Overclock.net - An Overclocking Community - Reply to Topic

Thread: Ryzen 3000 RAM Scaling strange behaviour [Lower scores] Reply to Thread
Title:
Message:

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


  Additional Options
Miscellaneous Options

  Topic Review (Newest First)
08-06-2019 02:33 AM
rul3s Finally I didn't do the tests in normal speed (3.6GHz) because it wouldn't be real, as my 3700x with PBO always goes up tu 41-41.5x, I've ensured stability on this fixed speed, without C-States, and then made the tests.

The results:


My custom config for 3666 is this:
08-02-2019 06:16 AM
rdr09
Quote: Originally Posted by rul3s View Post
The highest I can go with CL14-14-14-14-28-42 is 3333, thats my reference point.

Now I'm testing [email protected], with other subtimmings from another user, it's going very well, i'll post results
Can't wait. I suspect really you have to go that high to actually see any difference. The game cache probably has something to do with it.
08-02-2019 02:17 AM
rul3s
Quote: Originally Posted by rdr09 View Post
No 3466 CL14 setting? In Gen 1 and +, 3466 CL14 is the sweetest spot.
The highest I can go with CL14-14-14-14-28-42 is 3333, thats my reference point.

Now I'm testing [email protected], with other subtimmings from another user, it's going very well, i'll post results
08-02-2019 01:19 AM
rdr09
Quote: Originally Posted by rul3s View Post
Good advice, I'll try it and post results
No 3466 CL14 setting? In Gen 1 and +, 3466 CL14 is the sweetest spot.
08-02-2019 01:02 AM
rul3s
Quote: Originally Posted by almstsobur View Post
I would dial down the CPU, shoot for low temps, base speed. I know this CPU will throttle and not show a frequency drop when doing a negative Vcore. I'm not saying that is for sure the issue, but best to eliminate that variable. Verify that you get the results you expect and go from there. Also, try different interleaving sizes with each speed, it can be a detriment for higher speed higher latency ram to be using a tiny interleave size.
Good advice, I'll try it and post results
08-01-2019 11:04 AM
almstsobur I would dial down the CPU, shoot for low temps, base speed. I know this CPU will throttle and not show a frequency drop when doing a negative Vcore. I'm not saying that is for sure the issue, but best to eliminate that variable. Verify that you get the results you expect and go from there. Also, try different interleaving sizes with each speed, it can be a detriment for higher speed higher latency ram to be using a tiny interleave size.
08-01-2019 01:00 AM
rul3s
Ryzen 3000 RAM Scaling strange behaviour [Lower scores]

Hello guys,
I'm triing to find the sweet spof of my config about RAM, but I'm going crazy, my hardware is:
- Ryzen 3700x
- Asus x470 Strix F bios 5007 (agesa 1.0.0.2)
- KFA2 HOF DDR4 4000CL19 (B-Die)
To isolate RAM performance I've setted up fixed OC to 4.1GHz, windows 1903 and drivers 1.07.07 and always set to 1:1 FCLK/RAM.

Following the Ryzen Dram Calculator I've validated 3 configs, all them stable (Testmem5 + 1000%HCI):
- [email protected] (Fast Presset) 1.365vRam 1.025vSoc
- [email protected] (Fast Presset) 1.395vRam 1.1vSoc
- [email protected] (Safe Presset) 1.450vRam 1.1vSoc

Those are my results:



As you can see, I see little gains on Cinebench, more or less the same on 3DMark CPU TimeSpy (each run goes 100-200up/down) and losing perf on some tests of Realbench...

My first concert was about stability, but as far as I know, with 1000% HCI + TestMem5 1smus_v3 config is RAM stable, and about CPU is also stable for 30min of prime + 2h realbench stress test.

How can I understand this behaviour? Shouldn't be 3600 config outperforming all the rest?
[email protected] config is exactly the same timmings as 3200, but with more voltage, how can this be giving lower results on some test vs 3200 config?

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