bug check reboots

[ReadyMan]

Got a mystery problem on my hands:
Lately my PC has been hanging.
Specifically, when I'm playing a game (Red Solstice 2), it will stop, and the system will go black, requiring me to do a hard reset with the power button. Once the music continued to play even tho the monitors and everything else were not working. The system lights are still on, but not the kb/mouse lights.
Yesterday I played for a couple of hours with no problems, and then when I exited the game, it froze, but at the desktop. I brought up the control panel with cntrl/alt/del and was looking at the task manager when it then completely froze, requiring a reboot.

I tried doing a memory check, which found no errors but upon restarting, it kept looping with the restart until I did a hard reset with the power button and since then its run fine.
I tried updating the gfx driver to the latest, but still same issue
Under regular use (not gaming) I havent really had an issue, so I thought it was the game. I tried lowering settings, but still no joy.

The event view window logs show the same error each time there's a reboot.
Here it is:

The computer has rebooted from a bugcheck. The bugcheck was: 0x00000116 (0xffffb605613ea010, 0xfffff8007bc89b3c, 0xffffffffc000009a, 0x0000000000000004). A dump was saved in: C:\WINDOWS\MEMORY.DMP. Report Id: 08c11bf1-e255-4966-b780-36e6f824d8ec.
Event Xml:
<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
<System>
<Provider Name="Microsoft-Windows-WER-SystemErrorReporting" Guid="{ABCE23E7-DE45-4366-8631-84FA6C525952}" EventSourceName="BugCheck" />
<EventID Qualifiers="16384">1001</EventID>
<Version>0</Version>
<Level>2</Level>
<Task>0</Task>
<Opcode>0</Opcode>
<Keywords>0x80000000000000</Keywords>
<TimeCreated SystemTime="2021-09-14T02:35:34.3081484Z" />
<EventRecordID>33375</EventRecordID>
<Correlation />
<Execution ProcessID="0" ThreadID="0" />
<Channel>System</Channel>
<Computer>ReadyMan</Computer>
<Security />
</System>
<EventData>
<Data Name="param1">0x00000116 (0xffffb605613ea010, 0xfffff8007bc89b3c, 0xffffffffc000009a, 0x0000000000000004)</Data>
<Data Name="param2">C:\WINDOWS\MEMORY.DMP</Data>
<Data Name="param3">08c11bf1-e255-4966-b780-36e6f824d8ec</Data>
</EventData>
</Event>

I havent had issues with any other game (aoe2 or the aoe4 beta, with all settings on max) or at any other time, other than the desktop freeze mentioned above.
Oddly, I can play this same game with no issues sometimes, it's just every few days that it locks up.

Every !error marked in red in the event log is this bugcheck 0x00000116

Here's my specs:
I9-9900 @3.6
32 GB RAM
windows 10Pro
RTX 2080 TI


I'm hoping someone can help me.

Thanks!

RM

[Krom]

A dump was saved in: C:\WINDOWS\MEMORY.DMP.

Give me this file and I can probably narrow it down a little (compress it with 7zip first if it is big).

Although a quick google on 0x00000116 says its a TDR timeout (your video card had a hardware failure and couldn't be reset).

[ReadyMan]

Thanks Krom!

Running into a snag:
I zipped it, but it's still too big (313kb).

Tried to open it, but I dont have permission for some reason, and couldnt add myself to the permission list for some reason.
Do you have a place I can upload it (and if I did, could you read/open it?)

[Krom]

You could probably share it using google drive.

[ReadyMan]

I uploaded it to google drive.
tried to pm you, but you dont have anything listed

[admin edit: removed link]

[Krom]

I sent you a PM, check it.

[ReadyMan]

Sent you a pm back.

I'm wondering if it's my gpu fan, perhaps the card is overheating?
Yesterday I played for a few hours, and upped the fan speed, and had no issues....

[Krom]

Got it, as I suspected TDR fault in the nvidia display driver. Basically your GPU blew up, it could be heat, it could be a driver or game issue with a specific game, it could be a power issue. Though considering you didn't get any crashes to desktop with the typical "display driver nvlddmkm suffered an error and has recovered" it likely means the GPU locked up hard on the first go which isn't something that typically has a software cause. Backing off any factory overclock might sort it out, or just keeping it cooler with higher fan speeds, or going into the nvidia control panel and turning on debug mode (in the help menu). The problem is all of those fixes take some time to validate if they are actually working (you have to run the offending game for hours/days). Or find some other game/stress test that more quickly brings out the instability.

See also:
https://docs.microsoft.com/en-us/window ... dr-failure

And here is the analysis of the memory dump:

Code: Select all

Microsoft (R) Windows Debugger Version 10.0.17763.132 AMD64
Copyright (c) Microsoft Corporation. All rights reserved.


Loading Dump File [F:\Downloads\Torrents\MEMORYdumpfile\MEMORYdumpfile.DMP]
Kernel Bitmap Dump File: Kernel address space is available, User address space may not be available.


************* Path validation summary **************
Response                         Time (ms)     Location
Deferred                                       SRV*https://driver-symbols.nvidia.com/
Deferred                                       SRV*c:\symbols*http://msdl.microsoft.com/download/symbols
Symbol search path is: SRV*https://driver-symbols.nvidia.com/;SRV*c:\symbols*http://msdl.microsoft.com/download/symbols
Executable search path is: 
Windows 10 Kernel Version 19041 MP (16 procs) Free x64
Product: WinNt, suite: TerminalServer SingleUserTS
Built by: 19041.1.amd64fre.vb_release.191206-1406
Machine Name:
Kernel base = 0xfffff800`5b600000 PsLoadedModuleList = 0xfffff800`5c22a230
Debug session time: Mon Sep 13 21:34:31.573 2021 (UTC - 5:00)
System Uptime: 0 days 3:07:25.280
Loading Kernel Symbols
...............................................................
................................................................
................................................................
..........................
Loading User Symbols

Loading unloaded module list
.............
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck 116, {ffffb605613ea010, fffff8007bc89b3c, ffffffffc000009a, 4}

Page 7cd7a2 not present in the dump file. Type ".hh dbgerr004" for details
Page 7cd7a2 not present in the dump file. Type ".hh dbgerr004" for details
*** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys
Probably caused by : nvlddmkm.sys ( nvlddmkm+dc9b3c )

Followup:     MachineOwner
---------

2: kd> !analyze -v
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

VIDEO_TDR_FAILURE (116)
Attempt to reset the display driver and recover from timeout failed.
Arguments:
Arg1: ffffb605613ea010, Optional pointer to internal TDR recovery context (TDR_RECOVERY_CONTEXT).
Arg2: fffff8007bc89b3c, The pointer into responsible device driver module (e.g. owner tag).
Arg3: ffffffffc000009a, Optional error code (NTSTATUS) of the last failed operation.
Arg4: 0000000000000004, Optional internal context dependent data.

Debugging Details:
------------------


KEY_VALUES_STRING: 1


STACKHASH_ANALYSIS: 1

TIMELINE_ANALYSIS: 1


DUMP_CLASS: 1

DUMP_QUALIFIER: 401

BUILD_VERSION_STRING:  19041.1.amd64fre.vb_release.191206-1406

SYSTEM_MANUFACTURER:  ASUSTeK COMPUTER INC.

SYSTEM_PRODUCT_NAME:  ROG Strix GL12CX

SYSTEM_VERSION:  1.0       

BIOS_VENDOR:  American Megatrends Inc.

BIOS_VERSION:  GL12CX.307

BIOS_DATE:  07/05/2019

BASEBOARD_MANUFACTURER:  ASUSTeK COMPUTER INC.

BASEBOARD_PRODUCT:  GL12CX

BASEBOARD_VERSION:  1.0       

DUMP_TYPE:  1

BUGCHECK_P1: ffffb605613ea010

BUGCHECK_P2: fffff8007bc89b3c

BUGCHECK_P3: ffffffffc000009a

BUGCHECK_P4: 4

FAULTING_IP: 
nvlddmkm+dc9b3c
fffff800`7bc89b3c 48ff259d16b7ff  jmp     qword ptr [nvlddmkm+0x93b1e0 (fffff800`7b7fb1e0)]

DEFAULT_BUCKET_ID:  GRAPHICS_DRIVER_TDR_FAULT

CPU_COUNT: 10

CPU_MHZ: e10

CPU_VENDOR:  GenuineIntel

CPU_FAMILY: 6

CPU_MODEL: 9e

CPU_STEPPING: d

CPU_MICROCODE: 6,9e,d,0 (F,M,S,R)  SIG: B8'00000000 (cache) B8'00000000 (init)

BLACKBOXBSD: 1 (!blackboxbsd)


BLACKBOXPNP: 1 (!blackboxpnp)


BUGCHECK_STR:  0x116

PROCESS_NAME:  System

CURRENT_IRQL:  0

ANALYSIS_SESSION_HOST:  NATE

ANALYSIS_SESSION_TIME:  09-15-2021 16:56:37.0009

ANALYSIS_VERSION: 10.0.17763.132 amd64fre

STACK_TEXT:  
ffff9c01`e1ee7918 fffff800`7666665e : 00000000`00000116 ffffb605`613ea010 fffff800`7bc89b3c ffffffff`c000009a : nt!KeBugCheckEx
ffff9c01`e1ee7920 fffff800`76616e44 : fffff800`7bc89b3c ffffb605`56daa8a0 00000000`00002000 ffffb605`56daa960 : dxgkrnl!TdrBugcheckOnTimeout+0xfe
ffff9c01`e1ee7960 fffff800`7660f97c : ffffb605`58ea7000 00000000`01000000 00000000`00000004 00000000`00000004 : dxgkrnl!ADAPTER_RENDER::Reset+0x174
ffff9c01`e1ee7990 fffff800`76665d85 : 00000000`00000100 ffffb605`58ea7a70 00000000`57679e70 fffff800`5b8cf3fc : dxgkrnl!DXGADAPTER::Reset+0x4dc
ffff9c01`e1ee7a10 fffff800`76665ef7 : fffff800`5c324440 ffffb605`672c2d70 00000000`00000000 00000000`00000000 : dxgkrnl!TdrResetFromTimeout+0x15
ffff9c01`e1ee7a40 fffff800`5b8b8505 : ffffb605`5b68d040 fffff800`76665ed0 ffffb605`3b718720 ffffb605`00000000 : dxgkrnl!TdrResetFromTimeoutWorkItem+0x27
ffff9c01`e1ee7a70 fffff800`5b955845 : ffffb605`5b68d040 00000000`00000080 ffffb605`3b702080 000fa4ef`bd9bbfff : nt!ExpWorkerThread+0x105
ffff9c01`e1ee7b10 fffff800`5b9fe808 : fffff800`59524180 ffffb605`5b68d040 fffff800`5b9557f0 00000000`00000000 : nt!PspSystemThreadStartup+0x55
ffff9c01`e1ee7b60 00000000`00000000 : ffff9c01`e1ee8000 ffff9c01`e1ee1000 00000000`00000000 00000000`00000000 : nt!KiStartSystemThread+0x28


THREAD_SHA1_HASH_MOD_FUNC:  56a150c818aff348d17e35fc9e822916b6c0190b

THREAD_SHA1_HASH_MOD_FUNC_OFFSET:  566318dfd710cc1e0e9f44288bdfb7068acf4ad7

THREAD_SHA1_HASH_MOD:  3225689ea8bf23dca00cdfd517975e0703e07271

FOLLOWUP_IP: 
nvlddmkm+dc9b3c
fffff800`7bc89b3c 48ff259d16b7ff  jmp     qword ptr [nvlddmkm+0x93b1e0 (fffff800`7b7fb1e0)]

FAULT_INSTR_CODE:  9d25ff48

SYMBOL_NAME:  nvlddmkm+dc9b3c

FOLLOWUP_NAME:  MachineOwner

MODULE_NAME: nvlddmkm

IMAGE_NAME:  nvlddmkm.sys

DEBUG_FLR_IMAGE_TIMESTAMP:  6128de4f

STACK_COMMAND:  .thread ; .cxr ; kb

FAILURE_BUCKET_ID:  0x116_IMAGE_nvlddmkm.sys

BUCKET_ID:  0x116_IMAGE_nvlddmkm.sys

PRIMARY_PROBLEM_CLASS:  0x116_IMAGE_nvlddmkm.sys

TARGET_TIME:  2021-09-14T02:34:31.000Z

OSBUILD:  19041

OSSERVICEPACK:  0

SERVICEPACK_NUMBER: 0

OS_REVISION: 0

SUITE_MASK:  272

PRODUCT_TYPE:  1

OSPLATFORM_TYPE:  x64

OSNAME:  Windows 10

OSEDITION:  Windows 10 WinNt TerminalServer SingleUserTS

OS_LOCALE:  

USER_LCID:  0

OSBUILD_TIMESTAMP:  unknown_date

BUILDDATESTAMP_STR:  191206-1406

BUILDLAB_STR:  vb_release

BUILDOSVER_STR:  10.0.19041.1.amd64fre.vb_release.191206-1406

ANALYSIS_SESSION_ELAPSED_TIME:  b3a

ANALYSIS_SOURCE:  KM

FAILURE_ID_HASH_STRING:  km:0x116_image_nvlddmkm.sys

FAILURE_ID_HASH:  {c89bfe8c-ed39-f658-ef27-f2898997fdbd}

Followup:     MachineOwner
---------

2: kd> lmvm nvlddmkm
Browse full module list
start             end                 module name
fffff800`7aec0000 fffff800`7d325000   nvlddmkm   (no symbols)           
    Loaded symbol image file: nvlddmkm.sys
    Image path: \SystemRoot\System32\DriverStore\FileRepository\nvaki.inf_amd64_177579bbd21968a8\nvlddmkm.sys
    Image name: nvlddmkm.sys
    Browse all global symbols  functions  data
    Timestamp:        Fri Aug 27 07:45:03 2021 (6128DE4F)
    CheckSum:         023BD732
    ImageSize:        02465000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4
    Information from resource tables:

[ReadyMan]

Thanks Krom!

I will increase my fan speed and see how that goes.
I had success with it yesterday, and will be gaming a bit this week, so will keep my eye on it.
Really glad to narrow down the problem....hadnt had any issues with the RTSs that I usually play (even the AoE4 beta, which I played a lot of) until I got RS2.

I REALLY appreciate your help!
(if the card is bad, it'll be tough finding a replacement. Hopefully the chip shortage gets sorted out. I do have an old nvidia card packed away just in case of card failure,
so at least I'll have something if the card totally dies.)

[Krom]

If the higher fan speed fixes it, perhaps a quieter and more permanent solution would be to install MSI afterburner and just apply a negative clock offset of like 25-50 MHz or so, it would be barely relevant to performance but would probably be enough to pull the GPU back to a stable state for good.

Also just making sure the cooler is operating well and is free of excessive dust and debris, if that isn't the problem and you can still see unusually high core temperatures or excessive throttling then its possible the thermal paste application on the GPU heat sink is dried out and failing. In these days of obnoxious shrouds, back plates and RGB, pulling off a cooler to replace the thermal compound can be quite the ordeal, but if the paste is getting so bad you can't use the card anymore then its worth it.

[ReadyMan]

Thanks!
Good idea on cleaning out the case/fans/grills.
I'llkeep the GPU monitor open and watch for the temps.

I'll post back if I have any further issues.
(checked on amazon, and most of the 2080 cards are $1000 or more...even my old 1070 which I have boxed away is going for almost $500 on craigslist lol!)

[Krom]

Yeah, the pandemic + bitcoin (etherium actually) has made GPU prices insane. Just last week I finally got a queue spot notification from EVGA that I signed up for in early June and bought one of their FTW3 Ultra 3080 Ti 12 GB cards, but it cost almost $1500 (which being MSRP is still a lot better than the ebay price for the same card which would be well over $2000). Now I just need to get the water block for it because it is pretty noisy, which is to be expected from a card that regularly pulls 380w.

For a while there early this year I probably could have sold my GTX 1080 for more than I paid for it even though it just passed 5 years old this month.

[ReadyMan]

Cleaned out the case/fans etc....there wasnt much dust buildup.
Been monitoring the GPU temps and running the fans at 50%, which keeps things pretty cool.
Interestingly, just a few minutes ago, after gaming for an hour or so, with no issues, I quit all programs except my browser and email and got breakfast.
When I came back the system had rebooted.
Same 0x0000016 error, but a minute later had two ACPI errors:

Description:
AMLI: ACPI BIOS is attempting to write to an illegal IO port address (0x74), which lies in the 0x74 - 0x76 protected address range. This could lead to system instability. Please contact your system vendor for technical assistance.

and the 2nd one:

Description:
AMLI: ACPI BIOS is attempting to read from an illegal IO port address (0x75), which lies in the 0x74 - 0x76 protected address range. This could lead to system instability. Please contact your system vendor for technical assistance.


Nothing after that.

I gamed for a number of hours yesterday with no issues...

[ReadyMan]

Just an update:

Things seem to have settled down. I run the fans at 60% to 70% with no issues.
Every once in a while the system will reboot, then I just crank up the fans again, and no issues.
Guess it's the heatsink, or failing card.

I'd buy a new card, but seems rediculous right now, with the high prices, since I have a work around. (I have a 1070 sitting around that I could use if I needed)

[ReadyMan]

2nd update:

Running the fans at 60%-70% speed and about 1-2 reboots a week now.
Oddly, it's usually not when I'm running AoE4, but sometimes after, or just when the system has been
running all day (email/office/web) with little to no intensive programs running.
The bugcheck is always 0x00000116 (so graphics card related).
I see Errors in the event viewer system from time to time, usually something about nvidia, or a dump file creation failed

The card runs cool for the most part (47 C) during the day, and will hit the 80s while gaming.

The fan speed seems to be working alright, but I'm annoyed with the noise and the shutdowns.
I'm running the debug check in nvidia control panel now (per Krom's advice) with fans at normal speed. Gonna try this to see what happens.

Been looking for card upgrades. I have to find one that fits my case (it's a prebuilt system ASUS - ROG Strix GL12CX Gaming Desktop).
Been saving for a new card, and was going to wait, but I can make the purchase now if I can find a good card...
So...any suggestions? Running a 2070ti with 11gb now. Been looking at the 3070ti card with 8gb, seems like the 3080s are $500 more...and not sure if it's worth it?
Thanks for any advice!

--Edit--I've got a 2080Ti not a 2070Ti

[Krom]

Only the 3080Ti or 3090 is a meaningful upgrade from a 2080Ti right now and GPU prices remain insane because of cryptocurrency. It takes extraordinary effort to to buy a GPU right now and the shortage is expected to continue well into 2022. So basically I would highly recommend doing everything in your power to keep your existing 2080 Ti functional. Think about increasing the voltage, decreasing the clocks on both the core and memory by small amounts to see if there is a way to stabilize it.

[ReadyMan]

Good points. Thanks Krom!
Yes, a 3080 is going to be $2500 (I can actually get a new system for about $3000 with a 3080 included)
That's just crazy. I dont want to upgrade anything, as I'm happy with the system as it is.

Next step is to continue the debug mode with Nvidia control panel. (upgraded to win 11, and the newest nvidia drivers yesterday)
If that doesnt help, I'll drop my 1070 in and see if the system keeps hanging.
Then I'll start tweaking the voltage and clock speeds, and maybe try resetting the core?? lol....hate to do that, but if it's not working, its junk anyway
Last step I guess will be a new card (or new pc).

It doesnt seem like these prices can continue, as where there's money, a retailer will find a way to meet the need...

[ReadyMan]

Thought I'd post another update.
Been running the card fans at 50% with only the occasional lockup (every couple of weeks, but very inconsistant....sometimes it will lock right after startup).
It's pretty much a non issue actually.
2 days ago though, the screen started to flick on and off at startup, did it a few times, then started up no problem.
It did this a couple of times, but not every startup.

I was determined to power thru the pandemic with this card, until prices dropped.
I looked on amazon today and saw that the 3080 ti cards are about $1000-$1200, and the 3090s are about $1500.
I can afford one of these cards, and was thinking to put this current card and its annoyance behind me.
Any recommendations on current cards if I'm going to bite the bullet for the upgrade? (I have a small tower, so it'll have to be only 12"-13" or so long)

[Krom]

Since you live in the USA, EVGA is the most sensible option because of their vastly superior warranty and service compared to most other brands operating in the US. I have a EVGA 3080 Ti FTW3 Ultra that I put a full cover water block on, it was okay in stock configuration but a little noisy for my tastes.

Also keep in mind the 4000 series are likely right around the corner, so if you buy EVGA you could potentially take advantage of the step-up program: https://www.evga.com/support/stepup/

Also the monitor flickering at startup could be a problem with the monitor and not the video card (bad capacitors in the power board causing the back light to flicker until it warms up/stabilizes).

[TigerRaptor]

Krom with all the knowledge you possess, your head must be the size of a pumpkin by now. Stay inside for Halloween, we need to keep you around.

[Krom]

Heh, I'll think about it.

On the subject of video cards, these days you can sometimes snap up a 3000 series card for under the original MSRP if you catch the right deal. The crypto miners quit buying abruptly so now the supply side flooded and even lots of gamers are holding off on potential recession and or the 4000 series launching right around the corner. Basically if there is a 3000 series card that will keep you happy for the next 2-3 years, now is a good time to start looking.

Also because of the latest crypto bust, lots of video card vendors are getting trashed economically, EVGA for instance had to do a bunch of layoffs. Some other OEMs might even go bankrupt, just like the last time the crypto bubble popped. Lots of the GPU vendors were selling directly to crypto mining farms so a huge volume of 3000 series cards never even made it to the open market (and many that did make it were scalped). The volatility of crypto eventually hurts everyone and some people never learn.

[Tunnelcat]

I'm still running a GTX 2070 on my gaming rig. With prices dropping and the new cards coming around the corner, should I upgrade to a 3000 series Nvidia card NOW or wait until the new cards are eventually released? If so, who's got a decent product out there right now? I've bought EVGA in the past, but they've had some quality issues with their 3000 series lately. The 2070 in my rig (not EVGA) is still holding it's own, but it's noisy at full throttle and I might upgrade to something with better cooling and less noise.

[Krom]

There is only one way to make a >300w GPU (all the high end 3000 series cards) quiet and that is what I did with mine: full cover water block.

[ReadyMan]

Another update:
Well I've squeeked another year out of this 2080. Have the GPU fan set at 50% all the time. Still have the occasional lockups, sometimes while gaming, sometimes just using zoom.
Not getting the same error with the crashes in Event viewer....usually a bugcheck like this one which happened a couple hours ago.

Description:
The computer has rebooted from a bugcheck. The bugcheck was: 0x00000116 (0xffffe68f1bb80010, 0xfffff80183ee3d00, 0xffffffffc000009a, 0x0000000000000004). A dump was saved in: C:\WINDOWS\MEMORY.DMP. Report Id: 3fda2081-017f-46b9-b432-bd02df7030ea.


Not sure, but I think this is the same graphics error?
Either way, I'm ready for a new card. The 4080s are released, but dont see a lot of them.
The problem is my small case. The existing card is about 11", and I think there's a bit more room in the case for just under 12" of card.
The 4080s I've seen on amazon are too long.
So thinking of going for a 3090ti (I have $1500 saved for this card)....which are small enough in length, but paying almost as much as the msrp of the 4080

Disappointed in myself for not getting a bigger case (this was a pre-built pc from bestbuy at the start of the pandemic) so that card length wouldnt dictate the card options....

[Krom]

The 4080 is horribly priced for what it is, like it is so bad it makes the 4090 look reasonable.

A used 3080 or greater card may be a better option, for significantly lower cost. Although I have heard of 3090 ti cards going for as low as $900 new now that crypto crashed. Certainly don't pay $1500 for one when some patience could get you a 4090 from best buy for $100 more (though you may end up needing a new case and PSU to handle one, but the 3090 ti isn't much different).

Also probably time to send your PSU to be recycled, the 3000 series cards have really high transient loads that can trip over-current protection on PSUs pretty easily, and the 4090 is just plain a power hog even if it doesn't have as severe transients.

[ReadyMan]

Cant seem to find a card anywhere.
Had a couple more lockups and am frustrated with this whole issue.
No one has 4090s, not newegg, not amazon, not bestbuy.
BestBuy has no 3000 series online.
Amazon has a few but the 3080s and 3090s are rediculously priced.
Newegg doesnt have much of an inventory at all

Opened up my case and my current 2080 is 11" long and 4" wide. The case has enough room for just under 13" of card.

At this point, I could scrap this whole system and just buy a premade with a decent card already in it....which would be about double the cost of the new video card.

[Tunnelcat]

I wonder what happened. In September, I bought a ASUS Tuff RTX 3080ti and an LED ASUS ROG card support for around $980 total and plenty were in stock at Amazon. Now the price has skyrocketed again to over $1300 and the nice LED support for the card is not available now either. Just wow things are crazy.

[ReadyMan]

Well, after a year and a half of dealing with this video card, and random lockups, I decided to just build a new system.
The Asus pre-built system was great, but upgrading its components was really frustrating (small congested tower, with only enough room for a 11.5" GPU, a 650w PSU <---likely part of the lockup problem, but so packed into the system that I couldnt even get it out of the case...and didnt want to take everything out to do it).
I was able to get a 4080 GPU for $1269. Hunted and hunted for a 4090, but couldnt find one, and the 3090s were $1500+. Unreal.
A comparable pre-built system was about $4500-$5000 (with another HD, a 4090, and a i9 13900 CPU). For almost $2k less, I have a system that is almost as powerful, with the room for expansion if components have issues or I want to upgrade.
Thanks to Krom for helping me thru this whole process!

Here's my new specs:

*LIAN LI LANCOOL II MESH C RGB SNOW WHITE Tempered Glass ATX Case

*ASUS ROG Strix Z690-F Gaming WiFi 6E LGA 1700(Intel® 12th&13th Gen) ATX gaming motherboard(PCIe 5.0, DDR5,16+1 power stages,2.5 Gb LAN,Bluetooth v5.2,Thunderbolt 4)

*CORSAIR Vengeance 32GB (2 x 16GB) 288-Pin PC RAM DDR5 4800 (PC5 38400)

*WD_BLACK SN850X NVMe M.2 2280 2TB PCI-Express 4.0 x4 Internal Solid State Drive (SSD)

*GIGABYTE GAMING OC GeForce RTX 4080 16GB GDDR6X

*Noctua NH-D15S chromax.black, Premium Dual-Tower CPU Cooler with NF-A15 PWM 140mm Fan

*CORSAIR RMx Series (2021) RM1000x CP-9020201-NA 1000 W ATX12V / EPS12V SLI Ready CrossFire Ready 80 PLUS GOLD Certified Full Modular Active PFC Power Supply

I'm just waiting on 3 fans from Amazon to be delivered before putting the system together.
I appreciate any tips or input on assembling the new pc!

RM

[ReadyMan]

Finally got all my parts and am ready to put the build together.
Been watching LOTS of youtube videos on building pcs, and havent really found a good one for my particular mb,
but I think it's all pretty straight forward.
Only think I'm concerned about is connecting the right cables to all the connectors on the mb, and installing the cpu air cooler (not sure if I need an adapter?)....most of the builds on youtube are with water coolers.

Just a couple of questions:
I will have 8 fans connected (3 case fans in front, a noctura 120mm for the exhaust with no rgb, 2 blue 120 light fans on the top, and 2 red 120 fans on the bottom)...should I get a fan hub? Or not needed? Can I daisy chain the fans together (I think the mb has 3 connectors for fans...?)?

Many builds I've seen have the asus hyper m.2 ssd expansion pcie card with a cooler. But I just have a WD sn850x NVMe card. Do I need (or is it good to have) the m.2 cooler card? Or is this only for a raid setup? (The 2TB drive should be plenty for what I need...)
I dont really know much about these newer ssd cards. My last build had the small 2" (or so) ssd, so dont really know the difference.

Thanks for any input!

RM

[Krom]

Heatsinks for m.2 drives are generally not necessary, the drives would only need them under extreme conditions well outside the range of even power users. Ordinary productivity/gaming loads will never even get close to warming them up. Pretty much the only way to make them overheat is to subject them to a benchmark/torture test run for several minutes because that is the only thing that can keep them 100% busy for long enough.

Don't connect more than 3 fans to a single motherboard header.

[ReadyMan]

Don't connect more than 3 fans to a single motherboard header.

Ran into a snag. installed the cpu and cooler, ram, mb, psu, ssd, but then had to pull the mb back out because I couldnt get the cooler fan connector into the mb cpu fan slot.
The psu has a billion cables, but I think I only need the 24 pin atx connector, and 2 8 pin connectors to the mb, and then the one for the gfx card. Does that sound right? Now that I have the mb out of the case, I can wire the psu cables and connect them, so that I wont have to pull the mb again if it's too tight.

Lastly, fan connections? I have 3 front fans that came with the case that need to be plugged in, the exhaust that needs to be plugged in, the 2 top fans that need to be plugged in, and of course the cooler fan (that takes the spot on the mb). I have 2 bottom fans as well, but here's the problem:
How do I connect them all? The psu didnt come with any fan connectors. Do I need to get a fan hub, or buy fan splitters?

I've watched dozens of pc build videos, with my psu and my mb, but cant seem to find the answers.

Been working on this for 5 or so hours, so thought I'd better ask here.

Thanks in advance for any help!

RM

EDIT--
The GPU takes 3 8 pin connectors. I have 2 plugged into the mb, and only have 1 left. However, I have some 6 pin connectors with a 2 pin separated but tied to the 6pin. Can I use these 6 pin connectors and just plug in the other 2 pins, or do I have to get more 8 pin connectors (I might have a few older ones from my last build 8 years ago...if I didnt toss them already).
???

[Krom]

Motherboards usually have quite a few fan headers, for instance mine has 6, check your manual for how many you have.

Those PCIe 6 pin/8 pin connectors are for the GPU, they tie together to become 8 pin connectors. The PSU should be labeled (and it would also say in the manual) which plug is for what.

You absolutely cannot mix and match modular PSU cables from one PSU to another, they are not standard and you can damage or destroy hardware by mixing them. Only use the cables that came with the unit.

[ReadyMan]

Thanks Krom.
The PSU manual didnt say a thing, and the cables say "pci-e".
I took a long and close look at the mb schematics, and it has 7 or so fan slots.
The problem is the wiring to those spots for a couple of my fans.
I'd like to use the cpu opt slot for my exhaust fan (a noctura that I bought specifically for the exhaust), since it is close to that slot
and not long enough to reach the other slots.
I found a couple of old fan splitters (probably 10+ years old) and will use a couple of these to hook up the fans.

In retrospect, I should have gotten a fan hub, as well as some cool colored extension cables for the psu (the cables *barely* make it from the psu to the where they need to go).

[ReadyMan]

Update:
I have the system together, and was just waiting on Amazon to deliver some 3 way fan splitters.
They just arrived, so I should have the system up and running this weekend (the sag bracket from the gigabyte gfx card doesnt work with this LianLi case/Asus MB, so I have to drill into the case to provide support, which I will do once I pull the gfx card so that I can install the fan splitters.
Had to get some help installing the fans/cables, as this MB and case were very complicated (and the manuals didnt help a whole lot). The person helping me said I should have definitely gotten a water cooler instead of the air cooler.

I must say that I have really learned a lot with this build (it's been 9 years since my last build), and I'm enthused to keep modifying this existing build (thanks to Krom, I saved $1500+ on the build!), so any ideas are welcome!

I've got 3 great monitors, so my next upgrade might be LianLi fans/lights, and maybe some faster Ram? (32gb 4800 ddr 5 is what I have now)

Lockups continue on my existing system (IF fans are at 60%....I gamed on this system with fans on 80% for 3 weeks and no lockups...wierd), so not sure what to do with it. Keep it in the closet in case I have an issue with the new PC? Replace the gpu (when the prices drop for the 2080s/3080s, if I can find one that fits the case) and the underpowered 650 PSU?
Not sure. Nice to have an extra system for business in case something borgs on the current build.

[ReadyMan]

Another Update:

System is up and running! Posted first time.
updated the bios
updated the mb drivers thru armory crate
updated gfx drivers
dl and loaded Fan Control
only my gfx fans, and noctura exhaust fan are identifiable.
The 3 front fans that came with the LianLi case, and the 2 top fans, 2 bottom fans, and CPU fan are not noticed by Fan Control.

All fans are spinning (not sure about the cpu fan, as I cant see the fan itself...cpu temps are high running at low load 50c or so, my current system runs at 41c or so), but I cant ramp them up.

So yeah, this is lame, but I really would like to have control of all the fans. So what are my options?
Get a fan controller and better RGB fans? Suggestions?

I installed a 3 way fan splitter, so that explains one of the lower fans and one of the front case fans.

I havent done a stress test yet, will do that tomorrow.
Running both systems concurrently, and will migrate from current pc to new pc as time allows.

Anything else I should update besides bios, gpu, and mb drivers?
I watched a YT video on after build updates, but much of it was outdated.

RM

[fliptw]

Hello fellow modern ASUS MB owner!

Fan control and whatever BS armoury crate installs are probably fighting... That is holy BS, disable auto-installation from the bios and purge it from your system. Fan control and open RGB for your led strips is much lighter.

I'd just use the fan utility(qfan) in the bios myself.

that being said, control on fans plugged into the MB is per fan header, not per fan. so the fans on the splitter would be treated as one.

[Krom]

I handle all my fans directly in my BIOS fan control, but my system is a little different from normal in a way that makes it really simple: I have 7 fans and 1 pump on 3 channels. All 6 radiator fans are on a single channel (meaning the system sees them all as ONE fan, this is fine because they are all the same make and model), the pump is on its own channel, and the case fan is on its own channel. They are all controlled by a temperature probe I have set up in my water cooling loop with very simple linear speed curves because the coolant itself smooths out all the variability. They don't even come out of totally silent idle speeds till the coolant is over like 38C, the warmest the coolant will even reach is about 42-44C depending almost entirely on the temperature in the room, if the room is cold then it will level off at even lower than that.

Be aware newer/modern CPUs and GPUs run HOT, like do not be the least bit surprised if it spikes to over 90C in an instant, GPUs will also boost to around 70-80C (I believe the default throttle point on most NV GPUs is 83C). This is not harmful to the chips, they will all throttle before they get *too* hot. The thing you need to watch to make sure your cooling is working properly is a good CPU benchmark run that maxes out all the cores so you can make sure they aren't significantly throttling. I would suggest cinebench multi-threaded as a good starting point, its free, doesn't have to be installed and only takes a couple minutes to run, if it goes all the way through with performance within a few hundred points of benchmarks online of the same CPU then your cooling is working fine.

[ReadyMan]

Thanks for the help and suggestions.
The system is up and running. Had to drill into the case to put in a support for the gfx card sag bracket.
Have done a number of stress tests, and the system runs great.
CPU is at 38 degrees when idling, pumps to high 60's while gaming.
GPU is at 43 degrees at idle, pushing high 60's while gaming.

I set up the Armory Crate to do an Auto fan (AC gives me a quick and easy temp read), but will look for something similar in the bios.