• 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.

3080 / 3090 / 3070 Gigabyte Eagle Gaming OC & Vision Power Connector Concerns

Associate
Joined
24 Sep 2020
Posts
143
Yea It's not the GPU core temp but the card is thermally throttling and since it's a memory intensive load it should be VRM/GDDR6X temperature? I am betting on GDDR6X temps. The card boosted back to normal core clocks (temporarily) after pointing a fan to the backplate.

What makes you think that it's not issue with thermal pads on memory?
I'm not really sure its actually thermal capping was my point, do we know if PerfCap even considers the memory temp? Yeah Memory Clocks are low but so are GPU clocks so it doesn't seem like the GPU is struggling with whatever task you set it. The memory load is near 100% so its not going to get any higher than that anyway. I don't know maybe I've missed a key bit of information here I just don't see how you connected Thrml PerfCap to there being a problem with the thermal pads on specifically the memory. I guess the best option would be to establish some kind of bench mark for the memory test and get someone else with the same card to replicate and see if they get the same result.
 
Associate
Joined
7 Jul 2020
Posts
6
I'm not really sure its actually thermal capping was my point, do we know if PerfCap even considers the memory temp? Yeah Memory Clocks are low but so are GPU clocks so it doesn't seem like the GPU is struggling with whatever task you set it. The memory load is near 100% so its not going to get any higher than that anyway. I don't know maybe I've missed a key bit of information here I just don't see how you connected Thrml PerfCap to there being a problem with the thermal pads on specifically the memory. I guess the best option would be to establish some kind of bench mark for the memory test and get someone else with the same card to replicate and see if they get the same result.
Because the core clocks drop from 1200+ to 800-900 range and that's exactly the time gpu-z reports Thrm as perf cap which was otherwise pointing out Idle. Core clocks were set at 1200+ range.
I already tested it with a core intensive workload and it didn't throttle even past 58C.
 
Associate
Joined
21 Sep 2020
Posts
6
so after getting my 3080 back its running really badly, across multiple games. I did a full ddu before changing to the 3080. @GIGA-Man what should I do?

Edit: ah seems one of my ram sockets got damaged but getting the expected performance now soz!
 
Last edited:
Associate
Joined
26 Oct 2020
Posts
14
Hi,

I'm from France. I have received yesterday my brand new Gigabyte RTX 3070 Vision (s/n 2042) ... But sadly it doesn't work.
After many checkings and cables replacing, i see some issue on the 6 pin connector ... one of them is more pressed in than the others, as you can see on pictures below.



Then i found many pages on the internet about this issue ... As i can read on the first thread, issue seems to be still present on Gigabyte RTX 3070 ... Throught the backplate, it seems to be the old connectors type.

I made a RMA with my retailer , Materiel.net webshop (LDLC Group branch) ... Waiting for their reply early this next week ...

@GIGA-Man : Does Gigabyte have support for this issue with retailers or customers in France ?

Thanks for your replies and this helpfull thread !

Cordialement

JB

One of my 8 pin connectors (on the 3080 eagle) looked the same. Pin was pushed all the way inside. This is terrible that the 3070 has got this issue too.
 
Associate
Joined
2 Nov 2020
Posts
27
This is what I'm afraid of. I'm afraid of going through the effort of shipping my card out to them and them not doing anything to fix the issue and sending it straight back to me.

I'm not sitting here for 2-3 weeks without a GPU just to receive the same faulty GPU.

Are you in the UK? Seems wild that it wasn't handled properly over there. I thought for sure based on this thread the issue was being resolved appropriately over there.
Yeah, UK. I heard a bunch of good stuff about their RMA process. But it would seem my experience was contrary to that. I emailed them about the fact that the card was still faulty, they insisted that it was fine when they tested it, and suggested I either use a small screwdriver to straighten out the pins, or send it back to the retailer I got it from for a refund or replacement. So yeah, I'll be thinking twice before buying a Gigabyte GPU again in future. Which is a shame, since the 970 and 1080 G1 were both great and had no issues for me.
 
Associate
Joined
24 Sep 2020
Posts
143
Mine should be arriving today so fingers crossed I don't have any connector issues. I'll be doing a visual inspection when it arrives this afternoon and then try plugging it in this evening after work.
 
Associate
Joined
10 Oct 2015
Posts
218
Location
Essex
I think there's enough information coming in to safely say that this issue has not been addressed by Gigabyte. We really need some clear indication so the users can identify whether they have the modified connectors. This mish mash of cable ties and shrink wrap causes confusion and we don't seem to know what we've received.

I'm expecting delivery of my Vision tomorrow and will give an update once I have it in my hands.
 
Associate
Joined
23 Dec 2010
Posts
167
Location
Ireland
Thought I'd add my info to the other useful stuff here -
Received a week 43 card yesterday and was aware of the issues presented in this thread.
Card seems to have correct revised block around the power cable.
Power cable connected to the pins with no issues (the first time anyway - I had to disconnect them earlier to ensure card was postitioned correctly and when I tried reconnecting one of the pins was out of place and I had to move it with tweezers).

Card ran fine in Modern Warfare for a 1hr+ (full settings).
Card ran fine in Watch Dogs Legion until I reached the open world and it started crashing computer consistently after 2/3 minutes.
Card started crashing computer in Shadow Of The Tomb Raider after a few mins.
I can trigger a computer crash by running the SOTR benchmark, as soon as it tries loading the 2nd area it will crash.

After applying a -50 core clock using MSI Afterburner it has been able to finish the SOTR benchmark.
Have to test more...
 
Associate
Joined
24 Sep 2020
Posts
143
Thought I'd add my info to the other useful stuff here -
Received a week 43 card yesterday and was aware of the issues presented in this thread.
Card seems to have correct revised block around the power cable.
Power cable connected to the pins with no issues (the first time anyway - I had to disconnect them earlier to ensure card was postitioned correctly and when I tried reconnecting one of the pins was out of place and I had to move it with tweezers).

Card ran fine in Modern Warfare for a 1hr+ (full settings).
Card ran fine in Watch Dogs Legion until I reached the open world and it started crashing computer consistently after 2/3 minutes.
Card started crashing computer in Shadow Of The Tomb Raider after a few mins.
I can trigger a computer crash by running the SOTR benchmark, as soon as it tries loading the 2nd area it will crash.

After applying a -50 core clock using MSI Afterburner it has been able to finish the SOTR benchmark.
Have to test more...
This is post driver fix I presume?

Also my card has arrived. Pins look fine, will plug it in this afternoon and see if it works. It's a week 42 card.
 
Associate
Joined
25 Feb 2009
Posts
1,045
Location
GIGABYTE UK
Associate
Joined
25 Feb 2009
Posts
1,045
Location
GIGABYTE UK
I've talked to my HQ and they have assured me that all cards now produced are using only the new connector. If you do have any new cards from WK 40 onwards with issues please mail me with your serial number and images of the pins. I will make sure this is getting looked into.
 
Associate
OP
Joined
8 Jul 2013
Posts
2,089
Location
Middle age travellers site
Yeah, UK. I heard a bunch of good stuff about their RMA process. But it would seem my experience was contrary to that. I emailed them about the fact that the card was still faulty, they insisted that it was fine when they tested it, and suggested I either use a small screwdriver to straighten out the pins, or send it back to the retailer I got it from for a refund or replacement. So yeah, I'll be thinking twice before buying a Gigabyte GPU again in future. Which is a shame, since the 970 and 1080 G1 were both great and had no issues for me.

This is not good to hear at all . So just to clarify you had a problem with the 1st card you received.. then you RMA'd and now the replacement is having the same problem and now on top of that Gigabyte are asking you to straighten out the pins.

They should not be telling you to straighten out the pins with a screw driver on an expensive graphics card. It should have been fixed & the end USER the paying customer should not have to be bodging fixing things.

I would send it back and get a full refund at this point no matter how much you want the card and how difficult the cards are to buy.


Whichever option you go for hope this is sorted for you quickly.
I think there's enough information coming in to safely say that this issue has not been addressed by Gigabyte. We really need some clear indication so the users can identify whether they have the modified connectors. This mish mash of cable ties and shrink wrap causes confusion and we don't seem to know what we've received.

I'm expecting delivery of my Vision tomorrow and will give an update once I have it in my hands.

Totally agree with you on this. This whole thing is becoming a complete mess tbh .

I am that confused atm i do not even know what to put in the OP any more last thing i want to do is confuse peeps.

We do need more clarity re how does the paying customer identify if there on the new or old connector block. As it seems now there is many reports of cards over Wk39 and even cards that have been RMA'd and supposedly fixed coming back having the same problem.

And there is even reports of the 3070's being effected too.

I've talked to my HQ and they have assured me that all cards now produced are using only the new connector. If you do have any new cards from WK 40 onwards with issues please mail me with your serial number and images of the pins. I will make sure this is getting looked into.

Thanks for the Info so are we moving from Wk39 to Wk40 now of cards that could be on the old block ?

Also any update re the 3070's having the same connector block issues ?
 
Back
Top Bottom