cut IA supply.
A bad curve or degraded CPU will simply not pass. Some come bad from the factory (partial fault is bios SVID, like 12% blame)
Yep, now I have LLC6 with IA_AC 0.30
That's a lot for LLC6. :)
Under SFT the Vcore is 1.19V-1.199V, under VST-VT3 it's 1.208V-1.199V
I belive the CPU part is okay, something memory related changed, but I will investigate it.

Didn't try the 8600 yet, since I had some hard times with the CPU pressure. :)
 
Wenn Du diese Anzeige nicht sehen willst, registriere Dich und/oder logge Dich ein.
Didn't try the 8600 yet, since I had some hard times with the CPU pressure. :)
I want to bonk your head (gently), because you maybe did the direct-die half hearted.
I have a feeling you did, or just expect too much perfectionism.
NLju.gif

(a culture joke, its a soft hammer to ppl who dont listen)

Crystal needs lapping (they are not flat) and careful removal of solder without doing structural damage (no knife/blade)
// it needs a proof that a waterdrop or oil drop can hold it based on ringing - without having gravity issues.
Both surfaces actually need close inspection and height measurement, to prevent micro cracks.
And LM should not be used long term. Especially as dryout will happen and unmount will have a high chance to damage crystal again.

But ya
Shouldnt have any pressure issues at first
Frame does the pressure - and block sits ontop thumbscrews/fingertipps tight.
Too much pressure on the block has a very high chance to cause cracks and is an excuse of having height issues or flatness issues.

Anywho,
I expect from you 8800 stable :)
It will take time, but thats my goal for your random sample~
Around 8400 for zebra's "lottery loss" sample.
 
Zuletzt bearbeitet:
I want to bonk your head (gently), because you maybe did the direct-die half hearted.
I have a feeling you didnt, or just expect too much perfectionism.
Anhang anzeigen 980186
(a culture meme, its a soft hammer)

Crystal needs lapping and careful removal of solder without doing structural damage (no knife/blade)
// it needs a proof that a waterdrop or oil drop can hold it based on ringing - without having gravity issues.
Both surfaces actually need close inspection and height measurement, to prevent micro cracks.
And LM should not be used long term. Especially as dryout will happen and unmount will have a high chance to damage crystal again.

But ya
Shouldnt have any pressure issues at first
Frame does the pressure - and block sits ontop thumbscrews/fingertipps tight.
Too much pressure on the block has a very high chance to cause cracks and is an excuse of having height issues or flatness issues.

Anywho,
I expect from you 8800 stable :)
It will take time, but thats my goal for your random sample~
Around 8400 for zebra's "lottery loss" sample.
Hehe, I used car polish to polish the surface. It's very clean. ☺️
When it wasn't clean enough I got some bad temps under heavy load.
I think my CPU_VDDQ moved a little. Rtt could be temp sensitive, right?
So if it work from lookup table the 10C difference on the Dimms could make a difference in stability too.
Maybe with this cold environment it would worth to play with PLL's a little.
The error always comes around 12 minutes.
 
I think my CPU_VDDQ moved a little. Rtt could be temp sensitive, right?
So if it work from lookup table the 10C difference on the Dimms could make a difference in stability too.
Thermals on memory dont have a big discharge issue.
Wordline (PoD) opens gate, and lets discharge flow from capacitor to wordline x bitline design
This electrical discharge moves at a near predictable constant.
SenseAmps then detect that wordline reached X targets and amplifies either the loss or the higher voltage.

Soo on a higher voltage state, it flows back to 1transistor1Capactor Cell, recharges back to not lose the data
And later whatever state is on bitline and wordline, is read out as data.

Its like opening the valve of something, letting X fluid flow back
Probing it what it is (ph) and then routing it back or discharging it fully
The result of this action, is the data.

Thermals will not influence the accuracy of the "voltage state" it is on
VPP_MEM messes with this gate opening time (when it will happen)

Voltage moving factor is just not affected that much by such little thermal change.
Now if we talk superconductor state, things change around a lot.
But this is not -40 to -50°C . Its not even -20°C.
Copper wires dont just lose their electrical transfer properties with such little thermal change.


What could have happen, is that SNR improved due to a colder cpu which takes less voltage to reach same clock strap
Its dynamic after all.
Soo more headroom and more margins.

It "shifting" is too unlikely
I just dont think board's ODT is dynamic.
it is on a lookup table, thats certain, but its for sure not thermal sensor aware.
It will stay the place it was, based on SA state.
It may shift sliightly if V/F curve (fused) started to shift

What is a higher chance, is that pads of the substrate slowly got charred and a remount scratched them back clean
Tho , that and flatness check is common, if you dismount it.
Need to clean slightly burned pads and ya. More margins is expected but not due to to board or memory.
Only due to vCore behaving slightly differently.
Maybe with this cold environment it would worth to play with PLL's a little.
Not the time to skew all voltages and termination points.
Please no. There are couple of ways to go around the issue, but non of those resolve it.

You may try to reduce 1.9v AUX t 1.86 or back to 1.8v
And if no change happens, just increase IA supply slightly , and see if that resolves the loss taken by lower thermals (like +0.05 more)
And if also that does nothing (doubt) , then we play with VPP_MEM - either 1.7v or 1.86v. Tho issue is on (internal) CPU side , not on mem.
If it was on mem, TM5 will catch it.

EDIT:
RTT lookup table is by capacity and sometimes Vendor
Its not by thermals, and RTTs sitt on the dimms only.
ODT & RON is a CPU sided topic. But Board is also not sub zero, soo no RON change.

EDIT2:
Insolation of the Board, or paint over it ~ will mess with impedance.
But not thermals. Copper is a quite stable material. Fiberglass too.
Same goes to Polyurathene. Its stable.

If it was a mem thermal "issue" then dropping its VDDQ_MEM by -30mV more
And doing the same on VDDQ_CPU ~ would resolve the issue. But i doubt.
Higher chance that VPP_MEM messing will do it, and higher chance that TM5 will complain if it had anything to do with mem.
 
Zuletzt bearbeitet:
Thermals on memory dont have a big discharge issue.
Wordline (PoD) opens gate, and lets discharge flow from capacitor to wordline x bitline design
This electrical discharge moves at a near predictable constant.
SenseAmps then detect that wordline reached X targets and amplifies either the loss or the higher voltage.

Soo on a higher voltage state, it flows back to 1transistor1Capactor Cell, recharges back to not lose the data
And later whatever state is on bitline and wordline, is read out as data.

Its like opening the valve of something, letting X fluid flow back
Probing it what it is (ph) and then routing it back or discharging it fully
The result of this action, is the data.

Thermals will not influence the accuracy of the "voltage state" it is on
VPP_MEM messes with this gate opening time (when it will happen)

Voltage moving factor is just not affected that much by such little thermal change.
Now if we talk superconductor state, things change around a lot.
But this is not -40 to -50°C . Its not even -20°C.
Copper wires dont just lose their electrical transfer properties with such little thermal change.


What could have happen, is that SNR improved due to a colder cpu which takes less voltage to reach same clock strap
Its dynamic after all.
Soo more headroom and more margins.

It "shifting" is too unlikely
I just dont think board's ODT is dynamic.
it is on a lookup table, thats certain, but its for sure not thermal sensor aware.
It will stay the place it was, based on SA state.
It may shift sliightly if V/F curve (fused) started to shift

What is a higher chance, is that pads of the substrate slowly got charred and a remount scratched them back clean
Tho , that and flatness check is common, if you dismount it.
Need to clean slightly burned pads and ya. More margins is expected but not due to to board or memory.
Only due to vCore behaving slightly differently.

Not the time to skew all voltages and termination points.
Please no. There are couple of ways to go around the issue, but non of those resolve it.

You may try to reduce 1.9v AUX t 1.86 or back to 1.8v
And if no change happens, just increase IA supply slightly , and see if that resolves the loss taken by lower thermals (like +0.05 more)
And if also that does nothing (doubt) , then we play with VPP_MEM - either 1.7v or 1.86v. Tho issue is on (internal) CPU side , not on mem.
If it was on mem, TM5 will catch it.

EDIT:
RTT lookup table is by capacity and sometimes Vendor
Its not by thermals, and RTTs sitt on the dimms only.
ODT & RON is a CPU sided topic. But Board is also not sub zero, soo no RON change.

EDIT2:
Insolation of the Board, or paint over it ~ will mess with impedance.
But not thermals. Copper is a quite stable material. Fiberglass too.
Same goes to Polyurathene. Its stable.

If it was a mem thermal "issue" then dropping its VDDQ_MEM by -30mV more
And doing the same on VDDQ_CPU ~ would resolve the issue. But i doubt.
Higher chance that VPP_MEM messing will do it, and higher chance that TM5 will complain if it had anything to do with mem.
I see, thank you for the detailed explanation.
Then I have to check it somewhere else. I want to bring back the good old 8533 profile before moving forward.
If I can't stabilize this one it doesn't make sense to move to the next frequency.

Still working on it. Now it runs with 1.16V SA. It didn't drop error strangely on the beginning, runs fine.
Under 1.18V was impossible before.
Maybe it's not due to the thermals, it could be the contact pressure too.
 
I see, thank you for the detailed explanation.
Then I have to check it somewhere else. I want to bring back the good old 8533 profile before moving forward.
If I can't stabilize this one it doesn't make sense to move to the next frequency.

Still working on it. Now it runs with 1.16V SA. It didn't drop error strangely on the beginning, runs fine.
Under 1.18V was impossible before.
Maybe it's not due to the thermals, it could be the contact pressure too.
SA requirement state will be messed with once you change substrate thermals , like vcore requirement will change
But it will not shift delta's unless you touch SA.
// VDD2/VDDQ_CPU will not shift. It will only shift once SA shifts. Board and mem are solid materials. MemICs are more sensitive than PCBs.
// IMC may like less voltage, but will not stop liking same higher voltage - just because it got colder. It makes no sense.
// It only makes little sense, if supply started to be lacking due to VIDs being lower & one was borderline stable. Aka IA_AC voltage up please.

SA change please coldboot and verify that SA VID is reported correctly (bugs exist)
Only if you mess with SA you mess with the rest

It makes sense to target 8600 when 8533 is a 133 strap and training may round the decimals badly at any time.
But on the same setup, same bios, same voltages - it makes little sense to not search for a fix
Doing PLL skewing or slopes shifting is not the way. It can resolve it but its not the way.

Im not sure if even messing with SA makes sense, when your old SA is unstable
Its rather something else that messes with it, and making a change, will make your time harder later when SA needs slight shift up.
Clock has to run at high and low voltages. More margins, the better. Less margins due to lower thermals, is absurd :-)
 
Zuletzt bearbeitet:
SA requirement state will be messed with once you change substrate thermals , like vcore requirement will change
But it will not shift delta's unless you touch SA.
// VDD2/VDDQ_CPU will not shift. It will only shift once SA shifts. Board and mem are solid materials. MemICs are more sensitive than PCBs.
// IMC may like less voltage, but will not stop liking same higher voltage - just because it got colder. It makes no sense.
// It only makes little sense, if supply started to be lacking due to VIDs being lower & one was borderline stable. Aka IA_AC voltage up please.

SA change please coldboot and verify that SA VID is reported correctly (bugs exist)
Only if you mess with SA you mess with the rest

It makes sense to target 8600 when 8533 is a 133 strap and training may round the decimals badly at any time.
But on the same setup, same bios, same voltages - it makes little sense to not search for a fix
Doing PLL skewing or slopes shifting is not the way. It can resolve it but its not the way.

Im not sure if even messing with SA makes sense, when your old SA is unstable
Its rather something else that messes with it, and making a change, will make your time harder later when SA needs slight shift up.
Clock has to run at high and low voltages. More margins, the better. Less margins due to lower thermals, is absurd :-)
Should be something with the IVR due to the less Vcore.
I just ran a fast 30 minutes test, but seems okay to me, will run 90 minutes later with a new cold boot. (this one was cold boot too)
For TVB I set -5C offset to do not boost to 58X, I will change that later again.
Everything else is the same, used the old foundation.
1710247489348.png
 

Anhänge

  • 8000_Y_6H.png
    8000_Y_6H.png
    301,7 KB · Aufrufe: 68
  • 8000_TM5-HCI.png
    8000_TM5-HCI.png
    283,8 KB · Aufrufe: 619
Zuletzt bearbeitet:
Should be something with the IVR due to the less Vcore.
Generally IVR stuff only changes due to throttling reasons.
We shouldn't have any.
Our limiters are there to filter spikes and be another layer of protection if Thermal Management System fails.
No need for wattage limiters with fluid-based cooling.

Just IA_AC tiny bit higher and its ok.
If it was too close to comfort
I mean you even played with Loadlines, while you shouldnt need to.
Let LLC be droopy.
Also to have a bigger voltage difference between loadtypes.
Voltage jumping on the core is not a bad thing

Asus engineers know their Board the best.
 
Zuletzt bearbeitet:
Generally IVR stuff only changes due to throttling reasons.
We shouldn't have any.
Our limiters are there to filter spikes and be another layer of protection if Thermal Management System fails.
No need for wattage limiters with fluid-based cooling.

Just IA_AC tiny bit higher and its ok.
If it was too close to comfort
I mean you even played with Loadlines, while you shouldnt need to.
Let LLC be droopy.
Also to have a bigger voltage difference between loadtypes.
Voltage jumping on the core is not a bad thing

Asus engineers know their Board the best.
I know, this was also a part of my investigation. ☺️
I think I will go as low as LLC3. But maybe in the end it will get a slight OC, and need the higher loadline.
We will see. I can at least calm down now, because the unstable memory pissed me up a little. ☺️
 
We will see. I can at least calm down now, because the unstable memory pissed me up a little. ☺️
🤭
Never give up~~
Whatever was, is past
See you made it further than before.

You can't see how it will be tomorrow,
but you can say that its better than yesterday and seek on working hard

Your choice is only to continue with whatever happens.
Things change in life, same as in tech. They break, to make room for a better new start
Same as plants vanish to soil, soo something better can grow with past experience and become more beautiful.

Its not the event that sets the tone. Its only on you what you do with it~~
Don't become the soil, stay a strong growing flower 🌸
Keep on working hard with the head up to the sun! 💪
 
Hallo,

Ich habe einen 12700k, ein MSI MPG Z790i WIFI Edge Mini ITX-Motherboard und G Skill DDR5 7200MHZ RAM. Ich versuche herauszufinden, wie weit ich das RAM-Übertakten treiben kann. Wie weit denken Sie ist möglich? Ich bin bis zu 7800 MHz gegangen und wenn ich mich richtig erinnere, konnte ich 14 Durchläufe von VT3 machen, glaube ich? Das sind alle Durchläufe, die meine Konfiguration ausgeführt hat.

Ich habe mit den Spannungseinstellungen herumgespielt und kann VT3 jetzt nicht bestehen lassen. Was schlagen Sie vor, dass ich meine Spannungseinstellungen setze? Ich werde auch die primären Timings auf 38 48 48 115 lockern.

Mir ist durchaus bewusst, dass es wahrscheinlich ein Traum ist, einen 12700k auf 7800 MHz zu bekommen, aber ich möchte alles versuchen. Danke.

Entschuldigung, wenn mein Deutsch schlecht ist, ich benutze ChatGPT, aber jemand hat mir gesagt, dass die Antworten auf die DDR5-Übertaktungsrätsel hier sind.
 

Anhänge

  • Screenshot 2024-03-10 203453.png
    Screenshot 2024-03-10 203453.png
    29,5 KB · Aufrufe: 41
7800-8000 ist ende bei dem mainboard. Wenn die CPU es schafft, sollte es schon gehen.
 
"Disable vddq training, set 1.38MVDDQ and 1.2VDDQ, 1.15SA, 1.35VDD2, MVDD depends on your timings kinda, usually around 1.56 for 8000C36-48
Then start finding optimal CPU_VDDQ which hovers around 160-220mV from MVDDQ
You need to find optimal VDDQ offset between CPU and MEM without VDDQ training on. Otherwise vddq training is correcting mistakes and it's all over the place"

got this from madness in bzoid stream
im posting here cause i have a question

question is, does MVDD to MVDDQ need an correct delta aswell?
Thats dependent on sticks? i think in some post i heard from veii something like 60mv will always work but i have no idea what to think about it
whats the range and what influences it(curious if sticks change behavior)

**removed the reboot unstable 60min ycruncher pass, probably didnt cold boot correctly? or just wrong values**
 
Zuletzt bearbeitet:
"
Disable vddq training, set 1.38MVDDQ and 1.2VDDQ, 1.15SA, 1.35VDD2, MVDD depends on your timings kinda, usually around 1.56 for 8000C36-48
Then start finding optimal CPU_VDDQ which hovers around 160-220mV from MVDDQ
You need to find optimal VDDQ offset between CPU and MEM without VDDQ training on. Otherwise vddq training is correcting mistakes and it's all over the place"


got this from madness in bzoid stream
220mv tx seens to be very happy
im posting here cause i have a question


question is, does MVDD to MVDDQ need an correct delta aswell?
Thats dependent on sticks? i think in some post i heard from veii something like 60mv will always work but i have no idea what to think about it
whats the range and what influences it(curious if sticks change behavior)


also i bought 2 new rams just to seek for Class A, do u think big change? currently on class v 7800gskills 2x16

im asking this question cause
1.2tx <=220=> 1.42 mvddq didnt work but
1.28tx <=220=> 1.5 mvddq seens to work

image.png

atm it is
1.56 mvdd
1.5 mvddq
1.28 vddq
1.17 sa
1.35 mc
Sorry to hijack your own quesiton, but I'm new to DDR5 and trying to wrap my head around the correct methodology for finding stable voltages.

I've started out with this;
1.56 mvdd
1.38 mvddq

1.2 CPU Vddq
1.35 CPU Vdd2
1.15 SA

Now do I change change cpu vddq until I find the correct delta between mvddq? Or do I change both vddq to find this? When will I know I have found the correct delta? Until VT3 passes?
 
I **think** , who found out about deltas and training giving you better chance to have stability consistency (which i think is vei and vei friends?)
wont tell, put X, put Y value, but would prefer to explain, cause every computer part is different, even if same model

at the end of the day u will still test billion variables to start "knowing ur hardware"
thats also why even with their knowledge they probably cant remote access and give people good ocs
too much variety still of hw like x thing

I sincerely know nothing, But trying to help you, relax all timings or put something u know its safe and start playing with
VDDQ training off, and VDDQ x MVDDQ deltas yes.. If u new to ddr5 start by searching what is easily done in your motherboard/setup
also dont forget mrc fast boot/fast boot off and cold booting, and thats aparently easily replicable in apex and i dont know what to tell for other boards..
i only had apexes and darks and.. i can tell dark for example.. it is very different from apex, VERY . need an amount of values that wont work in apex(high imc for example and some rules)

edit: after 60min of ycruncher of that picture i sent, i cold rebooted and 1min vt3 stable XD
edit2: it seens to have improved findings when i started to respect cold boot after changes
 
Zuletzt bearbeitet:
I **think** , who found out about deltas and training giving you better chance to have stability consistency (which i think is vei and vei friends?)
wont tell, put X, put Y value, but would prefer to explain, cause every computer part is different, even if same model

at the end of the day u will still test billion variables to start "knowing ur hardware"
thats also why even with their knowledge they probably cant remote access and give people good ocs
too much variety still of hw like x thing

But trying to help you, relax all timings or put something u know its safe and start playing with
VDDQ training off, and VDDQ x MVDDQ deltas yes..
also dont forget mrc fast boot/fast boot off and cold booting, and thats aparently easily replicable in apex and i dont know what to tell for other boards..
i only had apexes and darks and.. i can tell dark for example.. it is very different from apex, VERY . need an amount of values that wont work in apex(high imc for example)

edit: after 60min of ycruncher of that picture i sent, i cold rebooted and 1min vt3 stable XD
Thanks for your help. I loosened timings and set SA to 1.185 as I remember that working really well a few nights ago. I managed to get VT3 to actually run for 22 mins or so but the test always exits after 22mins for some reason. I followed a thread on overclockers forum to get a shortcut for Y cruncher to run VT3 but seems it is only doing it for 22 mins. Any ideas on how I can get it to run longer?

Edit: Found out how to make it run longer and pause before finishing test

Also, these voltages that I have tried seem to be working well. Now what do I do from here? Stick with these voltages or keep trying some others? When do we approach tightening timings, once voltages seem stable at loose timings with desired frequency?

Current voltages are:
1.56 MVDD
1.385 MVDDQ

1.25 CPU VDDQ
1.35 CPU VDD2
1.185 SA




1710272240716.png
 
Zuletzt bearbeitet:
Thanks for your help. I loosened timings and set SA to 1.185 as I remember that working really well a few nights ago. I managed to get VT3 to actually run for 22 mins or so but the test always exits after 22mins for some reason. I followed a thread on overclockers forum to get a shortcut for Y cruncher to run VT3 but seems it is only doing it for 22 mins. Any ideas on how I can get it to run longer?

Also, these voltages that I have tried seem to be working well. Now what do I do from here? Stick with these voltages or keep trying some others? When do we approach tightening timings, once voltages seem stable at loose timings with desired frequency?

Current voltages are:
1.56MVDD
1,385 MVDDQ

1.25 CPU VDDQ
1.35 CPU VDD2
1,185 SA




Anhang anzeigen 980293
@echo off
cd %~dp0
.\y-cruncher.exe pause:1 stress -D:30 -TL:5400 VT3


this is .bat in folder of ycruncher
is what i use, -TL is the time limit in seconds 5400 = 90minutes, i prefer to use duration -D 30seconds because i think ON/OFF is heavier and easier to find errors
and pause:1 wont close the cmd window

so some people would sui cide after seeing ur timings
please put
tRDRD_dg 8
tWRWR_dg 8
tRRDs 8
tFAW 32
tRRDl auto

and everything that u dont have sure set to auto, ur making worse than auto.
my recommendation for you, if ur not trolling me, is that u decrease frequency and set all auto xmp, and first have a ram validation/usable pc first
first you be stable with super safe before trying something else
like i know for sure anything i throw in my board with 7800freq will prob be stable and i know timings that are safe, u need know same but for your board, dont be ashamed of decreasing frequency, u will be winning performance
 
@echo off
cd %~dp0
.\y-cruncher.exe pause:1 stress -D:30 -TL:5400 VT3


this is .bat in folder of ycruncher
is what i use, -TL is the time limit in seconds 5400 = 90minutes, i prefer to use duration -D 30seconds because i think ON/OFF is heavier and easier to find errors
and pause:1 wont close the cmd window

so some people would sui cide after seeing ur timings
please put
tRDRD_dg 8
tWRWR_dg 8
tRRDs 8
tFAW 32
tRRDl auto

and everything that u dont have sure set to auto, ur making worse than auto.
my recommendation for you, if ur not trolling me, is that u decrease frequency and set all auto xmp, and first have a ram validation/usable pc first
first you be stable with super safe before trying something else
like i know for sure anything i throw in my board with 7800freq will prob be stable and i know timings that are safe, u need know same but for your board, dont be ashamed of decreasing frequency, u will be winning performance

I have done all auto XMP and overnight with Kahru and VT3 and it was stable. XMP is 7200mhz, I then bumped up to 7400mhz and followed the old DDR4 bible to tighten timings and tested and it was all stable.

Right now, I want to see if I can push frequency further so what I have done in the screen shot there is completely loosen all timings and see if 7800mhz will be stable first of all. Like I say, I am new to DDR5 overclocking and you said to completely loosen timings so that is what I did XD.

I'm also on a 12700k so probably not going to be as good IMC as 14th gen or even 13th gen.
 
auto will be loose enough, ddr5 # ddr4
just stop putting random numbers in some places that has no sense
 
auto will be loose enough, ddr5 # ddr4
just stop putting random numbers in some places that has no sense
They are not random numbers in random places. You are prescribing random numbers to me which has less sense than me loosening timings that I have stepped through for each and every setting, and that I have spent hours testing for each step. All they are is each one is loosened quite a bit to see if 7800mhz will run stable. Whether they are less than auto or not right now doesn't really matter as I am testing out max frequency. If I cant lower timings at a high frequency of course I will settle for lower frequency. Plus, how would you even know they are less than auto without seeing what my system configuration sets them to on auto.

DDR4 and DDR5 are not the same, but the methodology of finding a maximum frequency with favourable voltages, and then moving on to tertiary, secondary or primary timings is more or less the same.
 
They are not random numbers in random places. You are prescribing random numbers to me which has less sense than me loosening timings that I have stepped through for each and every setting, and that I have spent hours testing for each step. All they are is each one is loosened quite a bit to see if 7800mhz will run stable. Whether they are less than auto or not right now doesn't really matter as I am testing out max frequency. If I cant lower timings at a high frequency of course I will settle for lower frequency. Plus, how would you even know they are less than auto without seeing what my system configuration sets them to on auto.

DDR4 and DDR5 are not the same, but the methodology of finding a maximum frequency with favourable voltages, and then moving on to tertiary, secondary or primary timings is more or less the same.
image.png

i failed in start using forums once again
 
image.png

i failed in start using forums once again
You cant just link a Buildzoid video and make out its a mic drop without explaining specifically what makes this timing bad when trying to experiment with how far DDR5 memory can clock to i,e implications for stability. IIRC in this video he explains that TRDRD_dg 16 limits bandwidth which would be bad in a situation where I am trying to optimize timings in a bench marking scenario, but I am not. So if you can explain the first part, then fair enough.
 
you know you can just reply to the post above you without fully quoting everything, right?
 
It's not about what you prefer, it's about being considered bad manners stupidly full quoting half a page without any reason.
 
It's not about what you prefer, it's about being considered bad manners stupidly full quoting half a page without any reason.
Scroll through the forums and you can see pages and pages of people doing exactly the same thing,

Where in the rule book does it specifically say quoting posts you are replying to is bad manners?
 
you know that's the thing about manners, they are not written in a rule book however still being adhered to by the regular folk on here. Your reply makes it seem like you either do not understand that concept or you just choose to be a cunt. This is not OCN.
 
You cant just link a Buildzoid video and make out its a mic drop without explaining specifically what makes this timing bad when trying to experiment with how far DDR5 memory can clock to i,e implications for stability. IIRC in this video he explains that TRDRD_dg 16 limits bandwidth which would be bad in a situation where I am trying to optimize timings in a bench marking scenario, but I am not. So if you can explain the first part, then fair enough.
there is not a single scenario in EARTH
that
tRDRD_dg 8
tWRWR_dg 8
tFAW 32
tRDRD_sg 8
will NOT hurt any test, dont make sense to stretch the unnecessary for testing
otherwise people would be recommending test frequency at "Load RAW mhz"(which is like 126+ all timings+)
QXE said once that ISNT good to test super relaxed stuff aswell.. you just type something safe...
u can safely put:
tRDRD_dg 8, tWRWR_dg 8, tFAW 32, tRDRD_sg 8, tRDRD_sg Auto, tRAS 60, tRTP 12, tRDWR both 20, at 7800 ez
 
Zuletzt bearbeitet:
Hardwareluxx setzt keine externen Werbe- und Tracking-Cookies ein. Auf unserer Webseite finden Sie nur noch Cookies nach berechtigtem Interesse (Art. 6 Abs. 1 Satz 1 lit. f DSGVO) oder eigene funktionelle Cookies. Durch die Nutzung unserer Webseite erklären Sie sich damit einverstanden, dass wir diese Cookies setzen. Mehr Informationen und Möglichkeiten zur Einstellung unserer Cookies finden Sie in unserer Datenschutzerklärung.


Zurück
Oben Unten refresh