Stock 3950x crashes with prime95 small FFTs - Page 13 - Overclock.net - An Overclocking Community

Forum Jump: 

Stock 3950x crashes with prime95 small FFTs

Reply
 
Thread Tools
post #121 of 130 (permalink) Old 07-30-2020, 03:12 PM
New to Overclock.net
 
samgu's Avatar
 
Join Date: Nov 2017
Posts: 30
Rep: 0
One more for the club, but mine is a R5 3600 batch May 2020. finished the build two weeks ago, ran all sorts of stress testing (CB20 several times, Prime95 for 20 minutes, Realbench for 30 minutes, CPU-Z for 30 minutes) and never got an error. Today, playing Assassin's Creed Syndicate for 15 minutes, it suddenly rebooted. I restarted the game, played for one hour without issues. I am running CPU at stock and the only OC is the XMP enabled for my RAM. By reading all of the comments in here, I am not sure what to think.

It's not the first time I play the game, I played for around 13 hours, 8 hours of Destiny 2, and no issues.

The highest temperature I got was 81 degrees during Prime95. While gaming, it goes up to 70 degrees. GPU never passed 79 degrees.

AMD Ryzen 5 3600 stock + Patriot Viper Steel 3600 CL17 stock
MSI B550 Gaming Edge Wi-Fi
Sapphire RX 5700 XT Pulse stock
Kingston A2000 250GB + Crucial P1 1TB
Seasonic Core GC 650W Gold
Fractal Design Meshify C White
Intake: 2 x 140mm intake -front- + Exhaust: 2 x 120mm -back-top-
samgu is offline  
Sponsored Links
Advertisement
 
post #122 of 130 (permalink) Old 08-07-2020, 05:09 PM
New to Overclock.net
 
Florida_Dan's Avatar
 
Join Date: Aug 2011
Location: Florida
Posts: 297
Rep: 9 (Unique: 9)
So just updating this thread 23 days after tendering my first sample of the 3950x to UPS for shipment back to the vendor who I purchased it from (with an RMA number.)

The replacement arrived this afternoon and went directly from the box to the motherboard. (Yes, the vendor accepted the return and sent a replacement.)

It's been running Prime95 small FFTs for 2.5 hours at this point. The first sample wouldn't run these settings for 2 seconds.

I'm completely stock (I cleared CMOS when I took the old chip out so I wouldn't forget to reset the LLC settings I'd put in to make it stable before.)

There will be more testing for sure, but I'm way further along than I was the first time.

Surprisingly, the sample I got this time is actually older than the first one. The new sample comes from batch 2003PGT. So reading that code based on information given earlier in this thread, the new sample was made in January 2020, at the facility in Malaysia from a chip diffused in Texas. My original chip was made at the same facility, also from a Texas wafer, but was manufactured in March.


Florida_Dan is offline  
post #123 of 130 (permalink) Old 08-07-2020, 09:04 PM
New to Overclock.net
 
Join Date: Apr 2020
Posts: 4
Rep: 0
Reading this thread really does not paint a positive picture of AMD binning and quality control. I went through 3 processors and 2 motherboards (X570 Aorus Xtreme). I am almost tempted to try replacing the processor again.

The previous poster mentioned one sample failing after 2 seconds while an older batch works for over 2 hours in the same motherboard.

In my initial motherboard, two processors failed after about 30 minutes (includes my current 3950x), while another unit would last less than 10 seconds and completely crash the PC with a reboot and bsod. Currently in the warranty replacement board I can only get about 1.5-2 hours before an error occurs.

Some people are having luck so I might do the AMD RMA again, otherwise what are we paying top dollar for.
rob-tech is offline  
Sponsored Links
Advertisement
 
post #124 of 130 (permalink) Old 08-07-2020, 09:07 PM
ٴٴٴ╲⎝⧹˙͜>˙⧸⎠╱
 
TK421's Avatar
 
Join Date: May 2011
Posts: 5,983
Rep: 165 (Unique: 128)
Quote: Originally Posted by rob-tech View Post
Reading this thread really does not paint a positive picture of AMD binning and quality control. I went through 3 processors and 2 motherboards (X570 Aorus Xtreme). I am almost tempted to try replacing the processor again.

The previous poster mentioned one sample failing after 2 seconds while an older batch works for over 2 hours in the same motherboard.

In my initial motherboard, two processors failed after about 30 minutes (includes my current 3950x), while another unit would last less than 10 seconds and completely crash the PC with a reboot and bsod. Currently in the warranty replacement board I can only get about 1.5-2 hours before an error occurs.

Some people are having luck so I might do the AMD RMA again, otherwise what are we paying top dollar for.

amd quality control seems to be looser than intel


see the oc potential of early sample zen2 compared to current

nͫٴiͤٴcͫٴeͤ੮Һ૯ ცɿ૭ ૭คעٴ
█▀█ █▄█ ▀█▀ ▀█▀


TK421 is offline  
post #125 of 130 (permalink) Old 08-08-2020, 07:50 AM
New to Overclock.net
 
samgu's Avatar
 
Join Date: Nov 2017
Posts: 30
Rep: 0
I just got the same error for the third time, and, again, after playing one hour of Destiny 2. I will start the RMA process.

AMD Ryzen 5 3600 stock + Patriot Viper Steel 3600 CL17 stock
MSI B550 Gaming Edge Wi-Fi
Sapphire RX 5700 XT Pulse stock
Kingston A2000 250GB + Crucial P1 1TB
Seasonic Core GC 650W Gold
Fractal Design Meshify C White
Intake: 2 x 140mm intake -front- + Exhaust: 2 x 120mm -back-top-
samgu is offline  
post #126 of 130 (permalink) Old 08-09-2020, 11:24 AM
New to Overclock.net
 
samgu's Avatar
 
Join Date: Nov 2017
Posts: 30
Rep: 0
I decided to do some more testing to try to solve it with what has been said, and I noticed now that during the blender test in Prime95, the line "Power Reporting Deviation (Accuracy)" on HWINFO64 dropped below 90% and was marked as red. That happened every time the temperature got close to 90 degrees. Is this a problem? I have PBO on and XMP.

By the way, Prime95 ran for 50 minutes and I got no error.

AMD Ryzen 5 3600 stock + Patriot Viper Steel 3600 CL17 stock
MSI B550 Gaming Edge Wi-Fi
Sapphire RX 5700 XT Pulse stock
Kingston A2000 250GB + Crucial P1 1TB
Seasonic Core GC 650W Gold
Fractal Design Meshify C White
Intake: 2 x 140mm intake -front- + Exhaust: 2 x 120mm -back-top-
samgu is offline  
post #127 of 130 (permalink) Old 08-11-2020, 03:19 AM
New to Overclock.net
 
Martin778's Avatar
 
Join Date: Feb 2012
Location: Central EU
Posts: 1,608
Rep: 18 (Unique: 18)
I've seen topics telling the same about the 3000 threadrippers, crashing on Small FFT AVX2. The case seems simple to me, the CPU's get wonky above 85-90*C. Have had it with almost all Intels as well, they cannot run SmallFFT AVX reliably above a certain temperature threshold.
They (both Intel and AMD) should implement something like a heavy offset when it sees SmallFFT's running, comparable to what Nvidia/Radeon do to Furmark lovers and also spank board manufacturers for trying to squeeze every last bit of these CPU's even on stock settings by setting agressive V/F curves etc.

3960X running 92-95*C with PBO on any cooler I've tried is insane, they really need to fix these CPU's but I doubt they would as it might induce slowing the CPU by a good bit by lowering the all-core clock.

Threadripper 3960X @ broken PBO EDC=40 | Aorus TRX40 Xtreme | Enermax Liqtech II 360 TR4 + EK Clear flush + 3x Noctua A12x25 | 4x8GB G.Skill TridentZ Neo 3800C14 @ 3733 CL14-15-15-30 CR1 | EVGA Kingpin 2080Ti @ 2115MHz / 8250MHz + Noctua A12-25 | 4x1TB 970 PRO raid0 | Corsair MP600 2TB | Seasonic Prime Platinum 1300W | In Win D-Frame 2.0 Gold | ASUS PG349Q

Last edited by Martin778; 08-11-2020 at 03:24 AM.
Martin778 is online now  
post #128 of 130 (permalink) Old 08-11-2020, 03:30 AM
New to Overclock.net
 
rdr09's Avatar
 
Join Date: Mar 2011
Location: From the US but lives in Africa
Posts: 20,042
Rep: 894 (Unique: 644)
Quote: Originally Posted by samgu View Post
I decided to do some more testing to try to solve it with what has been said, and I noticed now that during the blender test in Prime95, the line "Power Reporting Deviation (Accuracy)" on HWINFO64 dropped below 90% and was marked as red. That happened every time the temperature got close to 90 degrees. Is this a problem? I have PBO on and XMP.

By the way, Prime95 ran for 50 minutes and I got no error.
Kinda a close to the spec'ed temp limit. What cooler are you using?

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 ▲
rdr09 is offline  
post #129 of 130 (permalink) Old 08-11-2020, 04:20 AM
OG AMD
 
Redwoodz's Avatar
 
Join Date: Jan 2010
Location: Silicon Valley
Posts: 8,751
Rep: 600 (Unique: 465)
You guys do know that Windows OS has PowerPlans don't you? Have you even tried different power plans? Updated Windows? Wouldn't surprise me at all this is all about Windows +Motherboard Agesa growing pains. Who in the heck runs Prime95 at completely stock settings anyways? Never once have I heard of someone installing a cpu then running Prime95 for 24 hrs. They always do it after overclocking, or downclocking and undervolting or changing RAM settings. Just a few thoughts.

https://valid.x86.fr/cache/banner/4d8m5e-6.png
https://valid.x86.fr/cache/banner/bs72bx-6.png
AMD Athlon II 450 @4.74GHz http://valid.canardpc.com/show_oc.php?id=2204588
AMD MaxxMemm Rankings
960T @ 4.6GHz http://valid.canardpc.com/show_oc.php?id=2171450
RedwoodCustomPC
AM4
(6 items)
Redwoodz
(10 items)
Professional
(13 items)
CPU
Ryzen 5 2600
Motherboard
ASRock X470 Taichi
GPU
MSI RX570 Armor MKII 8GB
RAM
GSKill Trident Z D-die
RAM
TeamGroup Excalibur
Power Supply
GSKill Ripjaws PS850G
CPU
AMD FX Vishera 8350
Motherboard
ASRock Extreme 9 990FX
GPU
HIS X 7970 IceQ
RAM
Gskill Trident
Hard Drive
Samsung 840 Pro
Power Supply
Seasonic SS-760XP
Cooling
TT Water 2.0
Monitor
Samsung Series 9 27"
Keyboard
Mionix Zibal
Mouse
Mionix 5000
CPU
Phenom II X6 1100t
Motherboard
MSI 890FX GD65
GPU
MSI Radeon HD5670
RAM
GSkill RipjawsX DDR3 PC3 12800 2x4GB CL8
Hard Drive
WD Black 1TB SATA III
Optical Drive
Samsung BD
Power Supply
Kingwin Lazer Platinum 500w
Cooling
Zalman 9900MAX
Case
Fractal Design R3
Operating System
Windows 7 64 Professional
Monitor
AOC 22" LED
Keyboard
Logitech
Other
Samsung 470 SSD 128GB
▲ hide details ▲


Redwoodz is offline  
post #130 of 130 (permalink) Old 08-11-2020, 09:50 AM
New to Overclock.net
 
samgu's Avatar
 
Join Date: Nov 2017
Posts: 30
Rep: 0
Quote: Originally Posted by rdr09 View Post
Kinda a close to the spec'ed temp limit. What cooler are you using?
The power deviation line I was talking about shows percentages, not temps. The highest temp I reached running Prime95 is 81 degrees. I am using a Thermalright ARO-M14.

AMD Ryzen 5 3600 stock + Patriot Viper Steel 3600 CL17 stock
MSI B550 Gaming Edge Wi-Fi
Sapphire RX 5700 XT Pulse stock
Kingston A2000 250GB + Crucial P1 1TB
Seasonic Core GC 650W Gold
Fractal Design Meshify C White
Intake: 2 x 140mm intake -front- + Exhaust: 2 x 120mm -back-top-
samgu 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