Strictly technical: Matisse (Not really) - Page 68 - Overclock.net - An Overclocking Community
Forum Jump: 

Strictly technical: Matisse (Not really)

Closed Thread
 
Thread Tools
post #671 of 679 (permalink) Old 08-22-2019, 01:38 AM
Meddling user
 
gupsterg's Avatar
 
Join Date: Jan 2015
Location: Lurking over a keyboard
Posts: 7,423
Rep: 786 (Unique: 360)
Quote: Originally Posted by mongoled View Post
Thanks for that,

dont understand the reasons why der8auer and others are still actively promoting the use of 1344k while testing the stability of Ryzen CPU that are going to be sold into the retail chain.

Why would Casekings.de and others ignore the data and just peddle old truths ??
The thing is Silicon Lottery/Casekings.de/8 [email protected] and even any of us, all have our own differing methods. So I don't think Casekings.de and others is promoting old truths.

As I don't do manually set ratio OC on my R5 3600 I don't do the FFT test The Stilt stated. I run the version of P95 as he stated and disabled AVX/AVX2/FMA3 via local.txt, but I use 8K 4096K 27000MB to test:-

i) Stock CPU with RAM OC.
ii) PBO+xxxMHz with RAM OC.
gupsterg is offline  
Sponsored Links
Advertisement
 
post #672 of 679 (permalink) Old 08-22-2019, 08:26 AM
New to Overclock.net
 
mongoled's Avatar
 
Join Date: Oct 2011
Location: Ayia Napa, Cyprus
Posts: 690
Rep: 24 (Unique: 19)
Quote: Originally Posted by gupsterg View Post
The thing is Silicon Lottery/Casekings.de/8 [email protected] and even any of us, all have our own differing methods. So I don't think Casekings.de and others is promoting old truths.

As I don't do manually set ratio OC on my R5 3600 I don't do the FFT test The Stilt stated. I run the version of P95 as he stated and disabled AVX/AVX2/FMA3 via local.txt, but I use 8K 4096K 27000MB to test:-

i) Stock CPU with RAM OC.
ii) PBO+xxxMHz with RAM OC.

Yes, but us as individuals is completely different to when selling in retail channel.

Anyhow, its no biggie, just found it strange that they test with settings which also stress the memory subsystem..............

I generally do what you do with regards to prime95 settings, however for quick and dirty tests I wanted to use the fastest means of finding instability with the CPU and custom small FFTs achieves this.


SeaSonic X750 Gold 750W
MSI X370 XPOWER Gaming Titanium | Ryzen 5 3600 2010SUS @ 4400 mhz @ 1.325v (prime95 load)
Viper Steel Series PVS416G413C9K @ 3800/1900 mhz | 1.52v | 14-15-14-14-28-42-252-16-1T
Asus GTX 780 DCII OC @ 1332/3522 mhz | 1.3v | Samsung 960 EVO 500GB
EK-FB MSI X370 XPower RGB Monoblock 1/2" ID | Swiftech D5 MCP655 | EK-XRES 100
Alphacool NexXxoS UT60 280 | 4 x Scythe GT AP-15 @ 5v
Thermochill PA120.3 | 1 x NF-A12x15 | 2 x ML120 | PWM
mongoled is offline  
post #673 of 679 (permalink) Old 08-22-2019, 09:11 AM
Meddling user
 
gupsterg's Avatar
 
Join Date: Jan 2015
Location: Lurking over a keyboard
Posts: 7,423
Rep: 786 (Unique: 360)
Quote: Originally Posted by mongoled View Post
Yes, but us as individuals is completely different to when selling in retail channel.

Anyhow, its no biggie, just found it strange that they test with settings which also stress the memory subsystem..............

I generally do what you do with regards to prime95 settings, however for quick and dirty tests I wanted to use the fastest means of finding instability with the CPU and custom small FFTs achieves this.

Even the differing entities selling via retail channels use differing methods . What Silicon Lottery do, Caseking.de don't, what OCuk (which AFAIK is owned by Caseking) don't do as their resident overclocker is 8 Pack compared with der8auer over in Germany.

I have used small custom FFT on other Ryzen systems, just not with this one so far.
gupsterg is offline  
Sponsored Links
Advertisement
 
post #674 of 679 (permalink) Old 08-22-2019, 09:24 AM
New to Overclock.net
 
lordzed83's Avatar
 
Join Date: Mar 2013
Location: Nottingham
Posts: 1,628
Rep: 41 (Unique: 31)
Quote: Originally Posted by gupsterg View Post
The thing is Silicon Lottery/Casekings.de/8 [email protected] and even any of us, all have our own differing methods. So I don't think Casekings.de and others is promoting old truths.

As I don't do manually set ratio OC on my R5 3600 I don't do the FFT test The Stilt stated. I run the version of P95 as he stated and disabled AVX/AVX2/FMA3 via local.txt, but I use 8K 4096K 27000MB to test:-

i) Stock CPU with RAM OC.
ii) PBO+xxxMHz with RAM OC.
That's spot on. Everyone I jknow tests different. Like ME its ycruncher PI pass 10xveryhigh AVX IBT pass aida 1 hour stress test pass and nicehash 10+hours pass and got 3 hour long rendering project in power director for test. I cant remember when was last time I used prime.OFC HCI and ramtest pass. Thats what I consider stable for my use.
lordzed83 is offline  
post #675 of 679 (permalink) Old 08-22-2019, 12:52 PM
New to Overclock.net
 
Join Date: Aug 2019
Posts: 61
Rep: 1 (Unique: 1)
Anyone done much with BCLK? Here's what I'm getting:
Attached Thumbnails
Click image for larger version

Name:	4701_stress.jpg
Views:	369
Size:	703.4 KB
ID:	290840  

ver_21 is offline  
post #676 of 679 (permalink) Old 08-22-2019, 06:08 PM
New to Overclock.net
 
srg3037's Avatar
 
Join Date: Mar 2013
Posts: 11
Rep: 0
I have messed with BCLK on my 3900X as well. I do get the occasional WHEA correctable Code 17 error but I cant tell if that is a harmless error (Gigabyte has not released the updated BIOS for the x470 Gaming Wifi 7 yet that fixes the WHEA errors). Currently running 102.75 but I can go to 103.5 but the WHEA errors increase. Other than the WHEA errors everything runs stable and the benchmark score increase with the clock speed.
srg3037 is offline  
post #677 of 679 (permalink) Old 08-22-2019, 11:19 PM
Gaming enthusiast
 
Ironcobra's Avatar
 
Join Date: Apr 2011
Posts: 827
Rep: 36 (Unique: 30)
Quote: Originally Posted by polkfan View Post
Thank you finally someone who also puts this to rest!

I have been saying this over and over again lol.

1.5V is 100% safe as long as your temps are good, under load you should see 1.33V or so
I believe the higher idle voltages we are seeing near idle conditions makes it easier for the chip to respond faster when needed as higher voltage/potential makes for a faster response when current is needed, this chip is changing states so much it really makes sense for a higher potential to be there ready for action. Some people are really freaking out over this but if you watch current and watts during these idle conditions there very low and if theres no change in temps I think we might learn this is normal moving forward. My chip definitely drops to the low 1.3s under load. And in the electrical world higher voltage=lower current=more efficiency and less heat/resistance. I mean I could be wrong but thats basic ac/dc theory.

Origin: Ironcobra80
Steam: Ironcobra
Gaming Ryzen
(14 items)
CPU
3600
Motherboard
Gigabyte Auros Master x570
GPU
EVGA XC ULTRA 2080 TI
RAM
G.SKILL TridentZ RGB F4-3600C16D-16GTZR (16-15-15-28 3800)
Hard Drive
HP EX950 1TB
Hard Drive
SAMSUNG EVO 860 1TB
Power Supply
SEASONIC FOCUS+ GOLD 650W
Cooling
CORSAIR 115I PRO PUSH/PULL FANS-NOCTUA
Case
CORSAIR CARBIDE AIR 540
Operating System
W10 FAST RING
Monitor
LG OLED C6 65"
Keyboard
Varmilo VB87M
Mouse
LOGITECH G900
Mousepad
Zowie Gear P-SR
▲ hide details ▲

Last edited by Ironcobra; 08-22-2019 at 11:22 PM.
Ironcobra is offline  
post #678 of 679 (permalink) Old 08-23-2019, 12:21 AM
Meddling user
 
gupsterg's Avatar
 
Join Date: Jan 2015
Location: Lurking over a keyboard
Posts: 7,423
Rep: 786 (Unique: 360)
Quote: Originally Posted by Ironcobra View Post
I believe the higher idle voltages we are seeing near idle conditions makes it easier for the chip to respond faster when needed as higher voltage/potential makes for a faster response when current is needed, this chip is changing states so much it really makes sense for a higher potential to be there ready for action. Some people are really freaking out over this but if you watch current and watts during these idle conditions there very low and if theres no change in temps I think we might learn this is normal moving forward. My chip definitely drops to the low 1.3s under load. And in the electrical world higher voltage=lower current=more efficiency and less heat/resistance. I mean I could be wrong but thats basic ac/dc theory.
Doubt the higher vcore is to allow potential for improved CPU state change.

Some of my observations.

I had higher idle vcore on AGESA ComboPi-AM4 1.0.0.3AB when Ryzen Master is open, ~1.35V. I enabled Global C-State Control in UEFI and it then showed monitoring data as when using a UEFI with ComboPi-AM4 1.0.0.2, ~0.2V idle. But HWINFO would only show 3600MHz ~1V at idle on AGESA ComboPi-AM4 1.0.0.3AB, version 6.11.3890 in use at the time, UEFI with ComboPi-AM4 1.0.0.2 it would show 2200MHz ~0.9V.

Now on newer beta of HWINFO, even on AGESA ComboPi-AM4 1.0.0.3AB or AGESA ComboPi-AM4 1.0.0.3ABB, I see 2200MHz ~0.9V idle. Ryzen Master is showing ~0.2V idle even without enabling Global C-State Control on UEFI with AEGSA ComboPi-AM4 1.0.0.3ABB.

I think the increased sensitivity of boost mechanism, plus teething issues of FW/SW are cause for some of the oddities we're observing at present.
gupsterg is offline  
post #679 of 679 (permalink) Old 08-23-2019, 06:27 AM
New to Overclock.net
 
Join Date: Aug 2019
Posts: 61
Rep: 1 (Unique: 1)
Quote: Originally Posted by srg3037 View Post
I have messed with BCLK on my 3900X as well. I do get the occasional WHEA correctable Code 17 error but I cant tell if that is a harmless error (Gigabyte has not released the updated BIOS for the x470 Gaming Wifi 7 yet that fixes the WHEA errors). Currently running 102.75 but I can go to 103.5 but the WHEA errors increase. Other than the WHEA errors everything runs stable and the benchmark score increase with the clock speed.

I like that Master board. I think I saw last night that the latest BIOS might correct some/all of the WHEA for GB x570s? Not sure what version it was supposed to be.


I'm running on an MSI Ace, BIOS version 120 (releases are at 144 beta, atm). Frustrations with Boost Override made me determined to try BCLKing. I think it presents some additional issues with RAM and PCIe 4.0--it's hard to be sure, but I suspect not all GPUs tolerate BLCK OCing. And I don't think I can use such tight timings with RAM. We'll see, it's a work in progress, but it seems stable (so maybe I should leave well-enough alone).


I am wondering if HWiNFO is direct-measuring or calculating core frequencies achieved by BCLK. Ryzen Master does not seem to record frequencies correctly once BCLK is used.
ver_21 is offline  
Closed Thread

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