Доброго времени суток. Gопала по вкусной цене б/у карта MSI 3080 ventus x3 ОС Произведена 05 2021. У продавца повертел покрутил, погонял тесты... Карта обслужена, без подозрительных следов. Ушатаны вертушки, но я сразу нацелился на замену.
Дома сходу дал ей +1000 по памяти и сразу вылезли артефакты в 3d нагрузке. Без разгона все хорошо. Ну опыта у меня с GDDR6X не было и сразу подумал, что мол не держит разгон, да и все равно. Мне разгон не нужен. Позже, прочитал, что в этой памяти, есть коррекция ошибок и вообще она себя так вести не должна (Пишут мол, память умная и если не тянет, просто начинает скидывать сама частоту).
Начал разбираться глубже. Прогнал mods/mats. С dramclk 107% и выше, дает ошибку в тесте 166. 110%+ в 94ом тесте, та же ошибка. 106% - стабильна. Температура, оказывают влияние, но только на первый (холодный) прогон теста (166)... дальше ошибка.
Лог MODS, с ошибкой памяти
MODS arguments :
MODS start: Thu May 18 02:13:44 2023
Command Line : gputest.jse -oqa -dramclk_percent 108 -test 166 -run_on_error
CPU
Arch : x86_64
Name : 12th Gen Intel(R) Core(TM) i5-12600KF
Cores : 16
Version
MODS : 455.120
System
OperatingSystem: Linux (x86_64)
Kernel : 5.9.1-gentoo-x86_64
KernelDriver : 4.00
SBIOS Version : F4a
SBIOS Date : 04/14/2023
HostName : tinylinux
Available RAM : 31592/32082 MB (Free/Size)
Sys-uuid :
HDD-Serno :
GPU 0 [01:00.0] dev.sub 0.0
----------------------------------------
DevInst : 0
PCI Location : 0x00, 0x01, 0x00, 0x00
GPU DID : 0x2206
PDI : 0x8f0cc023bc50bd40
Raw ECID : 0x002001c00000005d78c305c4
Raw ECID (GHS) : 0x4705d78c30400000007000040
ECID : SAMSUNG-SNNZ31-07_x0_y1
Device Id : GA102
Revision : a1
Sub Revision : 1
NV Base : 0x52000000
FB Base : 0x40000000
IRQ : 16
WARNING: GPU 0 [01:00.0] PCIE speed capability (16000Mbps) higher than down stream port link speed (2500Mbps)
UUID : 7d04fc8e-6b25-629b-c548-3bd2c3e7e337
Platform : Hardware
Foundry : Samsung
Subsystem VID : 0x1462
Subsystem DID : 0x3896
Board ID : 0x023f
Project : G132-0030
Fuse File Fmt : JSON
Display : 0x00000400 (id)
SBIOS Init : Primary
Native Mode : 1920x1080
Memory Size : 10240 MB
FB Vendor : Micron
RAM Protocol : GDDR6X
RAM Config : 2
WARNING: WAR(2791626): 'Num Row Remaps' reporting is disabled
ROM Version : 94.02.26.40.99
ROM Type : Partner Production
ROM OEM Vendor : NVIDIA
ROM Partner : msi
ROM Project ID : 110182
ROM Timestamp : 2020-5-22 06:04:13
ROM Expiration : 2020-11-18 06:04:13
PState (mode) : 8 5 3 2 [0]
PState Version : 4.0
EDC : (RD,WR,REPLAY)
GPC Clock : 1995.000 MHz DEFAULT
DRAM Clock : 10261.968 MHz DEFAULT
Host Clock : 1350.000 MHz NAFLL
XBar Clock : 1890.000 MHz DEFAULT
Sys Clock : 1680.000 MHz DEFAULT
Power Clock : 540.000 MHz NAFLL
NVDec Clock : 1755.000 MHz NAFLL
Display Clock : 1350.000 MHz DEFAULT
NVVDD : 1081.25 mV
MSVDD : 1075.00 mV
GPC Mask : 0x3f (6 GPCs)
TPC Mask : [x 3f 3f 3f 3f 3e 3e] (34 TPCs)
FB Mask : 0x37 (5 FB Partitions)
L2 Mask : [3 3 x 3 3 3] (10 L2s)
L2 Slice Mask : [ff ff x ff ff ff] (40 L2 Slices)
PES Mask : [x 7 7 7 7 7 7] (18 PESes)
ROP Mask : [x 3 3 3 3 3 3] (12 ROPs)
FBIO Mask : 0x37 (5 FBIO Partitions)
FBIO Shift Mask: 0x00
XP Mask : 0x03 (2 3gio Pads)
Nvdec Mask : 0x01 (1 engine)
Nvenc Mask : 0x01 (1 engine)
Nvjpg Mask : 0x00 (0 engines)
Ofa Mask : 0x01 (1 engine)
PCE Mask : 0x30 (2 PCEs)
Syspipe Mask : 0x01 (1 syspipe)
Gpu Temp : 48 deg C
PEX Rx Lanes : 0xffff
PEX Tx Lanes : 0xffff
PEX Det. Lanes : 0xffff
PEX Width, ASLM: 16 lanes, Not Supported
PEX Link Speed : 16.0 Gbit/s
PEX BandWidth : 256.0 Gbit/s
ASPM, ASPM-CYA : (Disabled, Disabled)
ASPM L1SS, CYA : (Disabled, L1.1/L1.2)
LTR : Enabled
Chipset
VID : 8086 (Intel)
Chipset DID : 4648 (Unknown)
Chipset ASPM : Disabled
ASPM L1 SS : Disabled
Chipset LTR : Enabled
RM Version : rel/gpu_drv/r455/r455_00-263
testlist.js : 6
resumehandler.js: 1
gputest.js : 30
oqa.spc : 2
boards.js : 1
Running test(s) on GPU 0 [01:00.0] (DID: 0x2206)
Enter SetPState (test 0) Thu May 18 02:13:46 2023
Switched to PState 0 (0.max). Pcie Speed=16000, x16
ClkM = 10261.97 MHz
ClkHost = 1350.00 MHz
ClkDisp = 1350.00 MHz
ClkGpc = 1995.00 MHz
ClkXbar = 1890.00 MHz
ClkSys = 1680.00 MHz
ClkHub = 810.00 MHz
ClkPwr = 540.00 MHz
ClkNvd = 1755.00 MHz
ClkPexGen = 4.00
NVVDD = 1081.25 mV
MSVDD = 1075.00 mV
Exit 000000000000 : SetPState (test 0) ok [0.026 seconds]
Enter GLStressZ (test 166) Thu May 18 02:13:46 2023
New EDC errors have occured:
EDC total errors: 1 (New)
EDC read errors: 1 (New)
EDC write errors: 0 (New)
EDC read replay errors: 0 (New)
EDC write replay errors: 0 (New)
Partition A:
total: 1 (New)
read: 1 (New)
Byte 4 (Rd): 1 (New)
write: 0 (New)
read replay: 0 (New)
write replay: 0 (New)
SUBP0 Max Replay Cycles: 0
SUBP1 Max Replay Cycles: 256
SUBP0 Max Replay Retries: 0
SUBP1 Max Replay Retries: 1
Max Delta: 1
CRC Tick: 750 (LTCCLK)
dev 0: GLStressZ 1191200 Frames, DrawPct 100.0, avg Watts 298.342, max Watts 309.968.
INPUT_PEX12V avg Watts 27.983, max Watts 31.291
INPUT_EXT12V_8PIN0 avg Watts 120.441, max Watts 125.976
INPUT_EXT12V_8PIN1 avg Watts 149.918, max Watts 154.525
INPUT_MISC0 avg Watts 67.607, max Watts 72.605
INPUT_MISC1 avg Watts 24.996, max Watts 25.803
INPUT_MISC2 avg Watts 62.122, max Watts 66.164
OUTPUT_NVVDD avg Watts 118.906, max Watts 127.476
OUTPUT_SRAM avg Watts 43.227, max Watts 44.160
GLStressZ "Post-Run" callback failed (EDC detected a memory-bus error).
EDC Errors Exceeded Per-Test Threshold:
EDC total errors: 1 (New) (0 Allowed)
EDC read errors: 1 (New) (0 Allowed)
EDC write errors: 0 (New) (0 Allowed)
EDC read replay errors: 0 (New) (0 Allowed)
EDC write replay errors: 0 (New) (0 Allowed)
Partition A:
total: 1 (New)
read: 1 (New)
Byte 4 (Rd): 1 (New)
write: 0 (New)
read replay: 0 (New)
write replay: 0 (New)
SUBP0 Max Replay Cycles: 0
SUBP1 Max Replay Cycles: 256
SUBP0 Max Replay Retries: 0
SUBP1 Max Replay Retries: 1
Max Delta: 1
CRC Tick: 750 (LTCCLK)
GLStressZ POST_RUN callbacks failed.
Exit 020000166134 : GLStressZ (test 166) EDC detected a memory-bus error [15.116 seconds]
Error!
GPU tests completed.
Failure(s) :
LOOP TEST CODE MESSAGE
---- ------------------------ ------------ ---------------------------
1 GLStressZ 020000166134 EDC detected a memory-bus error
Error Code = 020000166134 (EDC detected a memory-bus error)
####### #### ######## ###
####### ###### ######## ###
## ## ## ## ###
## ## ## ## ###
####### ######## ## ###
####### ######## ## ###
## ## ## ## ###
## ## ## ######## ########
## ## ## ######## ########
MODS end : Thu May 18 02:14:02 2023 [18.083 seconds (00:00:18.083 h:m:s)]
MODS start: Thu May 18 02:13:44 2023
Command Line : gputest.jse -oqa -dramclk_percent 108 -test 166 -run_on_error
CPU
Arch : x86_64
Name : 12th Gen Intel(R) Core(TM) i5-12600KF
Cores : 16
Version
MODS : 455.120
System
OperatingSystem: Linux (x86_64)
Kernel : 5.9.1-gentoo-x86_64
KernelDriver : 4.00
SBIOS Version : F4a
SBIOS Date : 04/14/2023
HostName : tinylinux
Available RAM : 31592/32082 MB (Free/Size)
Sys-uuid :
HDD-Serno :
GPU 0 [01:00.0] dev.sub 0.0
----------------------------------------
DevInst : 0
PCI Location : 0x00, 0x01, 0x00, 0x00
GPU DID : 0x2206
PDI : 0x8f0cc023bc50bd40
Raw ECID : 0x002001c00000005d78c305c4
Raw ECID (GHS) : 0x4705d78c30400000007000040
ECID : SAMSUNG-SNNZ31-07_x0_y1
Device Id : GA102
Revision : a1
Sub Revision : 1
NV Base : 0x52000000
FB Base : 0x40000000
IRQ : 16
WARNING: GPU 0 [01:00.0] PCIE speed capability (16000Mbps) higher than down stream port link speed (2500Mbps)
UUID : 7d04fc8e-6b25-629b-c548-3bd2c3e7e337
Platform : Hardware
Foundry : Samsung
Subsystem VID : 0x1462
Subsystem DID : 0x3896
Board ID : 0x023f
Project : G132-0030
Fuse File Fmt : JSON
Display : 0x00000400 (id)
SBIOS Init : Primary
Native Mode : 1920x1080
Memory Size : 10240 MB
FB Vendor : Micron
RAM Protocol : GDDR6X
RAM Config : 2
WARNING: WAR(2791626): 'Num Row Remaps' reporting is disabled
ROM Version : 94.02.26.40.99
ROM Type : Partner Production
ROM OEM Vendor : NVIDIA
ROM Partner : msi
ROM Project ID : 110182
ROM Timestamp : 2020-5-22 06:04:13
ROM Expiration : 2020-11-18 06:04:13
PState (mode) : 8 5 3 2 [0]
PState Version : 4.0
EDC : (RD,WR,REPLAY)
GPC Clock : 1995.000 MHz DEFAULT
DRAM Clock : 10261.968 MHz DEFAULT
Host Clock : 1350.000 MHz NAFLL
XBar Clock : 1890.000 MHz DEFAULT
Sys Clock : 1680.000 MHz DEFAULT
Power Clock : 540.000 MHz NAFLL
NVDec Clock : 1755.000 MHz NAFLL
Display Clock : 1350.000 MHz DEFAULT
NVVDD : 1081.25 mV
MSVDD : 1075.00 mV
GPC Mask : 0x3f (6 GPCs)
TPC Mask : [x 3f 3f 3f 3f 3e 3e] (34 TPCs)
FB Mask : 0x37 (5 FB Partitions)
L2 Mask : [3 3 x 3 3 3] (10 L2s)
L2 Slice Mask : [ff ff x ff ff ff] (40 L2 Slices)
PES Mask : [x 7 7 7 7 7 7] (18 PESes)
ROP Mask : [x 3 3 3 3 3 3] (12 ROPs)
FBIO Mask : 0x37 (5 FBIO Partitions)
FBIO Shift Mask: 0x00
XP Mask : 0x03 (2 3gio Pads)
Nvdec Mask : 0x01 (1 engine)
Nvenc Mask : 0x01 (1 engine)
Nvjpg Mask : 0x00 (0 engines)
Ofa Mask : 0x01 (1 engine)
PCE Mask : 0x30 (2 PCEs)
Syspipe Mask : 0x01 (1 syspipe)
Gpu Temp : 48 deg C
PEX Rx Lanes : 0xffff
PEX Tx Lanes : 0xffff
PEX Det. Lanes : 0xffff
PEX Width, ASLM: 16 lanes, Not Supported
PEX Link Speed : 16.0 Gbit/s
PEX BandWidth : 256.0 Gbit/s
ASPM, ASPM-CYA : (Disabled, Disabled)
ASPM L1SS, CYA : (Disabled, L1.1/L1.2)
LTR : Enabled
Chipset
VID : 8086 (Intel)
Chipset DID : 4648 (Unknown)
Chipset ASPM : Disabled
ASPM L1 SS : Disabled
Chipset LTR : Enabled
RM Version : rel/gpu_drv/r455/r455_00-263
testlist.js : 6
resumehandler.js: 1
gputest.js : 30
oqa.spc : 2
boards.js : 1
Running test(s) on GPU 0 [01:00.0] (DID: 0x2206)
Enter SetPState (test 0) Thu May 18 02:13:46 2023
Switched to PState 0 (0.max). Pcie Speed=16000, x16
ClkM = 10261.97 MHz
ClkHost = 1350.00 MHz
ClkDisp = 1350.00 MHz
ClkGpc = 1995.00 MHz
ClkXbar = 1890.00 MHz
ClkSys = 1680.00 MHz
ClkHub = 810.00 MHz
ClkPwr = 540.00 MHz
ClkNvd = 1755.00 MHz
ClkPexGen = 4.00
NVVDD = 1081.25 mV
MSVDD = 1075.00 mV
Exit 000000000000 : SetPState (test 0) ok [0.026 seconds]
Enter GLStressZ (test 166) Thu May 18 02:13:46 2023
New EDC errors have occured:
EDC total errors: 1 (New)
EDC read errors: 1 (New)
EDC write errors: 0 (New)
EDC read replay errors: 0 (New)
EDC write replay errors: 0 (New)
Partition A:
total: 1 (New)
read: 1 (New)
Byte 4 (Rd): 1 (New)
write: 0 (New)
read replay: 0 (New)
write replay: 0 (New)
SUBP0 Max Replay Cycles: 0
SUBP1 Max Replay Cycles: 256
SUBP0 Max Replay Retries: 0
SUBP1 Max Replay Retries: 1
Max Delta: 1
CRC Tick: 750 (LTCCLK)
dev 0: GLStressZ 1191200 Frames, DrawPct 100.0, avg Watts 298.342, max Watts 309.968.
INPUT_PEX12V avg Watts 27.983, max Watts 31.291
INPUT_EXT12V_8PIN0 avg Watts 120.441, max Watts 125.976
INPUT_EXT12V_8PIN1 avg Watts 149.918, max Watts 154.525
INPUT_MISC0 avg Watts 67.607, max Watts 72.605
INPUT_MISC1 avg Watts 24.996, max Watts 25.803
INPUT_MISC2 avg Watts 62.122, max Watts 66.164
OUTPUT_NVVDD avg Watts 118.906, max Watts 127.476
OUTPUT_SRAM avg Watts 43.227, max Watts 44.160
GLStressZ "Post-Run" callback failed (EDC detected a memory-bus error).
EDC Errors Exceeded Per-Test Threshold:
EDC total errors: 1 (New) (0 Allowed)
EDC read errors: 1 (New) (0 Allowed)
EDC write errors: 0 (New) (0 Allowed)
EDC read replay errors: 0 (New) (0 Allowed)
EDC write replay errors: 0 (New) (0 Allowed)
Partition A:
total: 1 (New)
read: 1 (New)
Byte 4 (Rd): 1 (New)
write: 0 (New)
read replay: 0 (New)
write replay: 0 (New)
SUBP0 Max Replay Cycles: 0
SUBP1 Max Replay Cycles: 256
SUBP0 Max Replay Retries: 0
SUBP1 Max Replay Retries: 1
Max Delta: 1
CRC Tick: 750 (LTCCLK)
GLStressZ POST_RUN callbacks failed.
Exit 020000166134 : GLStressZ (test 166) EDC detected a memory-bus error [15.116 seconds]
Error!
GPU tests completed.
Failure(s) :
LOOP TEST CODE MESSAGE
---- ------------------------ ------------ ---------------------------
1 GLStressZ 020000166134 EDC detected a memory-bus error
Error Code = 020000166134 (EDC detected a memory-bus error)
####### #### ######## ###
####### ###### ######## ###
## ## ## ## ###
## ## ## ## ###
####### ######## ## ###
####### ######## ## ###
## ## ## ## ###
## ## ## ######## ########
## ## ## ######## ########
MODS end : Thu May 18 02:14:02 2023 [18.083 seconds (00:00:18.083 h:m:s)]
Я так понимаю, что дело в контроллере памяти на GPU? Он при смерти или это, скажем так "Умеренное снижение производительности"? Буду благодарен за любые комментарии по этому поводу. Я не умею грамотно читать такие логи, поэтому рассчитываю на помощь специалистов.
Еще вылезла ошибка в тесте 360 (CuvidNVDEC). Не очень понимаю, что это за тест, и что значат ошибка которую он выдает. Погуглил - не нашел. Что это такое?
Ошибка CuvidNVDEC (test 360)
MODS arguments :
MODS start: Thu May 18 02:22:01 2023
Command Line : gputest.js -oqa -run_on_error -test 360
CPU
Arch : x86_64
Name : 12th Gen Intel(R) Core(TM) i5-12600KF
Cores : 16
Version
MODS : 455.120
System
OperatingSystem: Linux (x86_64)
Kernel : 5.9.1-gentoo-x86_64
KernelDriver : 4.00
SBIOS Version : F4a
SBIOS Date : 04/14/2023
HostName : tinylinux
Available RAM : 31558/32082 MB (Free/Size)
Sys-uuid :
HDD-Serno :
GPU 0 [01:00.0] dev.sub 0.0
----------------------------------------
DevInst : 0
PCI Location : 0x00, 0x01, 0x00, 0x00
GPU DID : 0x2206
PDI : 0x8f0cc023bc50bd40
Raw ECID : 0x002001c00000005d78c305c4
Raw ECID (GHS) : 0x4705d78c30400000007000040
ECID : SAMSUNG-SNNZ31-07_x0_y1
Device Id : GA102
Revision : a1
Sub Revision : 1
NV Base : 0x52000000
FB Base : 0x40000000
IRQ : 16
WARNING: GPU 0 [01:00.0] PCIE speed capability (16000Mbps) higher than down stream port link speed (2500Mbps)
UUID : 7d04fc8e-6b25-629b-c548-3bd2c3e7e337
Platform : Hardware
Foundry : Samsung
Subsystem VID : 0x1462
Subsystem DID : 0x3896
Board ID : 0x023f
Project : G132-0030
Fuse File Fmt : JSON
Display : 0x00000400 (id)
SBIOS Init : Primary
Native Mode : 1920x1080
Memory Size : 10240 MB
FB Vendor : Micron
RAM Protocol : GDDR6X
RAM Config : 2
WARNING: WAR(2791626): 'Num Row Remaps' reporting is disabled
ROM Version : 94.02.26.40.99
ROM Type : Partner Production
ROM OEM Vendor : NVIDIA
ROM Partner : msi
ROM Project ID : 110182
ROM Timestamp : 2020-5-22 06:04:13
ROM Expiration : 2020-11-18 06:04:13
PState (mode) : 8 5 3 2 [0]
PState Version : 4.0
EDC : (RD,WR,REPLAY)
GPC Clock : 2010.000 MHz DEFAULT
DRAM Clock : 9501.996 MHz DEFAULT
Host Clock : 1350.000 MHz NAFLL
XBar Clock : 1905.000 MHz DEFAULT
Sys Clock : 1695.000 MHz DEFAULT
Power Clock : 540.000 MHz NAFLL
NVDec Clock : 1770.000 MHz NAFLL
Display Clock : 1350.000 MHz DEFAULT
NVVDD : 1081.25 mV
MSVDD : 1081.25 mV
GPC Mask : 0x3f (6 GPCs)
TPC Mask : [x 3f 3f 3f 3f 3e 3e] (34 TPCs)
FB Mask : 0x37 (5 FB Partitions)
L2 Mask : [3 3 x 3 3 3] (10 L2s)
L2 Slice Mask : [ff ff x ff ff ff] (40 L2 Slices)
PES Mask : [x 7 7 7 7 7 7] (18 PESes)
ROP Mask : [x 3 3 3 3 3 3] (12 ROPs)
FBIO Mask : 0x37 (5 FBIO Partitions)
FBIO Shift Mask: 0x00
XP Mask : 0x03 (2 3gio Pads)
Nvdec Mask : 0x01 (1 engine)
Nvenc Mask : 0x01 (1 engine)
Nvjpg Mask : 0x00 (0 engines)
Ofa Mask : 0x01 (1 engine)
PCE Mask : 0x30 (2 PCEs)
Syspipe Mask : 0x01 (1 syspipe)
Gpu Temp : 42 deg C
PEX Rx Lanes : 0xffff
PEX Tx Lanes : 0xffff
PEX Det. Lanes : 0xffff
PEX Width, ASLM: 16 lanes, Not Supported
PEX Link Speed : 16.0 Gbit/s
PEX BandWidth : 256.0 Gbit/s
ASPM, ASPM-CYA : (Disabled, Disabled)
ASPM L1SS, CYA : (Disabled, L1.1/L1.2)
LTR : Enabled
Chipset
VID : 8086 (Intel)
Chipset DID : 4648 (Unknown)
Chipset ASPM : Disabled
ASPM L1 SS : Disabled
Chipset LTR : Enabled
RM Version : rel/gpu_drv/r455/r455_00-263
testlist.js : 6
resumehandler.js: 1
gputest.js : 30
oqa.spc : 2
boards.js : 1
Running test(s) on GPU 0 [01:00.0] (DID: 0x2206)
Enter SetPState (test 0) Thu May 18 02:22:03 2023
Switched to PState 0 (0.max). Pcie Speed=16000, x16
ClkM = 9502.00 MHz
ClkHost = 1350.00 MHz
ClkDisp = 1350.00 MHz
ClkGpc = 2010.00 MHz
ClkXbar = 1905.00 MHz
ClkSys = 1695.00 MHz
ClkHub = 810.00 MHz
ClkPwr = 540.00 MHz
ClkNvd = 1770.00 MHz
ClkPexGen = 4.00
NVVDD = 1081.25 mV
MSVDD = 1081.25 mV
Exit 000000000000 : SetPState (test 0) ok [0.024 seconds]
Enter CuvidNVDEC (test 360) Thu May 18 02:22:03 2023
ERROR: cuvidCreateVideoSource returned error 300 (CUDA_ERROR_INVALID_SOURCE: device kernel image is invalid)
ERROR: Engine 0: No testing was done
ERROR: software error in engineIdx 0
Exit 020000360002 : CuvidNVDEC (test 360) software error [0.246 seconds]
Error!
GPU tests completed.
Failure(s) :
LOOP TEST CODE MESSAGE
---- ------------------------ ------------ ---------------------------
1 CuvidNVDEC 020000360002 software error
Error Code = 020000360002 (software error )
####### #### ######## ###
####### ###### ######## ###
## ## ## ## ###
## ## ## ## ###
####### ######## ## ###
####### ######## ## ###
## ## ## ## ###
## ## ## ######## ########
## ## ## ######## ########
MODS end : Thu May 18 02:22:04 2023 [3.556 seconds (00:00:03.556 h:m:s)]
MODS start: Thu May 18 02:22:01 2023
Command Line : gputest.js -oqa -run_on_error -test 360
CPU
Arch : x86_64
Name : 12th Gen Intel(R) Core(TM) i5-12600KF
Cores : 16
Version
MODS : 455.120
System
OperatingSystem: Linux (x86_64)
Kernel : 5.9.1-gentoo-x86_64
KernelDriver : 4.00
SBIOS Version : F4a
SBIOS Date : 04/14/2023
HostName : tinylinux
Available RAM : 31558/32082 MB (Free/Size)
Sys-uuid :
HDD-Serno :
GPU 0 [01:00.0] dev.sub 0.0
----------------------------------------
DevInst : 0
PCI Location : 0x00, 0x01, 0x00, 0x00
GPU DID : 0x2206
PDI : 0x8f0cc023bc50bd40
Raw ECID : 0x002001c00000005d78c305c4
Raw ECID (GHS) : 0x4705d78c30400000007000040
ECID : SAMSUNG-SNNZ31-07_x0_y1
Device Id : GA102
Revision : a1
Sub Revision : 1
NV Base : 0x52000000
FB Base : 0x40000000
IRQ : 16
WARNING: GPU 0 [01:00.0] PCIE speed capability (16000Mbps) higher than down stream port link speed (2500Mbps)
UUID : 7d04fc8e-6b25-629b-c548-3bd2c3e7e337
Platform : Hardware
Foundry : Samsung
Subsystem VID : 0x1462
Subsystem DID : 0x3896
Board ID : 0x023f
Project : G132-0030
Fuse File Fmt : JSON
Display : 0x00000400 (id)
SBIOS Init : Primary
Native Mode : 1920x1080
Memory Size : 10240 MB
FB Vendor : Micron
RAM Protocol : GDDR6X
RAM Config : 2
WARNING: WAR(2791626): 'Num Row Remaps' reporting is disabled
ROM Version : 94.02.26.40.99
ROM Type : Partner Production
ROM OEM Vendor : NVIDIA
ROM Partner : msi
ROM Project ID : 110182
ROM Timestamp : 2020-5-22 06:04:13
ROM Expiration : 2020-11-18 06:04:13
PState (mode) : 8 5 3 2 [0]
PState Version : 4.0
EDC : (RD,WR,REPLAY)
GPC Clock : 2010.000 MHz DEFAULT
DRAM Clock : 9501.996 MHz DEFAULT
Host Clock : 1350.000 MHz NAFLL
XBar Clock : 1905.000 MHz DEFAULT
Sys Clock : 1695.000 MHz DEFAULT
Power Clock : 540.000 MHz NAFLL
NVDec Clock : 1770.000 MHz NAFLL
Display Clock : 1350.000 MHz DEFAULT
NVVDD : 1081.25 mV
MSVDD : 1081.25 mV
GPC Mask : 0x3f (6 GPCs)
TPC Mask : [x 3f 3f 3f 3f 3e 3e] (34 TPCs)
FB Mask : 0x37 (5 FB Partitions)
L2 Mask : [3 3 x 3 3 3] (10 L2s)
L2 Slice Mask : [ff ff x ff ff ff] (40 L2 Slices)
PES Mask : [x 7 7 7 7 7 7] (18 PESes)
ROP Mask : [x 3 3 3 3 3 3] (12 ROPs)
FBIO Mask : 0x37 (5 FBIO Partitions)
FBIO Shift Mask: 0x00
XP Mask : 0x03 (2 3gio Pads)
Nvdec Mask : 0x01 (1 engine)
Nvenc Mask : 0x01 (1 engine)
Nvjpg Mask : 0x00 (0 engines)
Ofa Mask : 0x01 (1 engine)
PCE Mask : 0x30 (2 PCEs)
Syspipe Mask : 0x01 (1 syspipe)
Gpu Temp : 42 deg C
PEX Rx Lanes : 0xffff
PEX Tx Lanes : 0xffff
PEX Det. Lanes : 0xffff
PEX Width, ASLM: 16 lanes, Not Supported
PEX Link Speed : 16.0 Gbit/s
PEX BandWidth : 256.0 Gbit/s
ASPM, ASPM-CYA : (Disabled, Disabled)
ASPM L1SS, CYA : (Disabled, L1.1/L1.2)
LTR : Enabled
Chipset
VID : 8086 (Intel)
Chipset DID : 4648 (Unknown)
Chipset ASPM : Disabled
ASPM L1 SS : Disabled
Chipset LTR : Enabled
RM Version : rel/gpu_drv/r455/r455_00-263
testlist.js : 6
resumehandler.js: 1
gputest.js : 30
oqa.spc : 2
boards.js : 1
Running test(s) on GPU 0 [01:00.0] (DID: 0x2206)
Enter SetPState (test 0) Thu May 18 02:22:03 2023
Switched to PState 0 (0.max). Pcie Speed=16000, x16
ClkM = 9502.00 MHz
ClkHost = 1350.00 MHz
ClkDisp = 1350.00 MHz
ClkGpc = 2010.00 MHz
ClkXbar = 1905.00 MHz
ClkSys = 1695.00 MHz
ClkHub = 810.00 MHz
ClkPwr = 540.00 MHz
ClkNvd = 1770.00 MHz
ClkPexGen = 4.00
NVVDD = 1081.25 mV
MSVDD = 1081.25 mV
Exit 000000000000 : SetPState (test 0) ok [0.024 seconds]
Enter CuvidNVDEC (test 360) Thu May 18 02:22:03 2023
ERROR: cuvidCreateVideoSource returned error 300 (CUDA_ERROR_INVALID_SOURCE: device kernel image is invalid)
ERROR: Engine 0: No testing was done
ERROR: software error in engineIdx 0
Exit 020000360002 : CuvidNVDEC (test 360) software error [0.246 seconds]
Error!
GPU tests completed.
Failure(s) :
LOOP TEST CODE MESSAGE
---- ------------------------ ------------ ---------------------------
1 CuvidNVDEC 020000360002 software error
Error Code = 020000360002 (software error )
####### #### ######## ###
####### ###### ######## ###
## ## ## ## ###
## ## ## ## ###
####### ######## ## ###
####### ######## ## ###
## ## ## ## ###
## ## ## ######## ########
## ## ## ######## ########
MODS end : Thu May 18 02:22:04 2023 [3.556 seconds (00:00:03.556 h:m:s)]
Сначала хочу решить, буду ли я оставлять карту. В этом плане мне важна перспектива. Ждать полного выхода из строя контроллера памяти желания нету. С другой стороны, вряд-ли я буду над ним "издеваться". В общем буду очень благодарен, за любую помощь.