Announcement

Collapse
No announcement yet.

KVA2 RTX 2060 Super 8GB Desktop artifacts

Collapse
X
Collapse
 
  • Page of 1
  • Filter
  • Time
  • Show
Clear All
new posts
Previous template Next
    #1

    KVA2 RTX 2060 Super 8GB Desktop artifacts

    Hello (sorry for my bad English),

    I have a strange problem with my KVA2 RTX 2060 Super. As soon as Windows 11 is loaded, artifacts appear. Now the crazy thing: When I start Furmark the artifacts are gone and Furmark runs great without crashing. Mats test shows me no errors (PASS). Drivers can also be loaded without any problems.


    Mods shows me the following error:

    Code:
    MODS start: Thu Feb  1 19:16:00 2024

Command Line : gputest.js -oqa -old_gold -test 94 -dramclk_percent 100 -ignore_fatal_errors -run_on_error -fan_speed 70 -matsinfo

CPU
Foundry   : AuthenticAMD
Name      : AMD Ryzen 5 2600 Six-Core Processor
Family    : 15
Model     : 8
Stepping  : 2

Version
MODS           : 400.281
OperatingSystem: Linux (x86_64)
Kernel         : 4.17.4-gentoo
KernelDriver   : 3.87
SBIOS Version  : 4901
SBIOS Date     : 07/25/2022
HostName       : tinylinux

                 GPU 0 [08:00.0]  dev.sub 0.0            
                 ----------------------------------------
DevInst        : 0                                        
PCI Location   : 0x00, 0x08, 0x00, 0x00                  
GPU DID        : 0x1f06                                  
PDI            : 0x33d91aab89d9290c                      
Raw ECID       : 0x014045c00000007028442191              
Raw ECID (GHS) : 0x1647028442000000017020280              
ECID           : PS2X48-23_x04_y10                        
Device Id      : TU106                                    
Revision       : a1                                      
NV Base        : 0xfb000000                              
FB Base        : 0xd0000000                              
IRQ            : 59                                      
WARNING: GPU 0 [08:00.0] PCIE speed capability (8000Gbps) higher than down stream port link speed (2500Gbps)
ERROR: SMBus controller Not found
Setting Fan 1 of GPU 0 [08:00.0] to 70%
Foundry        : TSMC                                    
Subsystem VID  : 0x10de                                  
Subsystem DID  : 0x13a3                                  
Board ID       : 0x0160                                  
Chip SKU       : 410-0                                    
Project        : G160-0062                                
Fuse File Fmt  : JSON                                    
Display        : 0x00000100 (id)                          
SBIOS Init     : Primary                                  
Native Mode    : 1920x1080                                
Memory Size    : 8192 MB                                  
FB Vendor      : Micron                                  
RAM Protocol   : GDDR6                                    
RAM Config     : 1                                        
ROM Version    : 90.06.44.80.5d                          
ROM Type       : Partner Production                      
ROM OEM Vendor : NVIDIA                                  
ROM Partner    : galaxy                                  
ROM Project ID : 107293galaxy                            
ROM Timestamp  : 2019-6-12 17:09:31                      
ROM Expiration : 2019-12-9 17:09:31                      
ROM GUID       : 2D6AC2F846E34A8DBF8F581AA44D2AD8        
PPC Pri/Sec FW : Unknown                                  
PPC Driver Ver : Unknown                                  
PState (mode)  : 8 5 3 2 [0]                              
PState Version : 3.5                                      
EDC            : Disabled                                
GPC Clock      : 1890.0/1885.3 MHz NAFLL                  
DRAM Clock     : 7001.0/7014.0 MHz DEFAULT                
Host Clock     : 1350.0/1349.7 MHz NAFLL                  
XBar Clock     : 1800.0/1800.1 MHz NAFLL                  
Sys Clock      : 1890.0/1888.4 MHz NAFLL                  
Power Clock    : 540.0/540.5 MHz DEFAULT                  
NVDec Clock    : 1755.0/1746.2 MHz NAFLL                  
Display Clock  : 1330.0/1329.0 MHz DEFAULT                
NVVDD          : 1062 mV                                  
GPC  Mask      : 0x07 (3 GPCs)                            
TPC  Mask      : [3f 3f 3e] (17 TPCs)                    
FB   Mask      : 0x0f (4 FB Partitions)                  
ROP/L2 Mask    : [3 3 3 3] (8 ROP/L2s)                    
PES  Mask      : [7 7 7] (9 PESes)                        
FBIO Mask      : 0x0f (4 FBIO Partitions)                
FBIO Shift Mask: 0x00                                    
XP   Mask      : 0x03 (2 3gio Pads)                      
Nvdec Mask     : 0x07 (3 engines)                        
Nvenc Mask     : 0x01 (1 engine)                          
Gpu Temp       : 48 deg C                                
PEX Rx Lanes   : 0x00ff                                  
PEX Tx Lanes   : 0x00ff                                  
PEX Det. Lanes : 0x00ff                                  
PEX Width, ASLM: 8 lanes, Not Supported                  
PEX Link Speed : 8.0 Gbit/s                              
PEX BandWidth  : 64.0 Gbit/s                              
ASPM, ASPM-CYA : (L0s/L1, Disabled)                      
ASPM L1SS, CYA : (Disabled, L1.1/L1.2)                    
LTR            : Disabled                                


Chipset
VID            : 1022 (AMD)
Chipset DID    : 1450 (Unknown)
Chipset ASPM   : L1
ASPM L1 SS     : Disabled

RM Version     : rel/gpu_drv/r400/r400_00-823
resumehandler.js: 1
gputest.js     : 81
oqa.spc        : 3
boards.js      : 3

Running test(s) on GPU 0 [08:00.0] (DID: 0x1f06)
Enter SetPState (test 0)
Switched to PState 0 (0.max). Pcie Speed=8000, x8
Speed = 8.0Gb/s,  expected 8.0Gb/s
UpStreamWidth = 8,  expected 16
Error 020000000143 : Global.SanityCheckLink PCI Express bus error
Exit 020000000143 : SetPState (test 0) PCI Express bus error
Error!
Error 020000000143 : SetPState (test 0) PCI Express bus error
Enter NewWfMats (test 94)
Exit 000000000000 : NewWfMats (test 94) ok
GPU tests completed.

Failure(s) :
 LOOP           TEST                 CODE               MESSAGE
 ----  ------------------------  ------------  ---------------------------
   1   SetPState                 020000000143  PCI Express bus error

Error Code = 020000000143 (PCI Express bus error)

                                        
 #######     ####    ########  ###      
 #######    ######   ########  ###      
 ##        ##    ##     ##     ###      
 ##        ##    ##     ##     ###      
 #######   ########     ##     ###      
 #######   ########     ##     ###      
 ##        ##    ##     ##     ###      
 ##        ##    ##  ########  ########
 ##        ##    ##  ########  ########
                                        

MODS end  : Thu Feb  1 19:16:26 2024  [26.301 seconds (00:00:26.301 h:m:s)]
    What could be the problem?

    Many thanks
    Tags: None
    #2
    it is hard fault. simple mats shows it

    Comment

      #3
      Is there a chance to repair? If so, what should I look for? Equipment is available (microscope, soldering station...)

      Comment

        #4
        Originally posted by klugec View Post
        Is there a chance to repair? If so, what should I look for? Equipment is available (microscope, soldering station...)
        maybe. u have bad memory ic, bad solder joint under it or bad solder joint under gpu chip

        first u neet to run mats. to remove memory chip u need preheater, hotair station, new chip or stencil to reball original, solder balls, flux, soldering iron

        Comment

          #5
          Originally posted by ktmmotocross View Post

          maybe. u have bad memory ic, bad solder joint under it or bad solder joint under gpu chip

          first u neet to run mats. to remove memory chip u need preheater, hotair station, new chip or stencil to reball original, solder balls, flux, soldering iron
          Mats doesn't show me any errors, I also tried with 1024 MB. Mods shows me some errors at dramclk_percent 120. But I guess that has nothing to do with it? Should I reball every memory chip (stencil is available for DDR6x) or can I exclude the chips?

          Comment

            #6
            GPU might be unstable at lower voltages.
            OpenBoardView — https://github.com/OpenBoardView/OpenBoardView

            Comment

              #7
              Originally posted by piernov View Post
              GPU might be unstable at lower voltages.
              Would it help to increase the voltage? The best thing would be to write this directly into the BIOS. Can this be edited and flashed?

              Comment

                #8
                is that rtx only gpu in system? dont u have integrated gpu?

                Comment

                  #9
                  Originally posted by ktmmotocross View Post
                  is that rtx only gpu in system? dont u have integrated gpu?
                  i have a test bench, AMD 2600 without integrated gpu. I have also a riser cable for testing (pci-e x1).
                  I have a second GPU to test, an ATI card.

                  Comment

                    #10
                    Got the same problem with Asus 1080Ti Error Code = 020000000143 (PCI Express bus error), Furmark runs great without crashing, crashes only in games without artifacts, did you find the cause?

                    Comment

                      #11
                      your cpu cant support pcie x16 try it with other cpu.

                      Comment

                        #12
                        check memory marking. i bet it is older micron and u must replace whoe memory chips

                        Comment

                        Previous template Next
                        Working...
                        X