News Igor's Lab uncovers 'hotspot issue' affecting all RTX 50-series GPUs — says it could compromise graphics card longevity

Page 3 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
I bought the big box of popcorn from Sam's Club to watch the 50 series saga continue to play out, but I think I'm going to run out, lol. 🤣
 
I'd like to see some thermal images of equivalent, 30 and 40 series cards as a control. The 4090 is mentioned, but no real world data. Still, very interesting about the thermal design guide.
True, but also there’s no such thing as a Ryzen 5 5800x3D. Check your signature.
 
I always laugh at Nvidia fanboys talking about AMD GPUs having bad hotspots just because AMD actually gives you the hotspot temp and Nvidia doesn’t.
I always laugh at people who call other people fanboys because they dare saying something bad about their favorite brand. And what that makes you? Oh yes, an AMD fanboy!
 
  • Like
Reactions: KyaraM
I always laugh at people who call other people fanboys because they dare saying something bad about their favorite brand. And what that makes you? Oh yes, an AMD fanboy!
He is also wrong claiming no Nvidia cards does this, considering that both my 4070Ti and mobile 4060 do show hotspot temps... not the ones from the article, mind, but neither does AMD, so the point is very moot indeed.
 
Tell me in these pics where the power delivery system is located?
https://www.techpowerup.com/review/asus-geforce-rtx-5080-tuf-oc/5.html

On this page they all look to be under the heatsink
https://www.techpowerup.com/review/asus-geforce-rtx-5080-tuf-oc/4.html
you have be know that beside the actually running chips like the GPU die and the memory, the power delivery circuits do get hot as they deliver those 300+W power throught their circuits. although thte VRMs are usually attached to the heatsink via thermal pads, those are usually quite thick ones (1-2mm) and is much less effective than the thermal paste on the GPU.

Although those components are rated to work at 150C, they usually have a spec on heat vs life expectancy, like in caps, you will always want to run the 105C rated caps at much lower temp to let them survive expotentially longer, much like a car engine is rated to work till the red line, but if you kept it on red line good luck it will last you a few years.

Will it break your card instantly or really quickly? highly likely not.

But will it last like 3-5 years or more? it will be really skeptical, some caps at proximity could likely be baked to have much lower lifetime.
 
  • Like
Reactions: Moonstick2
Igor also shows that there is a difference when placing ThermalPads in the appropriate place, with obvious difference.

Unfortunately, the average Gamer hardly has the opportunity to measure this in his System itself,
such Thermal-Vision Devices are very expensive

And Nvidia is currently ignoring this

The best thing is, I just bought a 5070 a few days before Igor's video
 
I am sorry most youtube and madia ppl are now using click bait naming artciles this is not the issue of 5xxxx its the PNY issue of bad cooler which do not provide good cooling to the VRM, most cheap AIB and basic GPU have issue where VRM on load goes to 100+ not only for GPU but for many mainboards also, sure its not gr8 best to keep it up to 85C max

But we had this on most cheap sold cards, also gpu hotspot sensors in GPU HAVE NOTHING TO DO WITH VRM temperature and only if AIB put sensor next to VRM it would show its temperature!!!!

It would be nice if Igor was not mad or just want to earn some $$$ on clickbait article this is AIB issue and not 5xxx series issue
If he want to trash NV and AMD and AIB he can do 1000+ card review and find most cheap ones VRM is not cooled good, let say MSI ventus 4070 they had no pad on choke so choke goes to 100C, etc
 
Last edited: