Spoonii

Everyone is talking about poor case airflow like it is the end all and be all of overheating issues. What software are you using to monitor temps? different software produce different readings. Try different software and see what the readings say. According to HWmonitor, my CPU is ice cold. Yeah right... If you still think it is a overheating issue try this: open the side panel so the components can breathe and use the PC as you normally would. If the temps don't improve, point a desk fan in there and try again. If the temps still don't improve its another issue. Check the following. What software is used to monitor temps? (diff temps reported by diff software) Is it just incorrect reporting of temp numbers? Do you have an overclock running? multi-core enhancement (or whatever brand equivalent name) still counts as an OC. Thermal paste application, is there enough? Mounting pressure of cooler - how much? hopefully not too much. Where is the radiator mounted? a top mount is best, a front mount is ok, a bottom mount is a no no. Case fan and cooler fan orientation, intake/exhaust does matter. Case fans and CPU cooler plugged into right headers, different RPMs for each, refer to your mobo manual. What type of fans are mounted? Static pressure vs High flow matters BIOS fan curve - what is it set to? Check for fan control software that is conflicting with the bios fan control

Update. The underclock dosen't 100% solve the problem. It only delays the issue by a several hours but it will eventually crash again if you play a game long enough. Go figure.

I stumbled across a work around by lowering the power target and CLK to try and get it stable. Underclocking the card to offset Nvidia GPU Boost 3.0 so it doesn't go >1900Mhz stopped the errors. Apparently the 3000 series conk out when when the core CLK hits 1950~2000Mhz. A quick google search seems to indicate it affects 3060/70/80/90s with monitors shutting off and black screens. This is an RMA job for sure.

Well never mind, solved it myself.

The 3070, 1080ti, and 980ti all at the same time? Very unlikely.

replaced it with 2 new DP 1.4 VESA certified cables. No change.

Computer goes to black screen then monitor switches off with no signal, sometimes the game sounds stop but usually if there is music or a video playing I will still hear it. Only seems to happen when playing games with graphics turned up to high/max, extremely rare to crash when on low/min graphics. CPU: Item i9 7920X w/ Corsair h150i Pro (6 fans Push/Pull) MOBO: Rampage 6 Extreme RAM: G.Skill 8x8Gb 3200Mhz CL14 GPU: ASUS Rog Strix 3070 PSU: EVGA 1200P2 Monitor: Samsung Odyssey G7 (1440P @ 240Hz) This I have tried: DDU drivers, new windows install. Updated latest bios v3403. Took apart whole system and rebuilt. Disabled XMP and any other OC. Reset BIOS to default settings. Disabled 4G decoding and resizable bar. Reseated RAM and changed the order they are in (doesn't matter as all 8 slots are used) - Memtest came up 0 errors Replaced/tested with 1080ti and 980ti. Replaced/tested with ASUS Thor 850W PSU. Replaced/tested monitor with LG 51' TV and Samsung 24" monitor. Checked thermals under gaming load. CPU <60, GPU<70, RAM<50. Any ideas?

Try going to Windows 10 settings > Graphics Settings > Graphics performance preference > Add your preferred game .exe > click options on your game and set to high performance.

I'm looking for a new VRR monitor to go with my new RTX 3000 GPU. Namely ASUS ROG STRIX Curved XG43VQ People have advised to look at the list of Nvidia tested Gsync monitors to check for compatibility. This monitor does not appear on the list. https://www.nvidia.com/en-gb/geforce/products/g-sync-monitors/specs/ Other have said any AMD Freesync is practically guaranteed to work with Gsync, I have not seen any written materials from the monitor manufacturer to support this. Is this true or are they misinformed?

Well I haven't seen power mentioned yet so I'll throw it out there. Is the PSU sufficient to power the 3090? Also what temps is the GPU hitting? If you could log it until it it switches off again it might indicate something to do with sensors.

I couldn't get my rig to hold steady under load even though it passes memtest with no errors. I swapped it for a set on the QVL and it worked fine after that. I guess everyone's mileage may vary.

I would also check the QVL of the mobo to ensure the memory is compatible otherwise you may not get stability at full speed.

Could it be overheating issue? Is the GPU plumbed in properly or are you using it as ambient? Also does the GPU have an LN2/OC bios mode?

Didn't work. Seems either its this particular config of tv/monitor OR dying GPU I think. Should have bought that 3090 while it was still in stock...

Replaced the cable with another ultrahigh speed HDMI, but I don't have another model monitor/tv to try. I do have 2 other tvs that are of the same model.