Stock 3950x crashes with prime95 small FFTs - Page 9 - Overclock.net - An Overclocking Community
Forum Jump: 

Stock 3950x crashes with prime95 small FFTs

Reply
 
Thread Tools
post #81 of 100 (permalink) Old 06-03-2020, 06:41 AM
New to Overclock.net
 
smonkie's Avatar
 
Join Date: Sep 2013
Posts: 181
Rep: 4 (Unique: 4)
Quote: Originally Posted by rdr09 View Post
Do you mind showing a screenshot of HWINFO64 while running Prime95? Preferably all the info on the cpu and ram.
Quote: Originally Posted by Steven Stacy View Post
Has to be a reason why PBO does a 275Mhz drop from normal load for stressing the cache. The voltage is much lower and the temps is only 71c
I should be able to manually OC the lowest value above and work my way up and see when it starts failing.
I just notice it would be an instant 20deg jump between the tests.
What is your temps if you do a large FFT test at your current settings? Does it go above 80c?
I've just installed latest Aorus Master x570 BIOS beta version just posted yesterday (F12g), and without even entering BIOS after the flash, I've been running Prime95 for an hour now. Fingers crossed... Here's a screenshot:



https://imgur.com/3yATRb9

Edit: it failed in one worker at 1 hour 21 minutes mark, a few seconds after starting to watch a Youtube video. I wonder if there could be any relation.

Last edited by smonkie; 06-03-2020 at 06:54 AM.
smonkie is offline  
Sponsored Links
Advertisement
 
post #82 of 100 (permalink) Old 06-03-2020, 06:59 AM
New to Overclock.net
 
Steven Stacy's Avatar
 
Join Date: Feb 2020
Posts: 88
Rep: 3 (Unique: 1)
Does youtube use AVX lol.
I just ran the sfft without AVX and frequency increase 100Mhz to 4100Mhz.
I also notice CB20 always dropping 100Mhz on all cores to 4175 compared to other loads that are 4275 on all cores.
CB20 is always 100Mhz lower if I make changes. So If I topping out at 4250... CB20 is 4150.
I am starting to suspect a 100Mhz avx offset is part of the PBO logarithm.

Sure wish AMD had an avx offset for manual OC.

CPU
Ryzen 3950X
Motherboard
Aorus Wifi Elite
GPU
Gigabyte AORUS GeForce RTX 2080 Xtreme WATERFORCE WB
RAM
G.Skill Ripjaws 32GB (2 x 16GB) DDR4-3600 PC4-28800 CL16
Hard Drive
Gigabyte AORUS 1TB SSD 3D TLC NAND M.2 2280 PCIe NVMe 4
Power Supply
Seasonic USA Focus Plus 750 Watt 80 Plus Gold ATX
Cooling
Corsair Hydro XC7 RGB CPU Water Block
Cooling
Distrub Plate w/pump RGV-LAN-011-P
Cooling
Corsair XR5 360
Cooling
Corsair XR5 280
Cooling
FAN Artic Cool P14 X 4
Cooling
FANS Corsair LL120 x 3
Cooling
FANS LL140 X 2
Cooling
Corsair Commander Pro
Cooling
Phantek 8 port pwm controller
Case
Lian Li PC-O11 Dynamic
Monitor
Dell UltraSharp 49 Curved Monitor: U4919DW
Keyboard
Mastercool sk650
Mouse
corsair dark core rgb
Other
Lian LI diffused rgb strip
Other
Corsair LS100 diffused rgb strips pack of 2
▲ hide details ▲
Steven Stacy is offline  
post #83 of 100 (permalink) Old 06-03-2020, 04:18 PM
New to Overclock.net
 
DemonAk's Avatar
 
Join Date: Mar 2014
Posts: 28
Rep: 0
the interesting thing is, if you manually set PBO limits 142W 95A 95A then the prime95 seems to become stable, but the frequency drops to 2700Mhz (small fft) and 3200Mhz (smallest fft). CB20 Score 8762. The default settings (PBO disable) are 142W 95A 140A
if enable eco mode also seems stable =) limits: 87W 60A 90A

Ryzen 3950x (batch 2008), ASRock X370 Taichi, 4x32gb [email protected], Palit GTX 1080ti JetStream, Windows 10 x64 Pro

Last edited by DemonAk; 06-03-2020 at 04:23 PM.
DemonAk is offline  
Sponsored Links
Advertisement
 
post #84 of 100 (permalink) Old 06-03-2020, 04:59 PM
New to Overclock.net
 
Steven Stacy's Avatar
 
Join Date: Feb 2020
Posts: 88
Rep: 3 (Unique: 1)
Quote: Originally Posted by smonkie View Post
I've just installed latest Aorus Master x570 BIOS beta version just posted yesterday (F12g), and without even entering BIOS after the flash, I've been running Prime95 for an hour now. Fingers crossed... Here's a screenshot:


Edit: it failed in one worker at 1 hour 21 minutes mark, a few seconds after starting to watch a Youtube video. I wonder if there could be any relation.
I'll grab the beta. It must of just been released. I've looked for it everyday.\

Nope. only the non-wifi elite had f12g. No test bios for elite wifi.

CPU
Ryzen 3950X
Motherboard
Aorus Wifi Elite
GPU
Gigabyte AORUS GeForce RTX 2080 Xtreme WATERFORCE WB
RAM
G.Skill Ripjaws 32GB (2 x 16GB) DDR4-3600 PC4-28800 CL16
Hard Drive
Gigabyte AORUS 1TB SSD 3D TLC NAND M.2 2280 PCIe NVMe 4
Power Supply
Seasonic USA Focus Plus 750 Watt 80 Plus Gold ATX
Cooling
Corsair Hydro XC7 RGB CPU Water Block
Cooling
Distrub Plate w/pump RGV-LAN-011-P
Cooling
Corsair XR5 360
Cooling
Corsair XR5 280
Cooling
FAN Artic Cool P14 X 4
Cooling
FANS Corsair LL120 x 3
Cooling
FANS LL140 X 2
Cooling
Corsair Commander Pro
Cooling
Phantek 8 port pwm controller
Case
Lian Li PC-O11 Dynamic
Monitor
Dell UltraSharp 49 Curved Monitor: U4919DW
Keyboard
Mastercool sk650
Mouse
corsair dark core rgb
Other
Lian LI diffused rgb strip
Other
Corsair LS100 diffused rgb strips pack of 2
▲ hide details ▲

Last edited by Steven Stacy; 06-03-2020 at 05:14 PM.
Steven Stacy is offline  
post #85 of 100 (permalink) Old 06-04-2020, 04:28 AM
New to Overclock.net
 
Join Date: Feb 2017
Location: Sweden
Posts: 1,297
Rep: 54 (Unique: 42)
Well at least I can say I have no issues on the 3800X even if I have tortured it with excessive voltage since day one.
100% stable on stock on Gigabyte X570 Aorus Xtreme F12f. Took the occasion as I updated my BIOS from F12a to F12f version.

Just loaded optimized defaults and it was all OK, even went and enabled XMP @ 4000Mhz and all was well as I only needed increase stock DRAM voltage from 1.200 -> 1.350V for it to work there.
So I wonder what the issue is with the higher core counts. I've used "excessive" voltage on my 3800X a long time and as far as I can tell it's not changed a bit since start.

Gigabyte X570 AORUS XTREME, Ryzen 7 3800X PBO boost, 2x8GB 4866/1900 18.27.24.48.90.tRFC 680 1T 1.650V [Micron Rev.J]
4x8GB 3800/1900 13.23.17.30.66.tRFC 600 1T 1.700V Kingston HX434C19FB2K2/16 3466C19 1.2V (Micron Rev.E 19nm [D9VPP]),
2x8GB Kingston HX436C17FB3K2/16 3600C17 1.35V (Micron Rev.J 17nm),
2x8GB 3866/1933 15.21.20.32.60.tRFC 448 1T 1.590V Kingston HX436C17FB3K2/16 3600C17 1.35V (Hynix DJR 17nm)
Nighthog is offline  
post #86 of 100 (permalink) Old 06-04-2020, 05:25 AM
New to Overclock.net
 
rdr09's Avatar
 
Join Date: Mar 2011
Location: From the US but lives in Africa
Posts: 19,973
Rep: 892 (Unique: 643)
Quote: Originally Posted by smonkie View Post
I've just installed latest Aorus Master x570 BIOS beta version just posted yesterday (F12g), and without even entering BIOS after the flash, I've been running Prime95 for an hour now. Fingers crossed... Here's a screenshot:



https://imgur.com/3yATRb9

Edit: it failed in one worker at 1 hour 21 minutes mark, a few seconds after starting to watch a Youtube video. I wonder if there could be any relation.
Is the Vcore reading of 0.9XXv correct? Isn't too low? Sorry, I do not own same cpu. My 3600 jumps to 1.3XXv at stock running P95. Not much can be seen from that HWINFO screenshot. Was hoping to see stuff like SVI2 readings. See example.

EDIT: At post#61, the Stilt mentioned about what exactly should happen if cpu is at optimized default.
Attached Thumbnails
Click image for larger version

Name:	P95 at stock.PNG
Views:	24
Size:	935.0 KB
ID:	351638  


By Built-in Benchmark Queen:
My 2.5 years old delidded 7980xe @ 4600mhz all core with 49ns memorylatency (4000c16 twekaed) is faster than 3950x max overclocked on water in every scenario. Pretty good for a ancient cpu
In gaming it has no chance vs my new 10900k @ 5400mhz and 4600c16 tweaked memory. ~35ns memorylatency.
2nd AMD Build
(10 items)
CPU
2700
Motherboard
470
GPU
5700
RAM
3466/14
Hard Drive
1000
Power Supply
700
Case
212
Operating System
10/64
Monitor
55/2160
Keyboard
M100
▲ hide details ▲

Last edited by rdr09; 06-04-2020 at 05:56 AM.
rdr09 is offline  
post #87 of 100 (permalink) Old 06-05-2020, 11:11 PM
New to Overclock.net
 
LetterKilled's Avatar
 
Join Date: Mar 2012
Location: Sacramento
Posts: 74
Rep: 1 (Unique: 1)
I’ve tried to talk to gigabyte about the issue. The first reply took forever. None of my questions were answered -After that it was all really stupid replies like “is the software updated?” “We haven’t heard of any issues that relate to that at all” “maybe try an older bios” and when I ask to start an RMA they just jump around the question. I don’t know if I got unlucky and I’m talking with someone who doesn’t know what they’re doing or that’s how they are? It’s kind of frustrating. It’s almost to the point where I’m going to just cut the loss and get a different board.
LetterKilled is offline  
post #88 of 100 (permalink) Old 06-24-2020, 12:12 AM
New to Overclock.net
 
mongoled's Avatar
 
Join Date: Oct 2011
Location: Ayia Napa, Cyprus
Posts: 690
Rep: 24 (Unique: 19)
Just thinking out load....

Have any of those who have issues tried playing with the VDDG CCD (not IOD) voltage to "fix" this issue ?

As im troubleshooting my own Prime95 overclock, im seeing some weird anomolies I cant explain and have started playing with this voltage to see if I can stabilize my own setting with less CPU core voltage.

For those who can understand the Ryzen architecture better than myself, can we make an assumption that small FFT data is passed over the infinity fabric of the CCXs ?

Im currently doing my own testing to see what effects occur by increasing the VDDG CCD voltage while decreasing the CPU Core voltage while running prime95 small FFTs on my overclock.

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 online now  
post #89 of 100 (permalink) Old 06-24-2020, 03:10 AM
New to Overclock.net
 
Join Date: Feb 2017
Location: Sweden
Posts: 1,297
Rep: 54 (Unique: 42)
Quote: Originally Posted by mongoled View Post
Just thinking out load....

Have any of those who have issues tried playing with the VDDG CCD (not IOD) voltage to "fix" this issue ?

As im troubleshooting my own Prime95 overclock, im seeing some weird anomolies I cant explain and have started playing with this voltage to see if I can stabilize my own setting with less CPU core voltage.

For those who can understand the Ryzen architecture better than myself, can we make an assumption that small FFT data is passed over the infinity fabric of the CCXs ?

Im currently doing my own testing to see what effects occur by increasing the VDDG CCD voltage while decreasing the CPU Core voltage while running prime95 small FFTs on my overclock.
I had trouble with AVX2 LargeFFT when going for 4866/1900... It became utterly finicky for voltages on the VDDG/VDDP & vSOC. Lower speeds are more tolerable with most voltages, but if set wrong "work" but can give glitches. Not high Frequency like this, doesn't tolerate "wrong" settings.

Only ~950/950 VDDG_CCD/IOD works to make it pass, having enough VDDP is also critical, the recommended 900mv is to little, even 1000mv wasn't really enough.

I might be suspecting the 3950X is more taxing on FCLK? Maybe AUTO doesn't work? I know they are high when on AUTO settings. Above 1000mv on VDDG voltages are finicky in general from my testing. They make AVX2 fail more often than not.
Low VDDG_CCD voltage will make AVX2 fail also. My sample has a floor around 950mv to function without fail.

I noted a weird crossloading issue, AVX2 would work if I did nothing major but the minute I started a YouTube video they would crash. Had to adjust the CAD_BUS settings to fix that.

Gigabyte X570 AORUS XTREME, Ryzen 7 3800X PBO boost, 2x8GB 4866/1900 18.27.24.48.90.tRFC 680 1T 1.650V [Micron Rev.J]
4x8GB 3800/1900 13.23.17.30.66.tRFC 600 1T 1.700V Kingston HX434C19FB2K2/16 3466C19 1.2V (Micron Rev.E 19nm [D9VPP]),
2x8GB Kingston HX436C17FB3K2/16 3600C17 1.35V (Micron Rev.J 17nm),
2x8GB 3866/1933 15.21.20.32.60.tRFC 448 1T 1.590V Kingston HX436C17FB3K2/16 3600C17 1.35V (Hynix DJR 17nm)

Last edited by Nighthog; 06-24-2020 at 03:14 AM.
Nighthog is offline  
post #90 of 100 (permalink) Old 06-24-2020, 03:12 PM
New to Overclock.net
 
smonkie's Avatar
 
Join Date: Sep 2013
Posts: 181
Rep: 4 (Unique: 4)
Quote: Originally Posted by The Stilt View Post
The CPU will adjust the load-line and voltage autonomously, user actions are not required and can actually interfere the proper function.

If a CPU requires ANY changes to the default configuration (optimized defaults / failsafe) to run stable while operating within the spec (temperature / cooling wise), it is defective, period.

That CPU isn't running at stock thou, since the FCLK is higher than the rated max at any memory config (1600MHz, i.e. 3200MHz MEMCLK).

Also, based on the reported power draw, ASRock is still fooling around with the CPU power management.
A 3950X doesn't consume less power than its default limit (142W) during Prime95, unless the power or current limits are restricted to a lower value than stock, or the CPU is overheating...
So the FullScaleCurrent biasing alone might be the cause for the instability, since the power management of the CPU no longer works properly.
Just wanna update: you sir were right. A CPU which can't pass whatever test you throw at it with BIOS default settings, it's a defective unit. Period. Any time you spend trying to make it work, is time you are losing.

After receiving a new one by the dealer, I can confirm this one can pass anything with any BIOS version. It even passes Prime with negative voltage offset.

So yeah, I've been through TWO defective 3950X, which doesn't speak very well about Ryzen's QC. It had never EVER passed me before with a new CPU.
smonkie 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: 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