This is exactly what my GPU is doing. I have a 3090, Runs fine and will just randomly black screen. Restart the computer, pic comes back then randomly black screens again. I find it amazing that you can pull the GPU cores so clean with just basic reflow machine like a quick station and a under heater. When I was doing similar work, I was repairing PS4 and Xbox one gaming consoles, and I used a ERSA IR 550 for reworking their APU's, as the large heater for a IC that size was great, My work place latter upgraded to the HR 600 which was a BEAST but that's just an over kill machine for this kinda work. For the lols we bought a cheap Jovy Systems RE-7500 Infrared BGA Rework Station and gave that a whirl and the hilarious thing, was that it was just as good as the HR 600, for less than 10X the price. Anyway, Keep the videos coming, I love watching your repairs, because unlike a lot of the other "motherboard tech's" out there, you are the real deal.
What PSU are you running with the 3090? I had an 850w Corsair and PC would crash constantly when trying to run a game. Would restart instantly as soon as I started a game. Got a 1000w PSU and fixed it. Took me a month of testing everything. The transient power spikes where the culprit.
I hope you live forever my friend :) I don't marvel at many people but you impress me. Not that I am anyone worth impressing but, I do appreciate skill and talent and people that really GIVE something in areas that is hard to find people doing the things that need to be done. In the days of throwing it in the trash it is refreshing to see things fixed instead of thrown away.
just mount you GPU vertically and you'll be fine, avoiding sag and any of its consequences this way! Or at least use a GPU support "leg". It's not the preffered method imo, but it should work.
This card temp are lower because it is not high wattage card like the FTW3 Ultra or the others, iirc its limited to 350watt while the FTW ultra / Rog Strix and many others can easily go beyond 400Watt Look at the board power draw in the FTW3 Ultra screenshot
Suggestion: Change the response to the customer of “no fix” To “Beyond Economical Repair” The difference is that, you’re not refusing to do the work or charging the customer for “not fixing it” The diagnostic fee covered the analysis to then provide the customer the information to make a business decision. It’s subtle, but no one is paying to not have something fixed. It’s diagnosed, (as asked) and then a decision is made it’s beyond economical repair.
A very good point - being beyond economic repair is a diagnosis in itself. I've just had that diagnosis myself from a technician for a 2080ti with it's memory controller dying and artifacting under load which needs a new GPU entirely.
@@northwestrepair I think you deserve what you’re worth, and I think your repairs are worth it. Like going to the Doctor, you want to know what’s wrong and then have options. If the doctor says no fix, how would you feel? If the doctor says the prognosis is this, these are the options. You’re likely to feel better about the bill to go see the doctor. Saying this as a suggestion as you had one upset customer - trying to figure out why they’re upset, what did they pay for… The wording, believe it or not, it does matter.
The reason you have lower fps is that at such resolution, most likely, the gpu becomes cpu bound. Thus, in case with different cpu, you may get lower or higher results in fps.
Recently i brought my Asus TUF 4090 to RMA, since hotspot got to 110 'c, GPU itself got max 74 and mem 60. Was using it for almost a year. GPU never crashed.
I have a zotac 3080, compared to your temps I might need to repaste/ new thermal pads but I’m kind of scared to haha. It pulls a little over 300 watts and hangs out around low 70’s in degrees in an airflow case. Seems fine but prob need to because it was a local used buy
Another sucesful repaire watching Your videos it all looks so easy. Great job as always. I have been wondering, what kind of software do You use, I hear of mats and used it but it only works when Gpu is main never got it to work via mobo Hdmi ? I have about 10 or soo 10 serijes nvidia cards, all have the same issue, black screen with all voltages present and every singel one show no error on memory using 387 and 400 mods
Do you reball with real lead solder or do you use lead free solder ? Lead solder holds up better for stress on the pads. Leaded solder also has better heat transfer to the bottom of the board.
@@northwestrepairI presume they (manufacturers/vendors) do not use such in excuse of green Earth against the pollution and stuff. Haha. Can you confirm though? Also ty for content!
@@kreozello some say it's for worker safety but I don't buy it. Some say it's for greener environment. I think it's for more green $ in their pockets. Nothing else.
@@northwestrepair Some countries, especially in the EU, have strict regulations (Restriction of Hazardous Substances Directive, RoHS) when it comes to lead in consumer products. And they're not going to make different versions for EU and rest of the world, so everybody gets lead free solder
GPU load looked to be a bit lower on the MSI vs the zotac, which would definitely mean lower FPS. Probably should use a higher resolution or more demanding settings to try and get the load near 100%
I have an Asus Strix 3070, it runs cool …. Hotspot under 75C and deltaT around 5-6C. I specifically choose this model due to cooling perfomance as seen o techpowerup review
i always care for my pc to be cool, cooler it gets longer will last. have computers from 90s still working now :) being loud dont care that much, have a good headset block all outside sounds. almost no fan noise if under 60C but after starts going quite high, with new cpu i try to keep it under 70C. gpu same but usually is around 70C in a very intensive env
Both. Lead-free solder is very brittle, and A1 chip lives on the dangerous fault line of card sag. Next up, is GPU core, because the large flat chip can't flex when a PCB bends. There's a problem with super heavy cards having only two mounting contact points, back screw and PCI-E card edge. Everything hanging after the slot acts as a lever. I hope this is helpful in answering your question.🙂
So from a longevity standpoint, it makes sense to go with a waterblock? To keep the temperature difference lower and thus less thermal expansion and contraction. Assuming you dont fuck up the installation with incorrectly sized thermal pads and such
Having run most of my GPUs for more than a decade I would say the water-cooled GPU is mostly endangered by leaked water (dripping on GPU from other components above), less by incorrect installation. I did it mostly for overclocking but nowadays more for reduced (non-existent) fan noise. Water-cooling your PC requires more regular maintenance (check temps, water level, no leaks), and it's a pain if you want to change something as the water hoses will be in the way or you may need to drain and open the circuit, which is also a risk of spilling water.
@@Martinit0 Yeah, but what about longetivity? As for maintenance I run long flexible tubes to GPU&CPU blocks, so I can take them out and move them to the side without disconnecting hoses (enough so I can replace the motherboard), and have a couple of strategically placed valves for drain and fill when needed. I have the same reasons as you, fan noise and overclocking.
I miss my 3080 Ti.. I bought a 7900 XTX Hellhound, which is also nice, but oh my god the temp delta is crazy. Yesterday I managed to reach 59c global and 105c hot spot! I've asked around and many tell me its normal, but it doesn't make me comfy
Thats not normal, hotspot difference should not be higher than 17 degrees. From core temp. Your gpu either have thermalpads that are too big, or also has thermalpaste squished out from the core. Or just bad mounting pressure. Eitherway thats not good at all
Wasn't there an Error in the production of the vaporchamber on some of these 7900 cards? Maybe that's the case here, as the result was too high temperatures!
@@ascissordollynamedgwen9409 17 degrees might be the ideal spread for nVidia cards and lower end Radeon cards, but for Radeon in general it depends on the specific card what the acceptable TJ spread is, as it varies depending on what the edge temperature is and the amount of power being sent through the core. On water cooled and flagship overclocked models anything under a 30C delta is acceptable, while a 20-25C delta is very typical on high end air cooled cards. Radeons sensors are closers to the actual TJs in the cores than nVidia's sensors are, so they tend to represent the actual TJ temps more accurately. That being said, I agree that a 59c edge and 105c hot spot is a little too much delta, the Hellhound typically maxes at 60C edge and 90C TJ, right at or around the typical upper delta of 30C. It would probably benefit from a repaste, or better yet, a swap to a PTM7950 pad on the core.
Interesting comment about memory temp and hot spot temp's should be around the same. Is this an Nvidia thing? With my 7800 Nitro the memory temp is 16 degrees above the hot spot at idle. The gap closes on load, and the memory temp rarely exceeds low 80's. I've seen temps similar to this posted on reddit.
I have a 3080 turbo edition from gigabyte with the terrible turbine/blower fan setup. If I let my card run with factory settings it runs at 90c, and then thermal throttles like crazy. The fan sounds like a jet taking off for a few seconds. I have to set the power limit to around 60% and temp 75c. Thats as low as i can go before i just cant run games anymore. Do you recommend I try a pad swap? Im very disappointed in having spent so much on a card that is thermally terrible.
I'm beginning to think that all 30 series Ventus cards are garbage. I have a friend with a Ventus 3060 TI that ran hot, and I had a Ventus 3080 12G that I RMAed 3 times and the previous owner RMAed once. MSI ended up sending me a Suprim 3080 TI that I've had no problems with so far. 🤞
So basically, the bigger the chip, the bigger the PCB, the bigger the cooler... The more common GPU solder joint failure will be. A GPU riser is not an option anymore.
Question , I just bought a Zotac 3080ti extreme Halo and I get all the bought but zero display .I do have a 600 watt power supply so I'm not sure if that is the issue ? or do you think something is worst with the card . ? I don't get a boot screen or anything.
My EVGA 3080 ftw3 12gb is very similar to this in terms of temps and core and hot spot delta, if we compare at the same wattage, which is 350w in your test, at stock EVGA will have higher power limit. Delta is 6-7 degrees for hot spot and 10-12 for memory. And it's like that for more than year, nothing changed. I guess it's because EVGA were using phase change thermal paste in their cards.
@@TimberWulfIsHere No, it hasn't been repasted or repaded. In terms of delta between core and hot spot, I've seen some other evga ftw3 cards with similar delta. Like i said it's probably because of phase change thermal paste evga used in their cards. Most thermal pastes pump out quickly when using them on big dies, so phase change ones are the best for GPUs
@@Xayc__ dude they don't use that stuff. I would know, I've replaced all that crud myself with aftermarket pastes and pads. They use cheap, low thermal conductive pastes and pads. Most 3080s get 100+ degree mem temps and high core temps stock. It's with better pads and a repaste or ptm 7950 that you will have good core temps and low hotspot temps.
For FPS discrepancy, my guess is since you are running at a very low resolution, the performance is limited by cpu. As you’re saying that you switched cpu between benchmarking zotac and MSI, you would get different fps
"and that's a no fix...* No, that's a won't -fix, lol. What an arrogant SoB. And he knows he won't fix these cards but wants to charge that fee anyway. That's one of the big differences here.
Hello, im having a rtx 3080 zotak for repair.. Computer doesnt post if i have her connected, tried other pci port didnt changed. Computer works well when i try gtx750, Also tried a 850w power nothing changed. When i start with both graphis connected and hdmi on gtx 750 computer starts and detects other graphic card but doesnt work. Pretty sur gpu faulty. Can u look at it ? for express repair how much it can go ? Thank you
Simple choice. Use a card support bracket, or watch those A1s fly.🚀🌕 You really should take graphics cards out for dinner and a movie, or at least buy a few drinks, before you undress and deflower them.😁 And don't forget the love-glove. You never know where they've been.😱
😄 to the moon with A1s I prefer using a vertical mount bracket in conjunction with a PCIe riser and forget about sagging and damaging the card that way.
@@edalecu Good going.👍 One less graphics card has to visit St. Peter (Святой «Санкт» Пётр).☦️🪦🌷🕳️⚰️ Edit: There, UA-cam! At least i made the grave look more respectable, with flowers too.😁
@@edalecu I'm running the same setup; vertical GPU with PCIe cable, metal support bracket. I think it l looks better that way too, seeing the front and the fans instead of the top card-edge.
@northwestrepair Tony, how comes we have so many broken joints in gpus but not other electronics? Mechanical stress from sagging? Leaded solder should not suffer under normal operating temperatures of the card so what is it, temperature plus vibration from the fans maybe? Bit of both?
My best guess would be thermal cycling (expansion & shrinking) + physical stress (sagging) + borderline silent fan profiles (vendors often choose silence over thermals) + cards are voltage/frequency walled from the factory
Buy a super car, drive it hard on the track for most of it's life, compare it Joe bloggs commuter car maintenance wise, that's the difference. GPU's are driven hard and put away wet.
No major manufacturer of consumer equipment uses leaded solder anymore. It's banned from products sold in the EU (with some exceptions) and the industry decided it would be easier to use lead-free everywhere than to maintain seperate supply chains for EU and non-EU customers. Lead-free solder has a higher melting point than lead-based solder but tends to be more prone to cracking. As for GPUs I think they show up a lot in repair videos for a bunch of reasons. 1. High end GPUs are expensive enough to be worth fixing. 2. The PC expansion slot form factor sucks for high power devices. With limited mecahnical support (increasing flexing stresses on the PCB) and limited room for a good cooler (increasing thermal stresses). 3. Gamers buy based on performance with the result that many cards are sold "pre-overclocked".
@kreozello do you have any idea how much i paid for this card back in 2021? Do you think we have a choice back then. I could get a 4090 today and had some change for the price i paid for this reference card. Meh, it may be, but back then, unless you were willing to spend 3 grand on 3090, this was the most "affordable" and somewhat available.
@@ilovelimpfries oh, I know bro. These prices were ASS. Total ASS. I wanted GTX1660 Super from MSI (Gaming Z) back then at 2020 when it was around 16k rub (approx 200 usd?). Then it went 41k, 60k... 140k 💀 And now minimal bid is 20k lol. I blame my parents for fooling me at my first build and instead deciding to ditch helping me. Could've sat straight for 2 years & resell it with profit to upgrade for current RX6750xt. Damn my GT640....
MSI RTX 3080 10G LHR here 96c hotspot, 102c memory temp, 81c core temp. am i fucked? whats the resolution here? the card was bought two years ago, does it already need to be repasted and have new thermal pads applied? gonna try to apply a custom fan curve to try to remedy the situation in the meanwhile that said, i don't experience any instability or the like. occasional artifacting for a millisecond on youtube only, which hasn't been substantial enough yet for me to draw any conclusions about. other than that its very unlikely to be drivers, since i've practically went through an entire 100.xx worth of drivers.
I think you would be well served by repasting using a higher quality thermal paste and, more importantly, I'd get higher quality replacement pads. It was not uncommon for the 3080's (and 3090's) running Micron GDDR6x to hit triple digit temps (throttling is at 110c) right off the factory floor, but that doesn't mean it's good. It's not a matter of the thermal pads being dry, it's that many of the board partners used relatively low-spec thermal pads, and few 3080's and 3090's really dealt with memory cooling well. The EVGA's 3080's and the Asus TUF RTX 3080 seemed to be the exceptions that I noted - though (under heavy load) the coolest 3080 I've seen still has memory temps that peak in the mid 80's (for example, my TUF 3080 OC peaks at 75 hotspot, 84 Memory, 64 core temp).
@@northwestrepair You said before that you learned repair by watching all these repair videos. I swear you had some kind of an electronics background before you started.
This was my card! Can't praise the kindness and quickness of you enough with my card. Thank you for saving me time and money :)
how much was this repair?
Yea dude spill the beans
I remember when you hit 16,000 subscribers.
I said if this channel was stock I would buy it.
would have made 7 x my money.... good job
lol
You lift so many cores, you must be buff as hell!
Lol
Nothing is boring you put out man. All very educational! Thank you!!
This is exactly what my GPU is doing. I have a 3090, Runs fine and will just randomly black screen. Restart the computer, pic comes back then randomly black screens again. I find it amazing that you can pull the GPU cores so clean with just basic reflow machine like a quick station and a under heater. When I was doing similar work, I was repairing PS4 and Xbox one gaming consoles, and I used a ERSA IR 550 for reworking their APU's, as the large heater for a IC that size was great, My work place latter upgraded to the HR 600 which was a BEAST but that's just an over kill machine for this kinda work. For the lols we bought a cheap Jovy Systems RE-7500 Infrared BGA Rework Station and gave that a whirl and the hilarious thing, was that it was just as good as the HR 600, for less than 10X the price. Anyway, Keep the videos coming, I love watching your repairs, because unlike a lot of the other "motherboard tech's" out there, you are the real deal.
What PSU are you running with the 3090? I had an 850w Corsair and PC would crash constantly when trying to run a game. Would restart instantly as soon as I started a game. Got a 1000w PSU and fixed it. Took me a month of testing everything. The transient power spikes where the culprit.
I hope you live forever my friend :)
I don't marvel at many people but you impress me.
Not that I am anyone worth impressing but, I do appreciate skill and talent and people that really GIVE something in areas that is hard to find people doing the things that need to be done.
In the days of throwing it in the trash it is refreshing to see things fixed instead of thrown away.
i loved the temperature and performance comparison please do more of that 😍
It's not a fair comparison because of different motherboard and CPU
@@northwestrepair It's a stupid video bro. That card is known to be the worst 3080ti and here you are praising a broken POS.
You don't make boring videos! You inform and you share your knowledge. Thank you.
Watching this channel terrifies me from buying that GPU upgrade I've been wanting for a long time now.
Just buy new and you'll be safe for 3 years
just mount you GPU vertically and you'll be fine, avoiding sag and any of its consequences this way!
Or at least use a GPU support "leg". It's not the preffered method imo, but it should work.
GPU Riser and an appropriate PC Case.
Aorus Master 3080ti here. Now I feel an angel of doom peaking at me. Thanks for the video. Great work.
I have been taking notes and listen to every word. Soon I will get to wrecking stuff. Thanks for all the knowledge you share.
1:16 Were you also shouting "Press the key! Press the key!" while watching? :D
This card temp are lower because it is not high wattage card like the FTW3 Ultra or the others, iirc its limited to 350watt while the FTW ultra / Rog Strix and many others can easily go beyond 400Watt
Look at the board power draw in the FTW3 Ultra screenshot
Suggestion:
Change the response to the customer of “no fix”
To “Beyond Economical Repair”
The difference is that, you’re not refusing to do the work or charging the customer for “not fixing it”
The diagnostic fee covered the analysis to then provide the customer the information to make a business decision.
It’s subtle, but no one is paying to not have something fixed. It’s diagnosed, (as asked) and then a decision is made it’s beyond economical repair.
Then i will become northirdgefix and get rich
@@northwestrepairDon't forget to remind us of buying stuff from your shop
A very good point - being beyond economic repair is a diagnosis in itself. I've just had that diagnosis myself from a technician for a 2080ti with it's memory controller dying and artifacting under load which needs a new GPU entirely.
@@northwestrepairI think the difference is an important ones to distinguish you from him instead.
@@northwestrepair I think you deserve what you’re worth, and I think your repairs are worth it.
Like going to the Doctor, you want to know what’s wrong and then have options. If the doctor says no fix, how would you feel?
If the doctor says the prognosis is this, these are the options. You’re likely to feel better about the bill to go see the doctor.
Saying this as a suggestion as you had one upset customer - trying to figure out why they’re upset, what did they pay for…
The wording, believe it or not, it does matter.
I had an msi 3080 Ti Ventus 3X just like this. Great card!
i miss those parts of the videos where you put music for the part where youre lifting the core
The reason you have lower fps is that at such resolution, most likely, the gpu becomes cpu bound. Thus, in case with different cpu, you may get lower or higher results in fps.
MSI usually means to me Major Service Issues
Though MSI is not even leader on that topic, I sincerely appreciate the satire there! 😂
Literally had MSI motherboards that loved killing themselves for no reason. Never buying shit from MSI again
@@zetsubou3704 Tomahawk motherboards are the best for the money tho, since B450. Mine running strong since 2020
good luck with that since the only trustworthy brand EVGA is no longer with us. Asus and gigabytes aren’t anything better than MSI.
Hell no... they are the most reliable of the Big 3.
At 60% runtime and he's still opening the back plate, I'm sure there would be a good twist in this episode.
Recently i brought my Asus TUF 4090 to RMA, since hotspot got to 110 'c, GPU itself got max 74 and mem 60. Was using it for almost a year. GPU never crashed.
I get better on the trinnity 3080ti under load but that is for later 😊. Nice work boss
Can you please recommend card brand and model to buy that has low failure rate, starting to think that most of them are trash...
Have you ever seen a core that just malfunctions, but is not RIP? So there are constant artifacts or something?
Always the best kind of fix just needs to be realigned by the gurus wizard hands 😂
I have a zotac 3080, compared to your temps I might need to repaste/ new thermal pads but I’m kind of scared to haha. It pulls a little over 300 watts and hangs out around low 70’s in degrees in an airflow case. Seems fine but prob need to because it was a local used buy
With an aggressive custom fan curve in afterburner (linear in my case), the temps could have been reduced I suppose
Another sucesful repaire watching Your videos it all looks so easy. Great job as always. I have been wondering, what kind of software do You use, I hear of mats and used it but it only works when Gpu is main never got it to work via mobo Hdmi ? I have about 10 or soo 10 serijes nvidia cards, all have the same issue, black screen with all voltages present and every singel one show no error on memory using 387 and 400 mods
This is what you get with power hungry and hot GPUs.
8800 GTX was doing the same thing back in the day.
Glad I run mine at 300W power limit.
Do you reball with real lead solder or do you use lead free solder ? Lead solder holds up better for stress on the pads. Leaded solder also has better heat transfer to the bottom of the board.
lead solder only
@@northwestrepairI presume they (manufacturers/vendors) do not use such in excuse of green Earth against the pollution and stuff. Haha. Can you confirm though? Also ty for content!
@@kreozello some say it's for worker safety but I don't buy it.
Some say it's for greener environment.
I think it's for more green $ in their pockets. Nothing else.
@@northwestrepair Some countries, especially in the EU, have strict regulations (Restriction of Hazardous Substances Directive, RoHS) when it comes to lead in consumer products. And they're not going to make different versions for EU and rest of the world, so everybody gets lead free solder
@@northwestrepair yep, prob greenier in wallet rather than on planet Earth 😆
I have that same identical card. I'm watching this while biting my nails.
GPU load looked to be a bit lower on the MSI vs the zotac, which would definitely mean lower FPS. Probably should use a higher resolution or more demanding settings to try and get the load near 100%
When they removed lead from solder, they ruined all electronics, but in their favor. Things break easier, now.
Fr, not even phones are safe with many needing SoC reballing
Why did they remove the lead from the solder in electronic components now?
I have an Asus Strix 3070, it runs cool …. Hotspot under 75C and deltaT around 5-6C. I specifically choose this model due to cooling perfomance as seen o techpowerup review
The performance differences are not in the frequencies, they are in the powerlimit.
i always care for my pc to be cool, cooler it gets longer will last. have computers from 90s still working now :) being loud dont care that much, have a good headset block all outside sounds. almost no fan noise if under 60C but after starts going quite high, with new cpu i try to keep it under 70C. gpu same but usually is around 70C in a very intensive env
So why do we see so many broken solder joints and ripped pads? Do they make them too brittle or something?
Both. Lead-free solder is very brittle, and A1 chip lives on the dangerous fault line of card sag. Next up, is GPU core, because the large flat chip can't flex when a PCB bends. There's a problem with super heavy cards having only two mounting contact points, back screw and PCI-E card edge. Everything hanging after the slot acts as a lever.
I hope this is helpful in answering your question.🙂
So from a longevity standpoint, it makes sense to go with a waterblock?
To keep the temperature difference lower and thus less thermal expansion and contraction. Assuming you dont fuck up the installation with incorrectly sized thermal pads and such
Having run most of my GPUs for more than a decade I would say the water-cooled GPU is mostly endangered by leaked water (dripping on GPU from other components above), less by incorrect installation. I did it mostly for overclocking but nowadays more for reduced (non-existent) fan noise.
Water-cooling your PC requires more regular maintenance (check temps, water level, no leaks), and it's a pain if you want to change something as the water hoses will be in the way or you may need to drain and open the circuit, which is also a risk of spilling water.
@@Martinit0 Yeah, but what about longetivity?
As for maintenance I run long flexible tubes to GPU&CPU blocks, so I can take them out and move them to the side without disconnecting hoses (enough so I can replace the motherboard), and have a couple of strategically placed valves for drain and fill when needed.
I have the same reasons as you, fan noise and overclocking.
For some reason I keep thinking of Rick & Morty's Ball Fondlers whenever I see references to balls.
I miss my 3080 Ti.. I bought a 7900 XTX Hellhound, which is also nice, but oh my god the temp delta is crazy. Yesterday I managed to reach 59c global and 105c hot spot! I've asked around and many tell me its normal, but it doesn't make me comfy
Thats not normal, hotspot difference should not be higher than 17 degrees. From core temp. Your gpu either have thermalpads that are too big, or also has thermalpaste squished out from the core. Or just bad mounting pressure. Eitherway thats not good at all
Wasn't there an Error in the production of the vaporchamber on some of these 7900 cards? Maybe that's the case here, as the result was too high temperatures!
@@ascissordollynamedgwen9409 17 degrees might be the ideal spread for nVidia cards and lower end Radeon cards, but for Radeon in general it depends on the specific card what the acceptable TJ spread is, as it varies depending on what the edge temperature is and the amount of power being sent through the core. On water cooled and flagship overclocked models anything under a 30C delta is acceptable, while a 20-25C delta is very typical on high end air cooled cards. Radeons sensors are closers to the actual TJs in the cores than nVidia's sensors are, so they tend to represent the actual TJ temps more accurately.
That being said, I agree that a 59c edge and 105c hot spot is a little too much delta, the Hellhound typically maxes at 60C edge and 90C TJ, right at or around the typical upper delta of 30C. It would probably benefit from a repaste, or better yet, a swap to a PTM7950 pad on the core.
Do you have a video on that boot drive you use to test memory. I don't know whats it called.
I wonder if Alex is watching and taking notes… ;)
Love the vids man, i wish i could search various gpus and get the vidoe that corresponds
Interesting comment about memory temp and hot spot temp's should be around the same. Is this an Nvidia thing? With my 7800 Nitro the memory temp is 16 degrees above the hot spot at idle. The gap closes on load, and the memory temp rarely exceeds low 80's. I've seen temps similar to this posted on reddit.
Great job as always Tony 👍. Only 2K to go 😉
Hi. i can see you have a different type of mats which can test the rtx 4000 series aswell.
Where can i find it?
Cheers
Where can I find out that diagnostics sw? Its your private special one?
So what caused this issue? Too much heat? Too much use? I don’t get it, how you can prevent something like this?
screen went black in the video i thought my gpu went bad
By making videos he is getting double of the repair price per repair. Thats big brain move.
I have a 3080 turbo edition from gigabyte with the terrible turbine/blower fan setup. If I let my card run with factory settings it runs at 90c, and then thermal throttles like crazy. The fan sounds like a jet taking off for a few seconds.
I have to set the power limit to around 60% and temp 75c. Thats as low as i can go before i just cant run games anymore.
Do you recommend I try a pad swap? Im very disappointed in having spent so much on a card that is thermally terrible.
I'm beginning to think that all 30 series Ventus cards are garbage. I have a friend with a Ventus 3060 TI that ran hot, and I had a Ventus 3080 12G that I RMAed 3 times and the previous owner RMAed once. MSI ended up sending me a Suprim 3080 TI that I've had no problems with so far. 🤞
GPU usage should be at 99% to reach the max temp. but in that test usage is 86%. you should benchmark in 4k.
I’m so curious what does something like this cost?
So basically, the bigger the chip, the bigger the PCB, the bigger the cooler...
The more common GPU solder joint failure will be.
A GPU riser is not an option anymore.
Question , I just bought a Zotac 3080ti extreme Halo and I get all the bought but zero display .I do have a 600 watt power supply so I'm not sure if that is the issue ? or do you think something is worst with the card . ? I don't get a boot screen or anything.
Where can i download that Nvidia GPU testing utility?!?!
Im trying to rule out a potential fault with my GPU, and this would help me so much!
Fascinating skills once again! Russ
What kind of preheat bed is that T-8280 ?
It seems that the FPS are lower because the GPU is only about 75-85 utilized.
My EVGA 3080 ftw3 12gb is very similar to this in terms of temps and core and hot spot delta, if we compare at the same wattage, which is 350w in your test, at stock EVGA will have higher power limit. Delta is 6-7 degrees for hot spot and 10-12 for memory. And it's like that for more than year, nothing changed.
I guess it's because EVGA were using phase change thermal paste in their cards.
Delta from what? Tpackage?
@@VndNvwYvvSvv hot spot and memory delta from core temperature
Your card is an outlier or has been repasted/repaded
@@TimberWulfIsHere No, it hasn't been repasted or repaded. In terms of delta between core and hot spot, I've seen some other evga ftw3 cards with similar delta. Like i said it's probably because of phase change thermal paste evga used in their cards. Most thermal pastes pump out quickly when using them on big dies, so phase change ones are the best for GPUs
@@Xayc__ dude they don't use that stuff. I would know, I've replaced all that crud myself with aftermarket pastes and pads. They use cheap, low thermal conductive pastes and pads. Most 3080s get 100+ degree mem temps and high core temps stock. It's with better pads and a repaste or ptm 7950 that you will have good core temps and low hotspot temps.
Do you know how to fix motherboards as well? it'd be interesting
motherboards are not worth fixing in m opinion. They dont cost a lot of money so shipping and repair cost will not make much sense.
I have an asus tuf 3080ti it runs @ 84 C 100% fan, only undervolting doesn't help. Should I change the paste if it's still under warranty?
For FPS discrepancy, my guess is since you are running at a very low resolution, the performance is limited by cpu. As you’re saying that you switched cpu between benchmarking zotac and MSI, you would get different fps
Have you ever seen any issues with 3090Ti’s or are these more reliable than other 30 series cards?
As far as i remember 3090 ti's have a way better built vrm than regular 3090's on average so yeah i think they are more reliable
Did anyone see that Northridgefix video couple of weeks ago? 19 4090 GPU's all cracked, he deemed all of them a no fix 😂😂
"and that's a no fix...* No, that's a won't -fix, lol. What an arrogant SoB. And he knows he won't fix these cards but wants to charge that fee anyway. That's one of the big differences here.
The only thing you can do is salvage the the memory and the GPU... and the VRMs.
I've seen this guy fix many cracked boards, so you're talking nonsense...
You said “ i am sorry its just a boring re balling video “ 😂
Hello, im having a rtx 3080 zotak for repair.. Computer doesnt post if i have her connected, tried other pci port didnt changed. Computer works well when i try gtx750, Also tried a 850w power nothing changed. When i start with both graphis connected and hdmi on gtx 750 computer starts and detects other graphic card but doesnt work. Pretty sur gpu faulty. Can u look at it ? for express repair how much it can go ? Thank you
sounds like factory problem to me if reballing them fixes them all
Have you fixed any MSI GeForce RTX 3080 12GB GDDR6 Gaming Z Trio LHR ? Are they any good is the one in the video? Thanks!
why there are lots of gpus that need reball nowadays? do they use bad tin or what?
Surface mount devices and torsion twisting are not a good mix. There is a fundamental engineering flaw in this technology that needs to be addressed.
Maaaan.. I know somehow you have a bit negative sentiments about AMD is, but why most of the GPU that need to be repaired is always Nvidia tho?
If you specialize in (mostly) Nvidia repairs it should be no surprise when you end up fixing mostly Nvidia GPU's.
@@volentimeh I watched plenty(more than 20) of his video and I don't see or hear anything about "Specialize in Nvidia repair" interesting 🤔
is this sort of issue common with nvidia gpus ? or amd as well? im considering upgrading soon and these videos kinda discourage me xD
Nvidia is using bigger chips so it should be more common for them.
Simple choice. Use a card support bracket, or watch those A1s fly.🚀🌕
You really should take graphics cards out for dinner and a movie, or at least buy a few drinks, before you undress and deflower them.😁 And don't forget the love-glove. You never know where they've been.😱
😄 to the moon with A1s
I prefer using a vertical mount bracket in conjunction with a PCIe riser and forget about sagging and damaging the card that way.
@@edalecu Good going.👍 One less graphics card has to visit St. Peter (Святой «Санкт» Пётр).☦️🪦🌷🕳️⚰️ Edit: There, UA-cam! At least i made the grave look more respectable, with flowers too.😁
Nice of UA-cam to put my comment back.😁 Maybe i'll remove the one about calling them dicks.😆
@@edalecu I'm running the same setup; vertical GPU with PCIe cable, metal support bracket. I think it l looks better that way too, seeing the front and the fans instead of the top card-edge.
its it me or the sound volume of your recent videos are really low
you run it in pcie 3.0 (16x3.0) that's why
If the GPU is faulty and the serial number is not visible, how can I check it?!
Ever came acrosss a situation where NVIDIA gpus working fine on AMD Mobo But Not Giving display on Intel Mobo?
@northwestrepair Tony, how comes we have so many broken joints in gpus but not other electronics? Mechanical stress from sagging? Leaded solder should not suffer under normal operating temperatures of the card so what is it, temperature plus vibration from the fans maybe? Bit of both?
My best guess would be thermal cycling (expansion & shrinking) + physical stress (sagging) + borderline silent fan profiles (vendors often choose silence over thermals) + cards are voltage/frequency walled from the factory
Buy a super car, drive it hard on the track for most of it's life, compare it Joe bloggs commuter car maintenance wise, that's the difference.
GPU's are driven hard and put away wet.
No major manufacturer of consumer equipment uses leaded solder anymore. It's banned from products sold in the EU (with some exceptions) and the industry decided it would be easier to use lead-free everywhere than to maintain seperate supply chains for EU and non-EU customers. Lead-free solder has a higher melting point than lead-based solder but tends to be more prone to cracking.
As for GPUs I think they show up a lot in repair videos for a bunch of reasons.
1. High end GPUs are expensive enough to be worth fixing.
2. The PC expansion slot form factor sucks for high power devices. With limited mecahnical support (increasing flexing stresses on the PCB) and limited room for a good cooler (increasing thermal stresses).
3. Gamers buy based on performance with the result that many cards are sold "pre-overclocked".
Good job!
Awesome video
My friend bought msi laptop with rtx, and shutdown when use it for rendering
Brand new Gigabyte RTX 4070ti Super having temperature delta 15~19 degrees under load - is it OK?
No.
It's mx4 enough for GPU?
I see a lot of MSI on your channel, do you recommend MSI cards?
It's not MSI it's Nvidia.
My MSI 4090 is 27c at idle, and 34c while gaming, should I be concerned?
@@davehenderson6896 how is a 4090 at 34° while gaming??? Not even watercooling it would get that temp
My msi4090xtrio is 35c at idle and 65c when play cyberpunk 2k max RToverdrive(DLSS ultra quality)
Yeah those are good stats. Why do people poop on MSI?@@HoangDucNuyen
you see GPU¨s in your Dreams too xD
Hello internet
We're is the hot-spot measured from? Core or vrm?
Hotspot is one of several temp sensors built into the core these days
the x pattern of placing thermal paste is wrong.
Yes, you should only blast it on with a stucco sprayer, for that hippo spatter look.😁
I have a watercooled version of this card, never breach 60 degrees in sustain max load. And the delta between hotspot and gpu never exceed 6 degrees.
Is it Suprim X liquid? It's basically not the Ventus one, lol. Suprim is way better! And even better than the Gaming one.
@@kreozello Nah, it's the ventus. I watercooled my ventus with alphacool kit.
@@ilovelimpfries rlly? Why ventus? Cuz was affordable? Imo it's meh card.
@kreozello do you have any idea how much i paid for this card back in 2021? Do you think we have a choice back then. I could get a 4090 today and had some change for the price i paid for this reference card. Meh, it may be, but back then, unless you were willing to spend 3 grand on 3090, this was the most "affordable" and somewhat available.
@@ilovelimpfries oh, I know bro. These prices were ASS. Total ASS. I wanted GTX1660 Super from MSI (Gaming Z) back then at 2020 when it was around 16k rub (approx 200 usd?). Then it went 41k, 60k... 140k 💀
And now minimal bid is 20k lol. I blame my parents for fooling me at my first build and instead deciding to ditch helping me. Could've sat straight for 2 years & resell it with profit to upgrade for current RX6750xt. Damn my GT640....
What is RTX 5000?
Good work
Thanks
Поделитесь образом где mats
i have a 3080 10gb that i think needs this exact repair. how much does it cost to have this done? should i send you an email?
email of discord is fine
@GenocideJoozsometimes a 100$ repair is better as you know what you have instead of getting a another card for 300$ and perhaps get scammed. 😊
Well at least no pads fell off - lol -.
How to contact you? Looking for a single fan gtx 1650 for an old lenovo m83 mid tower. Looking for a trusted source. Thanks
I see your emai in the video. Sorry, thanks
Once again people just run the fans on high curves quite is stupid if it means the heat kills your gpu
MSI RTX 3080 10G LHR here
96c hotspot, 102c memory temp, 81c core temp.
am i fucked? whats the resolution here? the card was bought two years ago, does it already need to be repasted and have new thermal pads applied?
gonna try to apply a custom fan curve to try to remedy the situation in the meanwhile
that said, i don't experience any instability or the like. occasional artifacting for a millisecond on youtube only, which hasn't been substantial enough yet for me to draw any conclusions about.
other than that its very unlikely to be drivers, since i've practically went through an entire 100.xx worth of drivers.
I think you would be well served by repasting using a higher quality thermal paste and, more importantly, I'd get higher quality replacement pads. It was not uncommon for the 3080's (and 3090's) running Micron GDDR6x to hit triple digit temps (throttling is at 110c) right off the factory floor, but that doesn't mean it's good. It's not a matter of the thermal pads being dry, it's that many of the board partners used relatively low-spec thermal pads, and few 3080's and 3090's really dealt with memory cooling well. The EVGA's 3080's and the Asus TUF RTX 3080 seemed to be the exceptions that I noted - though (under heavy load) the coolest 3080 I've seen still has memory temps that peak in the mid 80's (for example, my TUF 3080 OC peaks at 75 hotspot, 84 Memory, 64 core temp).
NWR: GPU reballing is a booring video :|
Meanwhile some repairmen: GPU reballing is unpractical to do XD
Did you say RTX 5000 ?
Not sure. Maybe. Sometimes I just say things I don't hear
@@northwestrepair You said before that you learned repair by watching all these repair videos. I swear you had some kind of an electronics background before you started.