Overclock.net › Forums › AMD › AMD CPUs › [Official] AMD Ryzen DDR4 24/7 Memory Stability Thread
New Posts  All Forums:Forum Nav:

[Official] AMD Ryzen DDR4 24/7 Memory Stability Thread - Page 80

post #791 of 1172
Quote:
Originally Posted by happyluckbox View Post

So should I purchase the cas14 128gb kit, or just stick with my cas15 128gb kit? I really want to try and stabilize 3200mhz, and I wonder if the higher quality ram 14cas kit will allow me to reach it. I mean I have samsung b die already (lucked out i guess) So maybe its not worth the attempt. Also.... im out of my warranty/return period...
If you are out of your return period I would just stick with what you have. The CAS14 B-Die SHOULD have been binned higher than the CAS15 sticks, so you might have reached higher clocks, but nothing is ever guaranteed. Unless you are able to sell your CAS15 stuff for a good price to offset the cost of the new CAS 14 128GB kit, I would just stick with what you have in my view.
post #792 of 1172
Quote:
Originally Posted by Keith Myers View Post

Well I am not a gamer so anything related to that is over my head. I see a couple of things that draw my attention. First is VTTDDR is disobeying the rule. It should be 1/2 Vddr. So with your Vddr set at 1.4V, VTTDDR should be at 0.7V NOT 0.75V.

The other one is that stability in higher memory clocks has been observed with setting Vddp to equal Vsoc or slightly less. That should be at ~ 1.0V with your Vsoc at 1.02V. You have it too low at 0.9V Gupsterg is the expert here.
Quote:
Originally Posted by WarpenN1 View Post

Hmmm then this is super weird but when I set my vttddr from 0.75v to 0.706v memtest fails immediately. I have DRAM voltage at 1.41 so 0.7062v (can't set it more precisely) should be about half :/

And with loosened timings 16-16-16-34-50 got me 200% coverage in memtest before fail compared 14-14-14-28-42 when it failed in couple of minutes.

Bump from 0.9v vddp voltage to 1.0v helped quite a lot! Got 600% coverage without error compared to 200% with 3333mhz 16-16-16-34-50^^ Now i'm verifying it with prime95 if it catches any instability that memtest did not.

Gonna try getting 14-14-14-28-42 timings stable next week smile.gif

But I don't understand why changing vttddr from 0.752v to 0.7062 (DRAM voltage 1.41v) makes RAMs very unstable
post #793 of 1172
Quote:
Originally Posted by Keith Myers View Post

Well I am not a gamer so anything related to that is over my head. I see a couple of things that draw my attention. First is VTTDDR is disobeying the rule. It should be 1/2 Vddr. So with your Vddr set at 1.4V, VTTDDR should be at 0.7V NOT 0.75V.

The other one is that stability in higher memory clocks has been observed with setting Vddp to equal Vsoc or slightly less snip........ .
Cheers for this bit of info, I hadn't come across this up till now! I require 1.45v for 3066mhz to be stable prime95 custom 14000mb stable +6 hrs.

Will test this when I get a chance
Ryzen 1600X
(13 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 5 1600X 1716SUT @ 4.0 Ghz | 1.425v | LLC 3 MSI X370 XPOWER Gaming Titanium Asus GTX 780 DCII OC @ 1332/3522 mhz | 1.3v F4-3600C15D-16GTZ @ 3200 mhz | 14-14-14-14-30-5... 
Hard DriveCoolingCoolingCooling
2 x Samsung 840 Pro's 128GB RAID 0 EK-FB MSI X370 XPower RGB Monoblock | 1/2" ID EK-FC780 GTX Ti DCII | EK-FC780 GTX DCII Backplate Swiftech D5 MCP655 | EK-XRES 100 Revo D5 
CoolingCoolingCoolingPower
Alphacool NexXxoS UT60 280 | Push/Pull Thermochill PA120.3 | Push 6 x Scythe GT AP-15, 1 x SST-FN123 @ 5v SeaSonic X750 Gold 750W PSU 
Case
Corsair Obsidian Series 750D Airflow Edition 
  hide details  
Reply
Ryzen 1600X
(13 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 5 1600X 1716SUT @ 4.0 Ghz | 1.425v | LLC 3 MSI X370 XPOWER Gaming Titanium Asus GTX 780 DCII OC @ 1332/3522 mhz | 1.3v F4-3600C15D-16GTZ @ 3200 mhz | 14-14-14-14-30-5... 
Hard DriveCoolingCoolingCooling
2 x Samsung 840 Pro's 128GB RAID 0 EK-FB MSI X370 XPower RGB Monoblock | 1/2" ID EK-FC780 GTX Ti DCII | EK-FC780 GTX DCII Backplate Swiftech D5 MCP655 | EK-XRES 100 Revo D5 
CoolingCoolingCoolingPower
Alphacool NexXxoS UT60 280 | Push/Pull Thermochill PA120.3 | Push 6 x Scythe GT AP-15, 1 x SST-FN123 @ 5v SeaSonic X750 Gold 750W PSU 
Case
Corsair Obsidian Series 750D Airflow Edition 
  hide details  
Reply
post #794 of 1172
Quote:
Originally Posted by WarpenN1 View Post


Bump from 0.9v vddp voltage to 1.0v helped quite a lot! Got 600% coverage without error compared to 200% with 3333mhz 16-16-16-34-50^^ Now i'm verifying it with prime95 if it catches any instability that memtest did not.

Gonna try getting 14-14-14-28-42 timings stable next week smile.gif

But I don't understand why changing vttddr from 0.752v to 0.7062 (DRAM voltage 1.41v) makes RAMs very unstable
I don't know. Maybe that is just what your RAM needs. It IS operating well out of "design" spec so voltages and settings may not scale linearly from the JEDEC standard.
post #795 of 1172
Just did a sort of budged upgrade with room to grow, replaced my 3930k/RIVE with:
Ryzen 7 1700x
Asus Prime X370 Pro (bios is updated to latest release w/ agesa 1.0.0.6a)
1x16GB G.skill Ripjaws V 3200MHz

I ordered the ram because it was on sale at $130, and it was not until i received it that i realized I had ordered a single dimm 16GB instead of 2x8GB dimms. I didn't send it back though because i planned to get another 16GB for 32GB total at a later point.

this 16GB dimm is Dual rank Hynix M-die
3200MHz XMP Timings are 16-18-18-38. full timing list 16-18-18-18-36-56-6-8-39

Originally tried default XMP DOCP, no POST and board did not self-reset
Then tried Auto with 2933 and 2800MHz. no post or auto reset on 2933 but did auto reset on 2800mhz. I have not seen it auto-reset since

then i went back after reading a little and lowered the main timings on 3066 and put my SOC up to 1.1625. Posted and loaded windows fine and i have been using this as a base ever since:
16-17-17-17-37-55-6-8-35 soc1.1625v

Computer has been more stable with high ram usage than low ram usage, ive yet to run an actual stress test because i doubt any would pass.
Playing GTAV yeilds about 50% ram usage and no issues while in game, but regular web browsing or leaving the PC idle will randomly BSOD. common errors "Kernel Check Security Failure" and "Unexpected Kernel Mode trap"

tried raising tFAW from 35 closer to the XMP spec of 39 causes non-post (non-post at 37). lowering tFAW tRRDL and tRRDS seemed to help improve stability, currently at tFAW=34, tRRDS=5, tRRDL=7.
tried raising SoC voltage further. no apparent stability change between 1.1625-1.175v. SoC voltage 1.1825v+ causes non-post.

Trouble is i haven't really found a proper guide detailing what to try and where to go with the other subtimings. I would like to get to 3200mhz. given that I have gotten this far I think it is possible, but I dont know where to start with all these timings.

here are the timing readings from thaiphoon:
geVHC1e.png
and here are my current timings as reported from Aida64:
KuKDxGN.png
and the xmp timings according to aida64
yjIYZZ7.png

thanks for any and all help that can be provided
Edited by Nemesis158 - 9/24/17 at 11:54pm
post #796 of 1172
Hi folks,

Running the following hardware:

1800X
ASRock X370 Taichi
G.Skill f4-4266c19d-16gtzr


I finally managed to get the following stable (48 Hours HCI):

No CPU Overclock (yet)

VDDCR_SOC = 1.10625v
VDDCR_SOC_LLC = Level 3
DRAM Voltage = 1.325v
VTT_DDR = 0.670v
VDDP = 0.980v
ProcODT = 60 ohm
CLDO_VDDP Voltage = 935mV

Gear Down Mode = on
Bank Group Swap = off
Bank Group Swap Alt = off




I'm happy with the above settings just to start gaming on my new build, but I would like to go for better.

The main problem I'm currently facing is as soon as Gear Down Mode is disabled and 1T enabled, my PC fails to boot for three tries, and then resets the RAM to default 2133MHz.

The closest I came to a successful boot is when I under-volt the IMC considerably, then the PC boots up to the BIOS screen, and then freezes right there.

I've tried numerous voltage combinations on the IMC and DRAM, but could not get it to boot to windows yet.

The Taichi does not have DRAM Boot Voltage like the CH6, do you recon this is why I'm unable to boot with GDM = off + 1T @ 3200MHz?

2933MHz with GDM =off + 1T boots just fine though.
Edited by SlushPuppy007 - 9/25/17 at 5:56am
post #797 of 1172
Sharing my oc, Ryzen 5 1600, Gigabyte Gaming 3, KFA2(Galax) 3600C17 16GB




Anything higher or tighter would give me memtest errors.
Impressed by the performance gain from subtimings, i've seen 15% gain in some games.
Edited by Darkomax - 9/25/17 at 7:11am
post #798 of 1172
Hey guys, any tips on getting 4ghz stable on my setup? I tried upping my frequency from 3.9 to 3.925 then 3.95, then 3.975, and I didnt need to increase vcore at all! BUT going to 4.0 requires a massive leap in vcore from 1.368 to 1.41++ and even then its not quite stable. I'm starting to think its not vcore and some other voltage instead. Any ideas?
post #799 of 1172
Quote:
Originally Posted by Keith Myers View Post

The other one is that stability in higher memory clocks has been observed with setting Vddp to equal Vsoc or slightly less. That should be at ~ 1.0V with your Vsoc at 1.02V. You have it too low at 0.9V Gupsterg is the expert here.

No expert blushsmiley.gif . Still learning lots and do not have the depth of knowledge or experience to have that moniker blushsmiley.gif . More of a viral poster of info, from a) testing b) picked up from others shares smile.gif .

I had issues stabilising a 3466MHz RAM profile on C6H with UEFI 1403 onwards. I then had to reassess settings. I then opted to lower VDDP. I and another prior to me deciding to do this, had seen that UEFI upto 99xx were ~900mV on VDDP and later ones ~975mV. It had not posed an issue for me on UEFI between 99xx and 1403, but the ones after it was.

When I got Ryzen at launch and noted VDDP probe point on C6H I asked The Stilt what it was, link. Here is a later dated post by The Stilt concerning it again, link.

CLDO_VDDP I found helped with moving memory holes. On one CPU sample that I did more testing then others I had. I noted lowering it moved the hole down the frequency range and upping it the other way. Another C6H owner also experienced this. If all CPUs behave this way I have no idea. I also noted if I used say excessively low CLDO_VDDP, so all RAM MHz between 2133MHz and 3466MHz would post, I'd have stability issues in OS. So excessively low was no good, so what must be used is what is right for target RAM MHz.

ProcODT was a tricky one to get right IMO. For example I started using 60Ω mainly with my F4-3200C14D-16GTZ. One day when testing a custom setup of P95 I had dropped threads. Changing to 53.3Ω resolved issues. Again we may find a profile passes Memtest/GSAT and other tests but another may highlight a shortcoming in settings.

Another tricky one was CAD Bus. I had opt'd for 30Ω on higher than 3333MHz:-

ClkDrvStren
AddrCmdDrvStren
CsOdtDrvStren
CkeDrvStren

I found SuperPi was sensitive to changes to this, I'd have runs which failed.

On the whole I have never ever done as much RAM tweaking as I am for Ryzen. Has been nice to get into to pick up tweaks, test methods, but can be time consuming and tedious redface.gif .

SOC on all CPUs I have had I also tune. I start with default and then start increasing as needed for target RAM MHz, usually I'm going for 3200MHz, 3333MHz or 3466MHz.

I have only used 1 Ryzen board, the C6H and one RAM kit more than the 3 I had at one time, F4-3200C14D-16GTZ. With this combo I have used 3x R7 1700 and 1x R7 1800X. I shall be testing a 2nd R7 1800X soon smile.gif .
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
post #800 of 1172
Gupsterg, thanks for chiming in. You are more an "expert" than I. And done 10X the amount of RAM testing. Furthermore, your links back to the Stilt comments finally shed light on just "what the heck" is Vddp anyway. My searches for it had turned up nothing but noise in other science branches. Lots of good info in your post.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
Overclock.net › Forums › AMD › AMD CPUs › [Official] AMD Ryzen DDR4 24/7 Memory Stability Thread