Overclock.net › Forums › AMD › AMD CPUs › [OFFICIAL] FX-8320/FX-8350 Vishera Owners Club
New Posts  All Forums:Forum Nav:

[OFFICIAL] FX-8320/FX-8350 Vishera Owners Club - Page 519

post #5181 of 67342
Quote:
Originally Posted by shaxs View Post

Thanks for taking the time to reply!
I am running 4x4gb of ram. Before F9 bios, my computer defaulted them to 1333. With F9 it defaulted to 1600. I tried lowering the speed back to 1333 but that didnt solve the issues. I now have my 1090t BACK in and I am running at 1600 stable with an overclock of 3.9ghz. No issues.
I did boot into Memtest and ran with no issues.

How long did you let the memtest run ?
Quote:
The lab guys over at AMD got a hold of two 990 chipset boards. They tested the first one and said no issues. They are going to test on a Gigabyte and let me know the results.
Very odd how we can have vastly different experience. 1090t overclocked = stable. FX-8350 STOCK = sooo unstable.
BTW, what are your temperatures with your Noctua?

See my post at http://www.overclock.net/t/1318995/official-fx-8320-fx-8350-vishera-owners-club/5150#post_18776526, there is an image link that has my temps at 4.6 GHz , 17 hours of prime95.
That run was too hot, but it ran fine.
post #5182 of 67342
Quote:
Originally Posted by Red1776 View Post


Did you try the 'mem=ok' feature? and another set of ram?

Looks like my older PSU killed both the 990FXA-UD3 AND the Corsair XMS3 2000MHz..mad.gif All is well with the HyperX Predator 2400MHz (only running 2133 for now).

Not really.. 2248MHz..
Edited by kzone75 - 12/8/12 at 12:53am
Oxygène
(15 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 5 1600 MSI B350 Tomahawk Arctic VTX3D R9 390 8GB G.Skill Ripjaws  
Hard DriveHard DriveHard DriveHard Drive
Kingston Hyper X Kingston HyperX Fury ADATA SP550 ADATA SX8000 
CoolingOSMonitorKeyboard
Stock Windows 10 Dell U2711 Microsoft WK 800 
PowerCaseMouse
Chieftec A-135 Series 1000W NZXT. Phantom 410 Microsoft WM 1000 
  hide details  
Reply
Oxygène
(15 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen 5 1600 MSI B350 Tomahawk Arctic VTX3D R9 390 8GB G.Skill Ripjaws  
Hard DriveHard DriveHard DriveHard Drive
Kingston Hyper X Kingston HyperX Fury ADATA SP550 ADATA SX8000 
CoolingOSMonitorKeyboard
Stock Windows 10 Dell U2711 Microsoft WK 800 
PowerCaseMouse
Chieftec A-135 Series 1000W NZXT. Phantom 410 Microsoft WM 1000 
  hide details  
Reply
post #5183 of 67342
Quote:
Originally Posted by gertruude View Post

Id like to see a screenshot of u running 5ghz man with 20 ibt tests.

Well, I have kept a record of all the BIOS setting changes that I have done for every test in a spreadsheet.
As well as any hardware changes.

I have not run 20 IBT, just 10 times.

I have the full BIOS settings of the circumstance last week.

CPU clock ratio x19 4940
Northbridge freq x8 2080
CPB off
CPU host control auto
CPU frequency 260
PCIE clock 100

HT link width auto
HT link frequency x10 2600
DRAM EOCP Disabled
set memory clock Manual
memory clock X5.33 1385

voltage Manual
CPU PLL normal 2.5V
DRAM voltage normal 1.5V
DDR VTT normal 0.75V
NB voltage normal 1.10V
HT Link voltage control normal 1.2V
NB/PCIe/PLL Voltage Control normal 1.8V
CPU NB VID normal
CPU voltage +0.1V 1.4375V

Load Line Calibration auto

BIOS boot OK
Windows boot OK
IBT OK !

max temp 58°C

Prime 95 instant BSOD

I reset everything the same now.
But can't reproduce, IBT fails right now with above settings at 4940. And the temp goes higher than 58C.

The difference I believe is that I only had 2 sticks of 4GB RAM , Corsair DDR3-1600, last week.

On saturday, I switched over to 4 sticks of 8 GB DDR3-1866 , G;Skill.

With the 4 sticks , the mobo & CPU both run significantly hotter.

With the 4 sticks, 2 of them are under one of the Noctua CPU cooler fans.
When I had only 2 sticks, I had them both on the outside DIMM slots, ie. on separate channels...

My Corsair sticks are back in the Phenom II box on which I'm typing this. I switch back & forth with a keystroke on my KVM to test and record everything .
I don't feel like removing the sticks just to give you a screenshot.

Anyway, looking at my spreadsheet of all those records (now over 3000 lines long) I see that the highest clock that IBT passed with the 4 sticks of RAM in was 4776 MHz . That was 20.5 x 233 .
The max temp was 63C. Prime95 didn't BSOD for that test, but just stopped after 2 minutes. An occurrence with which I'm sure you are very familiar.
post #5184 of 67342
Quote:
Originally Posted by madbrain View Post

Well, I have kept a record of all the BIOS setting changes that I have done for every test in a spreadsheet.
As well as any hardware changes.
I have not run 20 IBT, just 10 times.

Anyway, looking at my spreadsheet of all those records (now over 3000 lines long) I see that the highest clock that IBT passed with the 4 sticks of RAM in was 4776 MHz . That was 20.5 x 233 .
The max temp was 63C. Prime95 didn't BSOD for that test, but just stopped after 2 minutes. An occurrence with which I'm sure you are very familiar.

I was only pulling your leg having a bit of a wind up over asking for a screenshot lol thanks for doing it though biggrin.gif
You did say you ran 20 times IBT @5ghz though in your previous post.

And yes over 4.6 i am having trouble with prime. I dont bsod or anything i just get the same core error I think most people do have problems with it after a certain point.
My Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
FX-8370 asus sabertooth 990fx R9 290 DC2 kingston hyper x beast 1866 
CoolingOSMonitorKeyboard
fractal design kelvin s36 windows 10 pro Acer G245HQ Saitek Cyborg V5 
PowerCaseMouse
EVGA SuperNOVA 1000 G2 Corsair Graphite series 780T corsair m90 
  hide details  
Reply
My Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
FX-8370 asus sabertooth 990fx R9 290 DC2 kingston hyper x beast 1866 
CoolingOSMonitorKeyboard
fractal design kelvin s36 windows 10 pro Acer G245HQ Saitek Cyborg V5 
PowerCaseMouse
EVGA SuperNOVA 1000 G2 Corsair Graphite series 780T corsair m90 
  hide details  
Reply
post #5185 of 67342
Hey guys and gals,

I'm wondering if any of you has tried OCing with AMD application power management ON. Though this seemed silly at first, I thought of the fact that in daily use the CPU seldom goes over 45C on the cores and 55C on the socket, even if stress tests take it to the upper limit (60C cores and 71C socket) and since my cooling is far from being the best, I thought I'd give it a try.

I'm now running 4.5 GHz with 1.425v which would normally take my temps way above 60C core / 71C socket mentioned above, so I can't stress test it without pushing over those temps which I'd like to stay under. But with AMD APM set to on, the CPU throttles down on several cores at a time (to 3.4GHz) in order to keep the TDP down (i guess) and to lower temps. After 50 mins of Prime temps have gone as high as 54C on the cores and 66C on the socket.

Judging by the AIDA logs, average CPU speed is ~4.2 GHz so that would mean that on average 6 cores are at 4.5GHz and 2 are at 3.4GHz at all times, cycling which cores are throttled down of course. This may vary and you get all cores at 4.5 one time then 6 cores at 3.4 after a while, but I'm talking average here.

During normal use the CPU doesn't throttle down as there's no need for it, so performance would not suffer (or only do so mildly) but there are two upsides that I can see:
- You have a sort of "fail-safe" in case you're doing something more CPU intensive, you don't have to keep an eye on the temps all the time, you know the APM will do it for you;
- You can stress test at higher speeds / voltages without having to worry about the temps;

Not really sure about how valid stress testing with the throttling down really is ... I'm looking for some input on this. We can only guess I presume, but considering that I'm stressing it at the exact same settings (APM enabled) that day to day use would be, if it can hold plenty of hours in a stress test with no issue, shouldn't one be able to call that "stable" ?

Any constructive input is welcome :-)
RyZen
(13 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen r7 1700 ASUS PRIME X370-PRO GTX 1070 Founders Edition 16GB GSkill RipV K2 3200MHz C16  
Hard DriveHard DriveCoolingOS
SanDisk X210 Solid State Drive WesternDigital Caviar Green 1TB Corsair H110i Windows 10 Pro 
MonitorKeyboardPowerCase
Acer XF270HU 144Hz IPS Microsoft Sculpt Comfort Keyboard Seasonic M12II-620 EVO Edition Bronze 620W Phanteks Eclipse P400 
Mouse
Microsoft Sculpt Comfort Mouse 
  hide details  
Reply
RyZen
(13 items)
 
  
CPUMotherboardGraphicsRAM
Ryzen r7 1700 ASUS PRIME X370-PRO GTX 1070 Founders Edition 16GB GSkill RipV K2 3200MHz C16  
Hard DriveHard DriveCoolingOS
SanDisk X210 Solid State Drive WesternDigital Caviar Green 1TB Corsair H110i Windows 10 Pro 
MonitorKeyboardPowerCase
Acer XF270HU 144Hz IPS Microsoft Sculpt Comfort Keyboard Seasonic M12II-620 EVO Edition Bronze 620W Phanteks Eclipse P400 
Mouse
Microsoft Sculpt Comfort Mouse 
  hide details  
Reply
post #5186 of 67342
Quote:
Originally Posted by gertruude View Post

I was only pulling your leg having a bit of a wind up over asking for a screenshot lol thanks for doing it though biggrin.gif
You did say you ran 20 times IBT @5ghz though in your previous post.
And yes over 4.6 i am having trouble with prime. I dont bsod or anything i just get the same core error I think most people do have problems with it after a certain point.

FYI .. Here is your screenshot .. Passed IBT at 5.0 GHz , 25 x 200 . With all 4 DIMMs . Lost 6 cores immediately in Prime 95 ...

https://doc-0c-c4-docs.googleusercontent.com/docs/securesc/lehnlc1vdlgktvn81jinll5sttj4aof8/l9lam45rb4ei43l8b600h31lfn41dsu1/1354968000000/14193508639955794529/14193508639955794529/0B1eSSO_7gwqeQmc3UHh2UHo1Z0U?e=download

Even though IBT reports it "passed", the result displayed is quite suspicious ! It's a NaN (not a number)...
But hey, at least it's the same suspicious looking number for all 10 passes, so who am I to argue ...

Edit: I ran it again with 20 passes, same result - IBT reports pass, and all 20 iterations have the same weird NaN number. Too lazy to take yet another screenshot.
Oh, and between the 19th and 20th iteration, the CPU temp hit 75C; and the Windows desktop process crashed and restarted iteslf. But no matter, IBT kept going. This does tell me the machine was definitely unstable. It wasn't just Prime95. But all OK according to IBT. I don't know how you guys can trust it alone.
Edited by madbrain - 12/8/12 at 4:47am
post #5187 of 67342
Quote:
Originally Posted by madbrain View Post

FYI .. Here is your screenshot .. Passed IBT at 5.0 GHz , 25 x 200 . With all 4 DIMMs . Lost 6 cores immediately in Prime 95 ...
https://doc-0c-c4-docs.googleusercontent.com/docs/securesc/lehnlc1vdlgktvn81jinll5sttj4aof8/l9lam45rb4ei43l8b600h31lfn41dsu1/1354968000000/14193508639955794529/14193508639955794529/0B1eSSO_7gwqeQmc3UHh2UHo1Z0U?e=download
Even though IBT reports it "passed", the result displayed is quite suspicious ! It's a NaN (not a number)...
But hey, at least it's the same suspicious looking number for all 10 passes, so who am I to argue ...
Edit: I ran it again with 20 passes, same result - IBT reports pass, and all 20 iterations have the same weird NaN number. Too lazy to take yet another screenshot.

Cool man. BTW the link dont work for me. Dont take my screenshot remark seriously man i was only pulling your leg a bit. rolleyes.gif

Im doing a 4.8ghz prime now and gone further than it usually does, i only raised cpu/nb volts a tiny bit so fingers crossed its gonna pass through the hour mark

Edit BTW what ya doing up so late? or early lol

I dont trust IBT alone if im stable with IBT and OCCT i figure its alright. Though i will be 100% happy if prime passes also
Edited by gertruude - 12/8/12 at 4:53am
My Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
FX-8370 asus sabertooth 990fx R9 290 DC2 kingston hyper x beast 1866 
CoolingOSMonitorKeyboard
fractal design kelvin s36 windows 10 pro Acer G245HQ Saitek Cyborg V5 
PowerCaseMouse
EVGA SuperNOVA 1000 G2 Corsair Graphite series 780T corsair m90 
  hide details  
Reply
My Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
FX-8370 asus sabertooth 990fx R9 290 DC2 kingston hyper x beast 1866 
CoolingOSMonitorKeyboard
fractal design kelvin s36 windows 10 pro Acer G245HQ Saitek Cyborg V5 
PowerCaseMouse
EVGA SuperNOVA 1000 G2 Corsair Graphite series 780T corsair m90 
  hide details  
Reply
post #5188 of 67342
Quote:
Originally Posted by gertruude View Post

Cool man. BTW the link dont work for me.

Hmm, that's odd, do you get any error ? It should let you download a PNG from my google drive ... I shared it.
Quote:
Im doing a 4.8ghz prime now and gone further than it usually does, i only raised cpu/nb volts a tiny bit so fingers crossed its gonna pass through the hour mark
Quote:
Edit BTW what ya doing up so late? or early lol

Still messing with the damn machine, that's what ...

I am very bothered by the huge variability in the temps. In my 17 hr run of prime 95 yesterday at 4.6 Ghz, the full load temp started around 40°C . Then went all the way to 65°C at some point during the night or day. That is just not right.

I have been trying to lower the voltage one notch in the hope of getting the temp to say below 60°C. It's a no go.
I have reduced vcore from 1.41 to 1.38, and upped every other voltage setting - but it's a no go at 4.6 Ghz, 1.41 seems to be the minimum voltage...
post #5189 of 67342
Quote:
Originally Posted by madbrain View Post

Hmm, that's odd, do you get any error ? It should let you download a PNG from my google drive ... I shared it.
Still messing with the damn machine, that's what ...
I am very bothered by the huge variability in the temps. In my 17 hr run of prime 95 yesterday at 4.6 Ghz, the full load temp started around 40°C . Then went all the way to 65°C at some point during the night or day. That is just not right.
I have been trying to lower the voltage one notch in the hope of getting the temp to say below 60°C. It's a no go.
I have reduced vcore from 1.41 to 1.38, and upped every other voltage setting - but it's a no go at 4.6 Ghz, 1.41 seems to be the minimum voltage...

That's exactly what ive been doing. Im just starting to think no matter how hard i try im not going to be stable with prime above 4.6. Which is disheartening, it would be icing on the cake. Though im the kinda guy to say NO Retreat NO Surrender! haha

the error for your link:
The page isn't redirecting properly
Firefox has detected that the server is redirecting the request for this address in a way that will never complete.
This problem can sometimes be caused by disabling or refusing to accept
cookies.

I cleared cookies also no change. Its no biggy though man like i said i was only having a laugh with ya smile.gif
My Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
FX-8370 asus sabertooth 990fx R9 290 DC2 kingston hyper x beast 1866 
CoolingOSMonitorKeyboard
fractal design kelvin s36 windows 10 pro Acer G245HQ Saitek Cyborg V5 
PowerCaseMouse
EVGA SuperNOVA 1000 G2 Corsair Graphite series 780T corsair m90 
  hide details  
Reply
My Rig
(11 items)
 
  
CPUMotherboardGraphicsRAM
FX-8370 asus sabertooth 990fx R9 290 DC2 kingston hyper x beast 1866 
CoolingOSMonitorKeyboard
fractal design kelvin s36 windows 10 pro Acer G245HQ Saitek Cyborg V5 
PowerCaseMouse
EVGA SuperNOVA 1000 G2 Corsair Graphite series 780T corsair m90 
  hide details  
Reply
post #5190 of 67342
For those that are having trouble getting stability either:
- on a module above a certain frequency
- on a module, but a rediculous amount of voltage helps and the temps are crazy high

I'd recommend using an overclock program to stagger the cores, you may get a higher overclock on average.

Example:
Cores 1-2 @ 5.3Ghz
Cores 3-4 @ 5.0Ghz
Cores 5-6 @ 4.7Ghz
Cores 7-8 @ 4.5Ghz
@ 1.475v - 4.875Ghz average

VS.

Full CPU Overclock of 4.6Ghz @ 1.5v
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
Overclock.net › Forums › AMD › AMD CPUs › [OFFICIAL] FX-8320/FX-8350 Vishera Owners Club