Ryzen 2600 performance down with 3333+ Ram at same CL - Overclock.net - An Overclocking Community

Forum Jump: 

Ryzen 2600 performance down with 3333+ Ram at same CL

 
Thread Tools
post #1 of 7 (permalink) Old 01-31-2019, 12:17 AM - Thread Starter
New to Overclock.net
 
rul3s's Avatar
 
Join Date: Dec 2014
Location: Spain
Posts: 120
Rep: 1 (Unique: 1)
Ryzen 2600 performance down with 3333+ Ram at same CL

Hi!
Yesterday I was doing some tests and I found a curious thing. I have my [email protected] paired with Samsung B-Die at [email protected] 1.4v, subtimmings setted up as Ryzen Dram Calculator says at 3200 Fast Presset. My test method is with Realbench v2.43 hwbot edition.
The thing is that until 3200, Realbench score was growing up like:

2133MHz: 132000
2933: 145000
3200:153000

But, when surpassing 3200 on RAM, performance is not going up anymore, it's going down like:
3400: 152000
3466:151000
3533: 148000

All this, with the SAME timmings as 3200... If I look at Realbench score, what goes down most of it it's image rendering, having 174000 points at 3200MHz and 162000 at 3533MHz for example... but on the other tests, like encoding and heavy multitasking it's winning for a little versus 3200.

Anyone know what's the reason for this behaviour? Maybe 3200 RAM gives the best divisor (1:16) to work on ryzen?

PhanteksR7
(16 items)
CPU
Ryzen R7 3700x
Motherboard
Asus x470 Strix F
GPU
RTX 2070 Windforce OC
RAM
KFA2 HOF Extreme DDR4 4000CL19 BDie
Hard Drive
Samsung 970EvoPlus 500GB NVMe
Hard Drive
Samsung 850Pro 512GB
Hard Drive
Samsung 860QVO 2TB
Power Supply
Seasonic Focus+ 650W Platinium
Cooling
Noctua NH-U12P
Case
Phanteks Enthoo Pro M SE
Operating System
Windows 10 PRO
Monitor
Samsung C24FG70
Keyboard
Corsair K70 RGB
Mouse
Zowie FK2
Mousepad
Razer Goliathus Cosmic Speed
Audio
Asus Xonar STX
▲ hide details ▲
rul3s is offline  
Sponsored Links
Advertisement
 
post #2 of 7 (permalink) Old 01-31-2019, 12:27 AM
The last VRM burner
 
Join Date: Mar 2018
Posts: 1,638
Rep: 24 (Unique: 13)
Secondaries/Tertiaries are getting worse.

PC for all
(18 items)
CPU
i7 5775c @4.2GHz/3.8GHz
Motherboard
Gigabyte Z97X Gaming 5
GPU
Soon to be 2070 Super FE
RAM
KHX24C11T3K2/16X 11-12-12-32-1T-250
Hard Drive
Corsair MP510
Hard Drive
Samsung Evo 860
Power Supply
SeaSonic Focus Plus 750W
Cooling
Raijintek Tisis lapped+ Thermal Grizzly Kryonaut
Case
LianLi 330X
Operating System
Windows 10 64-bit Pro
Monitor
Benq GL2250 80Hz
Keyboard
Corsair Strafe Red Silent
Mouse
SteelSeries Rival 310
Mousepad
Artisan FX Raiden
Audio
Audio in Motion SC808 + MUSES8820 + 2x 2xLME49990
Audio
AKG M220
Other
NZXT Hue+
Other
Mayflash F300 Arcade Stick
▲ hide details ▲
BroadPwns is offline  
post #3 of 7 (permalink) Old 01-31-2019, 12:33 AM
New to Overclock.net
 
neojack's Avatar
 
Join Date: Oct 2011
Posts: 153
Rep: 3 (Unique: 3)
i have seen posts with random problems with ryzen + ram OC, where the solution was to simply blow a fan to the ram. you may try this temporarly just for testing.
neojack is offline  
Sponsored Links
Advertisement
 
post #4 of 7 (permalink) Old 01-31-2019, 12:57 AM
DRAM Obsessor
 
CJMitsuki's Avatar
 
Join Date: Mar 2018
Posts: 1,019
Rep: 7 (Unique: 6)
Quote: Originally Posted by rul3s View Post
Hi!
Yesterday I was doing some tests and I found a curious thing. I have my [email protected] paired with Samsung B-Die at [email protected] 1.4v, subtimmings setted up as Ryzen Dram Calculator says at 3200 Fast Presset. My test method is with Realbench v2.43 hwbot edition.
The thing is that until 3200, Realbench score was growing up like:

2133MHz: 132000
2933: 145000
3200:153000

But, when surpassing 3200 on RAM, performance is not going up anymore, it's going down like:
3400: 152000
3466:151000
3533: 148000

All this, with the SAME timmings as 3200... If I look at Realbench score, what goes down most of it it's image rendering, having 174000 points at 3200MHz and 162000 at 3533MHz for example... but on the other tests, like encoding and heavy multitasking it's winning for a little versus 3200.

Anyone know what's the reason for this behaviour? Maybe 3200 RAM gives the best divisor (1:16) to work on ryzen?
Are you fully testing the ram after increasing the frequency to make sure there arent any memory errors? Also, are any of your timings set to AUTO? If so, then your AUTO settings/timings are being relaxed to allow for more overhead. This doesnt have to be timings, it can also be Cad_Bus, Proc_Odt, etc. If you arent testing the memory for errors after any change to memory settings then you may have some intermittent errors that will drop the score and could eventually corrupt system drivers resulting in the OS not being able to boot up.

CPU- AMD Ryzen [email protected] *
MOTHERBOARD- ROG Crosshair VII Hero
GPU- EVGA GTX 1080ti SC2 w/ XOC [email protected] (Morpheus Cooler)
RAM- G.SKILL TridentZ 32gb (4 x 8gb) @ 3800mhz (14-15-15-24-38-1T)
COOLING- CORSAIR Hydro Series H115i Pro w/ Noctua NF-A14 (x7)
CASE- Silverstone RL06 Pro (Modified)
MONITORS- VIOTEK GN34DB (x2)


CJMitsuki is offline  
post #5 of 7 (permalink) Old 01-31-2019, 01:04 AM - Thread Starter
New to Overclock.net
 
rul3s's Avatar
 
Join Date: Dec 2014
Location: Spain
Posts: 120
Rep: 1 (Unique: 1)
Quote: Originally Posted by BroadPwns View Post
Secondaries/Tertiaries are getting worse.
COuld you explain a little more please? I don't know what do you mean about secondaries tertiaries....

Quote: Originally Posted by neojack View Post
i have seen posts with random problems with ryzen + ram OC, where the solution was to simply blow a fan to the ram. you may try this temporarly just for testing.
My DRAM kit is KFA2 HOF Extreme DDR4 4000 that comes with ceramic heatsinks, they're not getting even hot at 1.4v

Quote: Originally Posted by CJMitsuki View Post
Are you fully testing the ram after increasing the frequency to make sure there arent any memory errors? Also, are any of your timings set to AUTO? If so, then your AUTO settings/timings are being relaxed to allow for more overhead. This doesnt have to be timings, it can also be Cad_Bus, Proc_Odt, etc. If you arent testing the memory for errors after any change to memory settings then you may have some intermittent errors that will drop the score and could eventually corrupt system drivers resulting in the OS not being able to boot up.
Yes, HCI testing for 100% of memory. Also, nothing set to auto, all parameters setted up as Ryzen Dram Calculator says... GearDown, BankSwap disabled and resistors setted up correctly also....

PhanteksR7
(16 items)
CPU
Ryzen R7 3700x
Motherboard
Asus x470 Strix F
GPU
RTX 2070 Windforce OC
RAM
KFA2 HOF Extreme DDR4 4000CL19 BDie
Hard Drive
Samsung 970EvoPlus 500GB NVMe
Hard Drive
Samsung 850Pro 512GB
Hard Drive
Samsung 860QVO 2TB
Power Supply
Seasonic Focus+ 650W Platinium
Cooling
Noctua NH-U12P
Case
Phanteks Enthoo Pro M SE
Operating System
Windows 10 PRO
Monitor
Samsung C24FG70
Keyboard
Corsair K70 RGB
Mouse
Zowie FK2
Mousepad
Razer Goliathus Cosmic Speed
Audio
Asus Xonar STX
▲ hide details ▲
rul3s is offline  
post #6 of 7 (permalink) Old 01-31-2019, 01:24 AM
DRAM Obsessor
 
CJMitsuki's Avatar
 
Join Date: Mar 2018
Posts: 1,019
Rep: 7 (Unique: 6)
Quote: Originally Posted by rul3s View Post
COuld you explain a little more please? I don't know what do you mean about secondaries tertiaries....



My DRAM kit is KFA2 HOF Extreme DDR4 4000 that comes with ceramic heatsinks, they're not getting even hot at 1.4v



Yes, HCI testing for 100% of memory. Also, nothing set to auto, all parameters setted up as Ryzen Dram Calculator says... GearDown, BankSwap disabled and resistors setted up correctly also....
What he meant was the secondary and tertiary timings but you have them set manually so they arent changing but that isnt to say that the timings are optimal for a higher frequency. I dont rely on Realbench to show true memory performance. Geekbench 3 is probably the best at showing if you gained any memory performance from a change in your setup. You can use the trial version of Geekbench 3 and just go into the Geekbench folder after it is installed and you will see two files that I have included in a screenshot below. Swap the name of the file with the other. So the geekbench_x86_32.exe will change to geekbench_x86_64.exe and vice versa. Then just open Geekbench 3 and run the 32bit test and it will run the 64bit test since you changed the file name earlier. Then just use the memory score to compare to other setups. That and firestrike Physics test are good choices with Geekbench being the better choice imo.
Attached Images
 

CPU- AMD Ryzen [email protected] *
MOTHERBOARD- ROG Crosshair VII Hero
GPU- EVGA GTX 1080ti SC2 w/ XOC [email protected] (Morpheus Cooler)
RAM- G.SKILL TridentZ 32gb (4 x 8gb) @ 3800mhz (14-15-15-24-38-1T)
COOLING- CORSAIR Hydro Series H115i Pro w/ Noctua NF-A14 (x7)
CASE- Silverstone RL06 Pro (Modified)
MONITORS- VIOTEK GN34DB (x2)


CJMitsuki is offline  
post #7 of 7 (permalink) Old 01-31-2019, 01:36 AM - Thread Starter
New to Overclock.net
 
rul3s's Avatar
 
Join Date: Dec 2014
Location: Spain
Posts: 120
Rep: 1 (Unique: 1)
Quote: Originally Posted by CJMitsuki View Post
What he meant was the secondary and tertiary timings but you have them set manually so they arent changing but that isnt to say that the timings are optimal for a higher frequency. I dont rely on Realbench to show true memory performance. Geekbench 3 is probably the best at showing if you gained any memory performance from a change in your setup. You can use the trial version of Geekbench 3 and just go into the Geekbench folder after it is installed and you will see two files that I have included in a screenshot below. Swap the name of the file with the other. So the geekbench_x86_32.exe will change to geekbench_x86_64.exe and vice versa. Then just open Geekbench 3 and run the 32bit test and it will run the 64bit test since you changed the file name earlier. Then just use the memory score to compare to other setups. That and firestrike Physics test are good choices with Geekbench being the better choice imo.
Ok, I'm trying geekbench for that so, but, what worries me is that in realbench scores goes down, I could understand if scores where almos the same going up from 3200, but going down I what keeps me frustrated...

I'll report geekbench scores later

PhanteksR7
(16 items)
CPU
Ryzen R7 3700x
Motherboard
Asus x470 Strix F
GPU
RTX 2070 Windforce OC
RAM
KFA2 HOF Extreme DDR4 4000CL19 BDie
Hard Drive
Samsung 970EvoPlus 500GB NVMe
Hard Drive
Samsung 850Pro 512GB
Hard Drive
Samsung 860QVO 2TB
Power Supply
Seasonic Focus+ 650W Platinium
Cooling
Noctua NH-U12P
Case
Phanteks Enthoo Pro M SE
Operating System
Windows 10 PRO
Monitor
Samsung C24FG70
Keyboard
Corsair K70 RGB
Mouse
Zowie FK2
Mousepad
Razer Goliathus Cosmic Speed
Audio
Asus Xonar STX
▲ hide details ▲
rul3s is offline  
Reply

Tags
3200 , divisor , Performance , ram speed

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: 1 (0 members and 1 guests)
 
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