• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

EVGA Pascal cards hotspot problem

I'd recommend reading through this thread from EVGA forums.

"The hotspot issue is a misconception based on a review where the reviewer in question was running Furmark, an extreme usage case, as most overclockers know. We believe this is a good approach to have some idea about the graphics card limit, and the thermal performance under the worst case scenario. EVGA has performed a similar qualification test during the design process, at a higher ambient temperature (30C in chamber) with a thermal coupler probe directly contacting the key components and after the Toms Hardware (Germany) review, we have retested this again. The results in both tests show the temperature of PWM and memory is within the spec tolerance under the same stress test, and is working as originally designed with no issues.

As for your card, it is not affected, as the "issue" has only been found to affect the FTW models, which are a non-reference PCB. Your card, the SC, and other reference PCB models do not exhibit this "hotspot". However, if you wish to obtain the thermal pads for your card, you may do so, as long as your product is ACX 3.0 and has a backplate. To request these, please go to ..
."

http://forums.evga.com/Is-the-EVGA-1070-SC-effected-by-the-VRM-overheating-issue-m2571376.aspx

Not checked in to see how this issue is playing out for a few days. I get what evega is saying. Would be nice to see some more information though. I have a classified rather than an ftw.

Last evega thread I read all I got from it was that the Evega mod. did not like the term problem as it indicated an issue.

Today I learn it also does not like the word affected.

Question: "Is the 1080 classified affected by this"

Answer: "Fortunately none of the cards are effected by an overheating/VRM issue."

I think I would have run with none of our card are effected by this but we are offering thermal pads for the 1080 classified so our custmers can enjoy lower temps.

Not learning anything other than Evega appear to be using The Parniod Troubleshooters Guide To The English Language. Which only seems to give a single worse case reading for each word.

Does not look like customer support it reads like damage limitation.
 
Well i think i may have fixed my cold boot issue through trial and error...

Disabled PLL overvoltage a few days back and it hasn't reset on cold boot since.

Of course this may be just a coincidence and it may be too early to say it's sorted.

Same with disabling quick startup, I wont know until more time has passed to be certain. Will keep you updated, hopefully we both have one of the issues resolved!

There's a few AIB cards that cool the VRM's with just a thermal pad contacting the fins of the cooler. The MSI Gaming X for one.

I think you will find that there are also going to be thermal pad/s fitted to the back of the PCB that then touch the backplate :)

Thanks for the updates, good to know and hopefully it will bring cooling in line with the competition.
 
I talked to EVGA and they said this hotspot issue has been blown up more then it should. They told me there is no need to replace GPU. If I really really wanted to they would, but they said no need at all. If your worried then put on the thermal pads. I said I never put pads on before. He told me it's only 4 screws and unplug led that's it. They'll have step by step with pads.They haven't sent out pads yet, but should do shortly for people that wanted them otherwise there is no need. I have 3 year warranty either way
 
Let's not make this AMD vs Nvidia, there is a difference here that it is a third parties actions in board design/cooling not the GPU manufacturer.

Out of those three paragraphs, the first one advised to get any card with this issue RMA'd ASAP (Due to some people playing down the issue). The second was about the fact that I like EVGA as a company and was surprised to see them drop the ball.

The third paragraph never mentioned Nvidia and was a statement more about the fickle nature of the Graphics card forum threads and it's clientelle.;)

I wouldnt wish this kind of issue on anyone regardless of choice of graphics card, hence my advice in paragraph one to get them RMA'd ASAP.
:)
 
So just tried calling evga on the below number as per an official rep post. The number is stated as being direct for UK customers... It isn't. Option 1 for English goes to Germany. Hopeless :D

This along with 24/7 EVGA Support, extended warranty services and the award winning EVGA Step-Up Program make it easy to see why EVGA support is considered among the best in the industry. EVGA customers located in United Kingdom can call 01788 247 298 for direct support from today onwards."
 
I talked to EVGA and they said this hotspot issue has been blown up more then it should. They told me there is no need to replace GPU. If I really really wanted to they would, but they said no need at all. If your worried then put on the thermal pads. I said I never put pads on before. He told me it's only 4 screws and unplug led that's it. They'll have step by step with pads.They haven't sent out pads yet, but should do shortly for people that wanted them otherwise there is no need. I have 3 year warranty either way

That's not true as to get a good contact and best possible temperatures you will need to replace the thermal paste on the core. If however it is just 4 screws and with the rest of the components so well protected due to the plate, I'll probably do it myself as its a 30 minute job at most. Might even get some lower core temps.

It's surprising that applying a thermal pad in a strip to the 12 vrm that that sits against very thin pieces of heatsink fin actually does much. The heatsink ideally wants a backplate attached to the fins in this area.
 
Last edited:
I talked to EVGA and they said this hotspot issue has been blown up more then it should. They told me there is no need to replace GPU. If I really really wanted to they would, but they said no need at all. If your worried then put on the thermal pads. I said I never put pads on before. He told me it's only 4 screws and unplug led that's it. They'll have step by step with pads.They haven't sent out pads yet, but should do shortly for people that wanted them otherwise there is no need. I have 3 year warranty either way

There are 2 cables to unplug from the headers on the PCB. The LED and the fan. Just to let you know however, that I don't think it is as simple as they have made it out to be and here is why..... the cable length is EXTREMELY short. I will take a pic of my ACX that I have removed from my 1080 SC and post it to show you.

What I am getting at, it when it comes to putting it back together, you need to re-attach these cables to the headers on PCB 1st. Then position the cooler, place it on and screw it in place. Now I might be very wrong on this and it is a very easy task. However, (and I have loads of experience of removing/ replacing coolers) it looks like it would be very awkward.
 
wmfn2r.jpg



2z3whnd.jpg


Apologies for quality of pics. That is the cables fully extended, so you will have to have the cooler pretty much seated to be able to fasten those to the headers. Like I said, it looks like it would be extremely awkward and frustrating.
 
Also with regards to the above video, he appears to turn the pc on then it flames up, so I'm guessing this can't be an overheat problem as the card hasn't had time to get hot?

Anyway I'm not sure what to do or what to believe about this whole saga. I got an EVGA 1070 SC back in August and it's been working fine since then. However there is clearly a problem as tests are showing pretty nasty high VRM temps on all the ACX 3.0 cards.....

I've ordered the thermal pads and will apply them. I did notice that I can extend my warranty up to 5 years or 10 years if I pay some extra cash. The warranty extension can only be bought within 90 days of the purchase date though, so I only have a few days to decide on that...

Might be worth emailing EVGA and seeing if they will extend me to 5 year warranty for free.

Certainly worth a go :)
Let us know how you get on.

Remember though... the warranty extension is only for the initial purchaser.
IE. Come re-sale time, only the original warranty is transferable.
 
Yeah will do mikeo. I've sent them a message. TBH a 5 year warranty is pretty much all anyone would need.

If I had a 10 year old card today it would be something like a 7800gtx with 512mb. So come 2026 my 1070gtx will be ready for the bin anyway :D
 
There are 2 cables to unplug from the headers on the PCB. The LED and the fan. Just to let you know however, that I don't think it is as simple as they have made it out to be and here is why..... the cable length is EXTREMELY short. I will take a pic of my ACX that I have removed from my 1080 SC and post it to show you.

What I am getting at, it when it comes to putting it back together, you need to re-attach these cables to the headers on PCB 1st. Then position the cooler, place it on and screw it in place. Now I might be very wrong on this and it is a very easy task. However, (and I have loads of experience of removing/ replacing coolers) it looks like it would be very awkward.

You are right. It is a bit fiddly. But in my opinion, it's going to be a lot easier to attach the two connectors "after" fitting the cooler back on and screwing that down. You do though need a decent pair of small long nosed pliers to do this. Use these to position each connector and push it carefully down with a small screw driver. Not done it on this card but it looks very similar to an MSI card that I removed and re-fitted the cooler on (applied better TIM).

I'll be interested to see their instructions when they get put up (hopefully later this week). If the pads are the same thickness as what I've got already, I might just push on and do it ASAP.

EDIT. Here you go. I thought by looking at the card, that you should be able to reconnect those fiddly little SOB's after refitting the cooler.
https://www.youtube.com/watch?v=N-brze9p_9E ... about 10 mins in.
Be interested to see what EVGA recommends.
 
Last edited:
You are right. It is a bit fiddly. But in my opinion, it's going to be a lot easier to attach the two connectors "after" fitting the fan back on and screwing that down. You do though need a decent pair of small long nosed pliers to do this. Use these to position each connector and push it carefully down with a small screw driver. Not done it on this card but it looks very similar to an MSI card that I removed and re-fitted the cooler on (applied better TIM).

I'll be interested to see their instructions when they get put up (hopefully later this week). If the pads are the same thickness as what I've got already, I might just push on and do it ASAP.

Yeah you might actually be right, and a decent set of long-nosed would be the least awkward method. There are decent cut outs in the fin area so should be able to access that area with tools reasonably well. That is the shortest length cables I have ever seen on a cooler though. I will have to reattach it at some point when it comes to selling the card as I am hopeful that I might be able to recycle my Hybrid on to the next generation card (assuming Nvidia dont change a) the mounting mechanism or b) the shroud design)
 
wmfn2r.jpg



2z3whnd.jpg

Apologies for quality of pics. That is the cables fully extended, so you will have to have the cooler pretty much seated to be able to fasten those to the headers. Like I said, it looks like it would be extremely awkward and frustrating.

Cheers for that Five.Stars. I hope instructions are easy to follow on installing pads because I've never done it before.
 
Hey Five.Stars,

I was considering the Hybrid kit myself, do you think it is worth the money?

It depends really. Yes it keeps the temps really low. Highest I have seen is 47c. However, it is fairly expensive (£120 plus P&P), on top of an already expensive card. It has increased my ability to overclock marginally and keeps the clockspeed much more consistent. I can now get close to 2100Mhz (not quite) when previously that was extremely distant. It now settles at 2050-2063Mhz where previously it was around the 2025Mhz. Nothing earth shattering by any stretch of the imagination.

Also something to take in to consideration is you may well see higher temps than this as the stock radiator fan is non controllable via software. It is only the blower fan on the shroud which is controllable. I have my rad fan (swapped it for a Scythe GT) plugged in to my Corsair Commander Mini which then adjust the fans speed based on temps via Corsair Link. The fan can also be plugged in to a motherboard header and adjusted either via the motherboards software, or preferably a program like speedfan. I believe plugging the rad fan in to the attached header spins the fan at a constant speed, although I can't 100% confirm this as I have never used the attached header.

Personally I love it, as having previously been a watercooler this gives me a number of the benefits, with much less of the work/ maintenance involved.

Installation is very simple, although routing the cables and then being able to seat the shroud properly to enable you to screw it in to position is a challenge. It probably didn't help that I had an extra cable inside the shroud that normally would not be there (the fan cable).
 
Last edited:
You are right. It is a bit fiddly. But in my opinion, it's going to be a lot easier to attach the two connectors "after" fitting the cooler back on and screwing that down. You do though need a decent pair of small long nosed pliers to do this. Use these to position each connector and push it carefully down with a small screw driver. Not done it on this card but it looks very similar to an MSI card that I removed and re-fitted the cooler on (applied better TIM).

I'll be interested to see their instructions when they get put up (hopefully later this week). If the pads are the same thickness as what I've got already, I might just push on and do it ASAP.

EDIT. Here you go. I thought by looking at the card, that you should be able to reconnect those fiddly little SOB's after refitting the cooler.
https://www.youtube.com/watch?v=N-brze9p_9E ... about 10 mins in.
Be interested to see what EVGA recommends.

Cheers for the video post. Pads ordered.
 
Same with disabling quick startup, I wont know until more time has passed to be certain. Will keep you updated, hopefully we both have one of the issues resolved!

Good to hear :)

I think this board does suffer with cold boot issues having been googling it for a while looking for answers.

The cause just seems to be random, down to the individuals hardware config, overclock and settings.

Mine hasn't reset on boot for quite a while now so for me, it seems the disabling of pll overvoltage has done the trick.

Glad the issue was not related to the video card directly.

Still waiting to hear back from evga regarding the hot spot issue though.
 
What is the "cold boot" problem and what does it mean exactly that the card resets?

When turning on the PC my card (1080 SC) seems to start with 100% fan for 1-2 seconds before going into fan stop mode. Is this related?
 
Back
Top Bottom