Overclock.net banner

10161 - 10180 of 10515 Posts

·
Overclock the World
Joined
·
2,101 Posts
hi @Veii

need ur help. :)

dram: 1.44v

View attachment 2466298
Simple Test size 16 (Error1) and 32mb (error 2)
It's a primary timing missmatch
What happens if you just try tRP 20 and drop tRAS to 32 ?
Thanks so much! Its MFR and i pushed the voltage at 1.4V so far in my light attempts to go beyond my current 3333mhz
Will try and report back, Danke nochmal
On the old days i wasn't very good with timings, but they are stable and perform - even tho are autocorrected up
MFR needs this high tRCD_RD , but they love voltage a lot
There is no issue dropping down to CL14 if you give it voltage
Mine where luckily on A1 PCB
very g**bage 2666CL16-18 kits
Hopefully you have a bitter bin :)

CAD_BUS timing on this target frequency you can replicate
i think cLDO_VDDP should be also replicatable, but better follow Matisse rulesets
SOC is different on 14nm than 12nm IMC

Edit:
Oh i ment also 25nm not 20nm 😐
25nm can run 1.6v without issues
 

·
Registered
Joined
·
6 Posts
Frendo,

Whats the reason you are running the Rtt values you have in the screenshot ?

I said previously to start with

RttNom / RttWr / RttPark @ 7/5/1
I got confused because i dont have RttWr 5 ... it have RZQ/2/1/3 and Hi-Rez.
For that reason i put Norm 7 / Wr 1 / Park 5 ,I thought you put it wrong 😅 sorry
 

·
Overclock the World
Joined
·
2,101 Posts
I got confused because i dont have RttWr 5 ... it have RZQ/2/1/3 and Hi-Rez.
For that reason i put Norm 7 / Wr 1 / Park 5 ,I thought you put it wrong 😅 sorry
RZQ=240ohm
visibly written on the DRAM calculator :)
Because this changes are low-level and borrowed from AMD CBS, not every brand translates them to ohm values
Causes less bugs that way
usually memory timings run in HEX
 

·
Registered
Joined
·
20 Posts
Noob 5900X owner here... coming from a 3900X and a stable system running a MSI Meg Ace x570 with GSkill Trident Royall 4000mhz CL15 kit. All was good until 5900X haha. I'm able to run memory at 3200mhz with a 1600fclk, but anything else is just not stable. Anyone have any basic/stable settings they can suggest? I've tried following a few setting listed here, but just am not having any luck? Should I just wait until a non-beta bios or can this actually work with the right settings?
 

·
Overclock the World
Joined
·
2,101 Posts
Noob 5900X owner here... coming from a 3900X and a stable system running a MSI Meg Ace x570 with GSkill Trident Royall 4000mhz CL15 kit. All was good until 5900X haha. I'm able to run memory at 3200mhz with a 1600fclk, but anything else is just not stable. Anyone have any basic/stable settings they can suggest? I've tried following a few setting listed here, but just am not having any luck? Should I just wait until a non-beta bios or can this actually work with the right settings?
settings ?
dram calculator?
cpu voltages ?
usually nearly identical to what matisse runs on ~ voltage wise
OC'ing T-Force 4133 cl18 bottom half of the post

Whole list of settings people run on Matisse ~ most should work on Vermeer
 

·
Registered
Joined
·
20 Posts
settings ?
dram calculator?
cpu voltages ?
usually nearly identical to what matisse runs on ~ voltage wise
OC'ing T-Force 4133 cl18 bottom half of the post

Whole list of settings people run on Matisse ~ most should work on Vermeer
Thanks, DRAM Calc keeps kicking a "not supported" error for me. All voltages I've left at AUTO for now.

I'll take a look at those links.
 

·
Registered
Joined
·
6 Posts
RZQ=240ohm
visibly written on the DRAM calculator :)
Because this changes are low-level and borrowed from AMD CBS, not every brand translates them to ohm values
Causes less bugs that way
usually memory timings run in HEX
Thaks Vei ... im new with this.

When you say RZQ = 240 , you mean the sum of the RTT Ohms equals need to equal to 240 ?
 

·
Registered
Joined
·
1,140 Posts
Thaks Vei ... im new with this.

When you say RZQ = 240 , you mean the sum of the RTT Ohms equals need to equal to 240 ?
Starting points is 240 ohms.

1 x 240 = 240
2 x 120 = 240
3 x 80 = 240
4 x 60 = 240
5 x 48 = 240
6 x 36 = 240
7 x 34.3 = 240

So from here we deduce Rtt values

A value of 1 is 240 ohms
A value of 2 is 120 ohms
A value of 3 is 80 ohms
A value of 4 is 60 ohms
A value of 5 is 48 ohms
A value of 6 is 36 ohms
A value of 7 is 34.3 ohms
 
  • Rep+
Reactions: bohemiocaa and Veii

·
Overclock the World
Joined
·
2,101 Posts
Thaks Vei ... im new with this.

When you say RZQ = 240 , you mean the sum of the RTT Ohms equals need to equal to 240 ?
Think simple :)
X divided by Y = real value
RZQ / something = something
 

·
Registered
Joined
·
59 Posts
Wow that close.
It would question here if the thermal pads on these really touch the heatsink and probably remove one side of it like:
Keep in mind, they often use double sided-glue thermal pads
very sticky ones at least and you don't want to rip appart the ICs
Else , alone comparing both A2 vipers ~ if they really are A2 dimms - would be a worthful thing to do
Although once removed from the DIMM Slots, i'd probably wipe them once with isopropanol
As strong current causes an oxidation layer on their pins. Same goes for older 3xx series boards (CPU and DIMM pins)

tWR 12 is awkward for this tRFC, 14 would be better - just a bit slower
But is this really a good set 🤔
It's a bit too unstable at this high voltage with yet "high" tRFC
My vipers run 1.45 without any airflow at all - open bench with their "heatsink"
Really questioning how well yours touch the thermal pads of them

Well anywho, go for GDM off 2T now :D
Then we shall see how stable the CAD/RTT setup really is :)
Your tRFC alone with GDM on, should go one step further down at least

Probably get them out first and make a close up picture on the bottom of the dimm - the traces
OC'ing T-Force 4133 cl18 visible like here, just a bit further away
Who knows, some old 4400 vipers apparantly where A0 not A2
It would be bad luck overall, but you might have lucked out - as A0 are less sensitive and can be placed on the slave set
What I can tell you right now is 1.52v or less with fan full speed these settings will error, but 1.53v or more they will be stable. So regardless of temps, 1.52v is just not enough.
1.50v 3733C16 without fan was stable overnight, could be stable less than 1.50v but I didn't get around to trying.

If I keep GDM on is there anything else I can change to reduce latency without having to increase voltage? Or anything I can change to reduce voltage without losing latency?
Basically is there any such thing as free performance I'm leaving on the table, or any settings I have requiring excessive voltage for little latency improvement?
And what do you suggest for tRFC? I have tRC * 6 = 252
2466330
 

·
Overclock the World
Joined
·
2,101 Posts
If I keep GDM on is there anything else I can change to reduce latency without having to increase voltage? Or anything I can change to reduce voltage without losing latency?
Basically is there any such thing as free performance I'm leaving on the table, or any settings I have requiring excessive voltage for little latency improvement?
And what do you suggest for tRFC? I have tRC * 6 = 252
View attachment 2466330
Free performance, tCWL 12
Should be stable with this set
later tWRRD 1 instead 3 and SCL drop to 3
Could be stable up to PCB
Another one is tRRD_L to 5 and tWTR_L to 10
~ maybe stable but needs perf testing with SiSoftware Sandra MultiCore Efficiency Test
it's a bit unclear, if it's better, just because Aida64 result will be better ~ even if stable

There is more on the table to do
But it's unlikely to work without a voltage increase
Well maybe :D
And what do you suggest for tRFC? I have tRC * 6 = 252
231 tRFC could work (use the calculator on manual mode & only try at the very end)
but it would only work with GDM enabled - because of the half speed mux units
120ns so far only worked on GDM on - and only a rare set of kits with 1.55v

Try it out
Sigh i miss my ryzen already ~ annoying itx board
Thank you for info.
I'm surprised that I booted with one SCL=2 not the two. I'm currently at 1.38v.

Seems I can do only 3733MT/s with my either CPU/Mobo, tried many voltages to boot 3800MT/s but no luck even at loose timings and with latest bios.. Not sure if there is any other trick to work on.
View attachment 2464295 View attachment 2464297
Oho , this is interesting
I think you might have discovered something 🧐
I wasn't always too sure on SCL , except that it's a PCB dependent thing
& either it will work or it won't even post
while tRRD_ & tWTR_ are IC depending + dual or single rank PCB

There is no way for you to drop tRCD_RD to 18 down ?
Even with a voltage increase ?
If this is 4x8GB you can use tRAS 26 without issues - then just adjust tRC or add as much tRP till tRC+tRFC are identical again
If this is 2x 16gb single rank, the same applies
- but if dual rank and not new 2028mb chips which are single rank
Then the lowest tRAS is 28 for you here
Depends really :D

I will keep this awkwardness in mind and let people try soon something
Don't think it's just random or board dependent 🤔

Still a bit unclear on SD DD ones
Some timings are just plain different from intels standartization
Hard to "learn" it really without comparable data online :)

EDIT: i'm sorry for the 19days lat reply
The forum bugged out on writing, but i'm happy i discovered the post nonthereless 🙇‍♂️
 

·
Registered
Joined
·
20 Posts
@Veii
Ok first attempt... figured I'd start with 3800mhz with a 1900fclk at 2T (XMP Profile 2 selected). Memory is Gskill Trident Royal 4000mhz CL15.

I ran TestMem5 with zero errors with the below config.

2466405


2466378


So DRAM kicks me an error when I set to 4000mhz, however, it will provide info on 3800mhz. Should I try the below settings next? I've had trouble with 1T since installing the 5900X.
2466406
 

·
Registered
Joined
·
66 Posts
Simple Test size 16 (Error1) and 32mb (error 2)
It's a primary timing missmatch
What happens if you just try tRP 20 and drop tRAS to 32 ?

On the old days i wasn't very good with timings, but they are stable and perform - even tho are autocorrected up
MFR needs this high tRCD_RD , but they love voltage a lot
There is no issue dropping down to CL14 if you give it voltage
Mine where luckily on A1 PCB
very g**bage 2666CL16-18 kits
Hopefully you have a bitter bin :)

CAD_BUS timing on this target frequency you can replicate
i think cLDO_VDDP should be also replicatable, but better follow Matisse rulesets
SOC is different on 14nm than 12nm IMC

Edit:
Oh i ment also 25nm not 20nm 😐
25nm can run 1.6v without issues
seems ok at first. but after 24th cycle of TM5 a ton of error came out on all test numbers. so im revert to old setting fclk 1900 1:1 which most stable so far :) . anyway i have install new kit 2x16gb (dual rank). below is default setting (no xmp). what are the best setting for twrwr/trdrd & procodt? 1st time got this dual rank. so the value to set a bit confusing me. thanks

2466403
 

·
Registered
Joined
·
6 Posts
RZQ=240ohm
visibly written on the DRAM calculator :)
Because this changes are low-level and borrowed from AMD CBS, not every brand translates them to ohm values
Causes less bugs that way
usually memory timings run in HEX

Vei.
Mongoled help me tu understand what are u saying to me 😅.... but i check every option in the Bios of B550-F , and no setting to manual offset in Data Bus.
Can be possible this mother + Bios dosent have? and the maximum ohms for RTTWr is 80 (Rzq/3)
 

·
Old crazy guy
Joined
·
2,277 Posts

·
Registered
Joined
·
586 Posts
BZ was not really pleased about the outcome :)
Well, for start his ram wasn't too good (he also stated that), yet he managed to go up to 5 ghz with cl15, which isn't bad at all if you ask me. Also, both cpu and memory overclock records are on this mobo, so it can't be by any means worse than Gigabyte. My 0.02$.

It's VRM is second to none, with 90A power stage (yes I know we'll most probably never use it at fully, but man, it's insanely sexy !) and the 2 dimm setup is also very sweet. Bios also feels superior and the looks wise... **** me, but that's the most beautiful board I ever seen lol.

Dunno about you, buddy, but I'll buy it in a split, the second I'll see it listed.
 

·
Old crazy guy
Joined
·
2,277 Posts
Well, for start his ram wasn't too good (he also stated that), yet he managed to go up to 5 ghz with cl15, which isn't bad at all if you ask me. Also, both cpu and memory overclock records are on this mobo, so it can't be by any means worse than Gigabyte. My 0.02$.

It's VRM is second to none, with 90A power stage (yes I know we'll most probably never use it at fully, but man, it's insanely sexy !) and the 2 dimm setup is also very sweet. Bios also feels superior and the looks wise... **** me, but that's the most beautiful board I ever seen lol.

Dunno about you, buddy, but I'll buy it in a split, the second I'll see it listed.
I'll buy as well in a split second :p

BZ was disappointed he couldn't do the same he had done with his beloved Master.
I wonder if he has a unique version because it was looking like when I was struggling with my Master...

But he also never changed once the termination parameters. Don't know... he doesn't like ram oc for sure!
BIOS just looks awesome, so many options that are missing on the Master... I feel that knowing every corner anything is possible.
 

·
Registered
Joined
·
586 Posts
I'll buy as well in a split second :p

BZ was disappointed he couldn't do the same he had done with his beloved Master.
I wonder if he has a unique version because it was looking like when I was struggling with my Master...

But he also never changed once the termination parameters. Don't know... he doesn't like ram oc for sure!
BIOS just looks awesome, so many options that are missing on the Master... I feel that knowing every corner anything is possible.
Yeah, most probably memory oc isn't his niche :)
 
10161 - 10180 of 10515 Posts
Top