*Official* Intel DDR4 24/7 Memory Stability Thread - Page 1362 - Overclock.net - An Overclocking Community

Forum Jump: 

*Official* Intel DDR4 24/7 Memory Stability Thread

Reply
 
Thread Tools
post #13611 of 13916 (permalink) Old 08-01-2020, 09:12 AM
Facepalm
 
Falkentyne's Avatar
 
Join Date: Dec 2007
Location: Riverside
Posts: 8,092
Rep: 465 (Unique: 322)
Quote: Originally Posted by XGS-Duplicity View Post
for 4500, trrd_L @ 6 and trfc @ 320 consistantly would not go past 37% in karhu and would not let me play games, even with air conditioning blasting and cardboard box vent cooling.


for 4500, trrd_l @ 4 and trfc @ 280 made it to 300% in karhu(with air blasting/vent box and tertiary on auto) and let me play starcraft with no crashing at 80F ambient- without the air conditioning and without cardboard box vent cooling.

trrd_L @ 6 @ 4200 and trfc @ 378 for 4200 is memtest stable, trrdL @ 4 and trfc lower was not memtest stable for 4200(same cas 17-18-18-38-2T). volts for stable 4200 c17 is 1.45v bios set, +50mv for every 100mhz higher or +50mv for each cas lower-in theory.



Higher trfc was less stable for 4500.




But trrdL @ 6 couldn't even play anything at 4500 even though its looser.



It doesn't make sense to me why worse ambient and tighter timings was more stable for starcraft. It literally goes against everything I learned so far.


I tried txp @ 4, d6 or no train.
trefi 65534 is no go completely, gives system service exception, either more vcore or more vdimm or more sa/io, not willing to go anymore higher..
trefi 32768 gives system service exception, either more vcore or more vdimm or more sa/io, not willing to go anymore higher.
Tuned tertiaries crash in starcraft.


I wonder if system service exception with air blasting/vent cooling made chip fall below vcore ram requirements for 4500mhz due to utilizing turbo TVB for the chip?


I think c16 or c17 4400 with good subs would probably be more feasible but this kit will not go past the very beginning of a training sequence at all for 4300 or 4400 on this board ever. i've tried a hundred times. But it can train 4500 fine lol. Can someone explain why this is?
There's a bug with DDR VTT where what you set in BIOS ends up big time incorrect, and you have to check in health settings to make sure the value you entered is actually correct and not 50-100mv higher or lower. Not sure why that bug happens and I can't be bothered with it anymore. Haven't seen this happen on Z490 Master at all, except the DDR VTT is always like 25mv too low, even on auto (so you need to manually raise it 25mv).

[email protected] ghz, RX Vega 64, 32GB DDR4, Gigabyte Aorus Master, Seasonic Platinum 1000W, Corsair 760T
Alt: MSI GT73VR Throttlebook with 7820HK @ 4.7 ghz, GTX 1070 MXM TDP mod to 230W, 32 GB RAM

Quote: Originally Posted by sakete View Post
Well, I want you to know I have an academic degree in speculation.
Falkentyne is offline  
Sponsored Links
Advertisement
 
post #13612 of 13916 (permalink) Old 08-01-2020, 10:15 AM
What's your "stable"?
 
skullbringer's Avatar
 
Join Date: Apr 2013
Posts: 291
Rep: 35 (Unique: 21)
Quote: Originally Posted by Falkentyne View Post
There's a bug with DDR VTT where what you set in BIOS ends up big time incorrect, and you have to check in health settings to make sure the value you entered is actually correct and not 50-100mv higher or lower. Not sure why that bug happens and I can't be bothered with it anymore. Haven't seen this happen on Z490 Master at all, except the DDR VTT is always like 25mv too low, even on auto (so you need to manually raise it 25mv).
is turbovcore readout of vttddr reliable, or only health section sensors in bios, do you know?

Main Rig
(10 items)
Test Bench
(10 items)
CPU
Intel core i7-8086K
Motherboard
Asus Maximus XI Gene
GPU
EVGA GeForce GTX 1080 Ti Gaming SC Black Edition
RAM
G.Skill TridentZ 3600 C15-15-15-35 1.35V
Hard Drive
Samsung 960 Evo
Power Supply
Corsair RM1000
Cooling
Custom Loop
Case
Lian-Li O11 Dynamic
Monitor
Acer XB270HU
Keyboard
Ducky One 2 SF
CPU
Intel core i9-10900KF
Motherboard
Asus Maximus XII Apex
GPU
Asus GeForce GTX 750 DCU2
RAM
G.Skill TridentZ Neo 3800 C14-15-15-35 1.35V
Hard Drive
Sandisk Extreme 240GB
Power Supply
Seasonic Focus GX-650
Cooling
Supercool Computer Direct Die Block Mk 6, 6x 120mm Radiators
Case
Open Benchtable
Monitor
Benq XL2720
Keyboard
KBC Poker 2 Cherry MX Clear
▲ hide details ▲
skullbringer is offline  
post #13613 of 13916 (permalink) Old 08-01-2020, 10:45 AM
New to Overclock.net
 
fly1ngh1gh's Avatar
 
Join Date: Nov 2018
Posts: 42
Rep: 1 (Unique: 1)
Hi guys.
Here is another test after disconnecting the PC. I had to raise the SA by 1 step for karhu.
4500Mhz 16-16-16
Attached Thumbnails
Click image for larger version

Name:	TeamGroup 4500Mhz 16-16-16 - 1.jpg
Views:	74
Size:	494.0 KB
ID:	363472  

Click image for larger version

Name:	TeamGroup 4500Mhz 16-16-16 - 2.jpg
Views:	67
Size:	475.6 KB
ID:	363474  


/i7 8086k 5.0Ghz(Avx-0), 1.200v+ASRock z370 Taichi/
/Corsair Vengeance LpX ReD 16Gb 2*8Gb 4500Mhz/
/Palit 1080 GameRock Premium 2100/12500Mhz/
fly1ngh1gh is offline  
Sponsored Links
Advertisement
 
post #13614 of 13916 (permalink) Old 08-01-2020, 10:49 AM
Facepalm
 
Falkentyne's Avatar
 
Join Date: Dec 2007
Location: Riverside
Posts: 8,092
Rep: 465 (Unique: 322)
Quote: Originally Posted by skullbringer View Post
is turbovcore readout of vttddr reliable, or only health section sensors in bios, do you know?
Referring to Z390 Aorus Master. Nothing to do with Asus.
Asus readout is correct, however I have seen bugs where changing DDR Voltage on Asus does not change DDR Termination voltage sometimes (saw setting 1.5v DDRV once and DDR Termination was still 0.675v, and yes it was on "Auto")...

[email protected] ghz, RX Vega 64, 32GB DDR4, Gigabyte Aorus Master, Seasonic Platinum 1000W, Corsair 760T
Alt: MSI GT73VR Throttlebook with 7820HK @ 4.7 ghz, GTX 1070 MXM TDP mod to 230W, 32 GB RAM

Quote: Originally Posted by sakete View Post
Well, I want you to know I have an academic degree in speculation.
Falkentyne is offline  
post #13615 of 13916 (permalink) Old 08-01-2020, 02:33 PM
New to Overclock.net
 
reflex75's Avatar
 
Join Date: Aug 2016
Location: Paris, France
Posts: 189
Rep: 3 (Unique: 3)
Quote: Originally Posted by skullbringer View Post
people advertise karhu over hci because it's "faster", when in reality a 10 minute karhu run just says nothing about stability. So "faster" testing is actually counterproductive, as it is inconsistent and may lead to false security, even when you take heat out of the equation.

so "calling something stable" should not just be about testing to a certain percentage, but testing for a certain percentage and time.

like "test for 1000% hci or 1 hour", whichever takes longer
Memory OC is not straightforward like CPU OC, and it's why it's more rewarding when it works.
You have millions of timings and frequency combinations that match more or less better.
And then you have both PCB and ICs quality to prevent loosing signal at higher temperature.
And moreover, all memory cells are not equal.
There are always weak points that throws errors first when stressed.
And when testing, you can not stress all memory at once.
If the weak point is at the beginning of the test, you can have errors sooner...
reflex75 is offline  
post #13616 of 13916 (permalink) Old 08-01-2020, 04:36 PM
New to Overclock.net
 
Join Date: Sep 2019
Posts: 167
Rep: 12 (Unique: 7)
Quote: Originally Posted by zGunBLADEz View Post
Thats cool, what board i mean what io/sa and cooling used? :rollseyes: nah i don't care for that..



at this point really looking the last 20pages back this topic has become a top 10 3dmark type of topic....


But, Specially...
wheres the hci run?
As far as I remember, there was IO / SA = 1.25 / 1.33 and stable in Karhu, LinX, TM5 Extreme1. BIOS BETA.
Gen. is offline  
post #13617 of 13916 (permalink) Old 08-02-2020, 06:58 AM
New to Overclock.net
 
Nizzen's Avatar
 
Join Date: Apr 2013
Location: Norway
Posts: 2,220
Rep: 142 (Unique: 62)
On topic?

From #1 post in this thread:

"HCI
HCI consider 1000% to be the 'golden standard' however for larger densities this can be time consuming. A minimal coverage of two laps (200%) is required to be added to the table for HCI for density over 16GB. 16GB or less requires a minimum of 4 laps (400%)"

10900k direct die cooled
Team T-Force Extreem 4500c18 2x8GB aircooled
Asus Apex xii
Attached Thumbnails
Click image for larger version

Name:	20200802_155709.jpg
Views:	98
Size:	3.59 MB
ID:	363644  


AMD systems: 1950x - 1800x - 3900x
Intel systems: 7980xe - 10900k - 9900k
Storage: 3xOptane 900p 480GB - 4x Adata sx8200pro 1TB - Areca Raid-6 144TB- 4x samsung 850pro raid-0- 3x sm961 nvme 1TB
MB's : Asrock x399 pro - Asorck Taichi x370 - Asrock x570 Taichi - Apex x299 - Apex z490- Apex z390

Last edited by Nizzen; 08-02-2020 at 07:52 AM.
Nizzen is offline  
post #13618 of 13916 (permalink) Old 08-02-2020, 07:08 AM
New to Overclock.net
 
reflex75's Avatar
 
Join Date: Aug 2016
Location: Paris, France
Posts: 189
Rep: 3 (Unique: 3)
Quote: Originally Posted by Nizzen View Post
On topic?
4700 CR1 is pretty impressive!
Apex really shines for higher mem frequency thanks to its 2 dimm only slots than are closer to the CPU (for shorter traces).
Nice if you don't need more than 16GB, because 2x16 are too slow...
reflex75 is offline  
post #13619 of 13916 (permalink) Old 08-02-2020, 08:29 AM
hazed is lazed
 
eminded1's Avatar
 
Join Date: Feb 2009
Location: Rhode Island, US
Posts: 267
Rep: 4 (Unique: 4)
i got a 64gb kit of gskill 3600mhz ram at c18 im looking to see how far i can take it. the name of kit is F43600c18D64GVK anyone have any expirernce with this kit and how far you can OC it.. it is 2x32gb Dimms.. i have them in a ASrock z490 Taichi Motherboard. im gona try to OC it and ill post back

haze is lazed

Last edited by eminded1; 08-02-2020 at 08:34 AM.
eminded1 is offline  
post #13620 of 13916 (permalink) Old 08-02-2020, 12:20 PM
What's your "stable"?
 
skullbringer's Avatar
 
Join Date: Apr 2013
Posts: 291
Rep: 35 (Unique: 21)
pretty happy as a stop off point for today
Attached Thumbnails
Click image for larger version

Name:	Screenshot (90).png
Views:	70
Size:	851.1 KB
ID:	363714  


Main Rig
(10 items)
Test Bench
(10 items)
CPU
Intel core i7-8086K
Motherboard
Asus Maximus XI Gene
GPU
EVGA GeForce GTX 1080 Ti Gaming SC Black Edition
RAM
G.Skill TridentZ 3600 C15-15-15-35 1.35V
Hard Drive
Samsung 960 Evo
Power Supply
Corsair RM1000
Cooling
Custom Loop
Case
Lian-Li O11 Dynamic
Monitor
Acer XB270HU
Keyboard
Ducky One 2 SF
CPU
Intel core i9-10900KF
Motherboard
Asus Maximus XII Apex
GPU
Asus GeForce GTX 750 DCU2
RAM
G.Skill TridentZ Neo 3800 C14-15-15-35 1.35V
Hard Drive
Sandisk Extreme 240GB
Power Supply
Seasonic Focus GX-650
Cooling
Supercool Computer Direct Die Block Mk 6, 6x 120mm Radiators
Case
Open Benchtable
Monitor
Benq XL2720
Keyboard
KBC Poker 2 Cherry MX Clear
▲ hide details ▲
skullbringer is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 16 (1 members and 15 guests)
munternet
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off