Rtx 3080 lower quality capacitor Issue

[Chris Ecob]

Gigabyte is also using better capacitors which may mean it negates this issue for some reason, the gigabyte is using 470 and the zotac is 330.
Also gigabyte might have tuned there OC better than other AIB's so it doesn't really hit the dreaded 2Ghz.

I have a 3090 eagle OC arriving next week and i feel pretty okay with it, but time will tell!

I saw that too, 470uF will offer quite a bit more filtering than 330uF. Not sure how it would compare to a bank of MLLC's at 2GHz+ but perhaps this was Gigabytes solution to the problem instead of going to MLLC's like ASUS/EVGA.

My Palit has 1 bank of the smaller "more expensive" caps and it has not crashed yet.
To be fair though, the highest I have seen it boost is 1980, I might try a small OC or place an intake fan under it to see if it crashes above 2,000

If your hitting 1980 MHz and getting performance better than the FE then I'd be happy enough with that.

 

[Shiari]

I saw that too, 470uF will offer quite a bit more filtering than 330uF. Not sure how it would compare to a bank of MLLC's at 2GHz+ but perhaps this was Gigabytes solution to the problem instead of going to MLLC's like ASUS/EVGA.

And it's still not certain that this is actually the problem, or even part of the problem.

 

[kins888]

My guess probably just down to an early driver issue. If it's not, I think we should be happy with a 2000MHz clock - at the end of the day what's another few hundred MHz translate to in terms of FPS? I'd rather have a stable clock than be on the limit with occasional crashes. I know this is a OC forum but let's get real.

 

[Chris Ecob]

I have yet to see a 3080 fe which boosts to 2000mhz unless you overclock it

I think most of them are boosting to around 2GHz. I guess I'm more interested in actual benchmark figures but there's a direct correlation in clock speed vs performance when looking at the same architecture,

mine will set at 2020mhz playing hours of Call of duty.. on the cards out of the box settings (no overclock)

This is what I would have thought considering most cards are performing very similar to each other. They all seem to boost to around 2GHz but some are crashing for some reason. Could be the Caps or maybe it's just silicon lottery and the clocks have been set too high to begin with.

 

[SeeNoWeevil]

Because people have waited long enough for something they are paying good money for. If they finally get one and it crashes they could be at the back of the queue waiting for a replacement.

It’s not acceptable, which is why people are cancelling until the problems are solved.

Don't get me wrong, if my card crashes on stock settings, I am going to be ******. However, RMAs go to the front of the queue (rightly bloody so!) so you're not losing anything by getting it and testing it. What if it turns out to be a driver issue and a new revision fixes it and you cancelled your card and are now at the back of the queue?

 

[ZedEx48K]

Ok, been lurking here for a while and thought I would throw my 2p in.

Gigabyte Eagle card, performance has been rock solid so far, a mixture of gaming and game benchmarks etc, I did also buy a new 750w Gold rated PSU to go with it. I have not ventured anywhere near OC'ing it past what it came with, the leap from a 1080 card is more than enough for me to not waste my time arsing around for a few more fps.

 

[azazl187]

Ok, been lurking here for a while and thought I would throw my 2p in.

Gigabyte Eagle card, performance has been rock solid so far, a mixture of gaming and game benchmarks etc, I did also buy a new 750w Gold rated PSU to go with it. I have not ventured anywhere near OC'ing it past what it came with, the leap from a 1080 card is more than enough for me to not waste my time arsing around for a few more fps.

Dont you want to try just so you know that it can handle it? It would also be helpful to others

 

[el-nino]

Does anyone know if the msi gaming trio is effected by this issue? Jayz2cents shared a tweet from EVGA who said they became aware of this during testing and upgraded their cards that's why they're not available yet. I doubt other manufactures will I think they will do bios fixes limiting the cards.

 

[Locky]

Dont you want to try just so you know that it can handle it? It would also be helpful to others

I pushed mine a bit today no crashing. In fact ive not had a single crash since owning it but my system never crashed with my 1080ti either.

This this post when I overclocked it a bit.

https://www.overclockers.co.uk/forums/posts/33990945/

 

[ZedEx48K]

Not really, CPU overclocking I'm ok with, but for now I will ride this out as every time a theory pops up, it gets disproven, far too early to say what the issue is, and if it does end up affecting my card, I'm sure an RMA will follow.

 

[Joxeon]

Because people have waited long enough for something they are paying good money for. If they finally get one and it crashes they could be at the back of the queue waiting for a replacement.

It’s not acceptable, which is why people are cancelling until the problems are solved.

Doesn't cancelling put you at the back of the queue anyway?.

My guess probably just down to an early driver issue. If it's not, I think we should be happy with a 2000MHz clock - at the end of the day what's another few hundred MHz translate to in terms of FPS? I'd rather have a stable clock than be on the limit with occasional crashes. I know this is a OC forum but let's get real.

50mhz means about 1fps.

 

[azazl187]

Not really, CPU overclocking I'm ok with, but for now I will ride this out as every time a theory pops up, it gets disproven, far too early to say what the issue is, and if it does end up affecting my card, I'm sure an RMA will follow.

Im sure that it will, but better to know now and RMA than a few months down the line, dont you think? Although getting a replacement may not be easy

 

[azazl187]

I pushed mine a bit today no crashing. In fact ive not had a single crash since owning it but my system never crashed with my 1080ti either.

This this post when I overclocked it a bit.

https://www.overclockers.co.uk/forums/posts/33990945/

Looks like you got a good one there

 

[Chris Ecob]

Ok, been lurking here for a while and thought I would throw my 2p in.

Gigabyte Eagle card, performance has been rock solid so far, a mixture of gaming and game benchmarks etc, I did also buy a new 750w Gold rated PSU to go with it. I have not ventured anywhere near OC'ing it past what it came with, the leap from a 1080 card is more than enough for me to not waste my time arsing around for a few more fps.

This is great info thanks. Would be cool to know what OC/boost clock you can get before it crashes if your happy to overclock.

I pushed mine a bit today no crashing. In fact ive not had a single crash since owning it but my system never crashed with my 1080ti either.

This this post when I overclocked it a bit.

https://www.overclockers.co.uk/forums/posts/33990945/

Looks like you got a good Gigabyte Gaming OC there mate. I'd be well happy with 2130mhz. Surely can't be an issue with the Caps if your achieving those clocks. Silcon lottery and Nvidia/AIB's setting too higher boost clocks for the average silcon quality is my bet.

 

[ZedEx48K]

Im sure that it will, but better to know now and RMA than a few months down the line, dont you think? Although getting a replacement may not be easy

If I panic and return it now, what am I left with? The 970 nicked from my media PC If in a few months it turns out there were issue it will get replaced anyway, be it shop or direct with Gigabyte, and by then the issues will have been ironed out. If the card dies on it's arse at stock, then I will be straight onto the case to return it, but for now I won the release day lottery and may as well enjoy it.

 

[Lastof]

Does anyone know if the msi gaming trio is effected by this issue? Jayz2cents shared a tweet from EVGA who said they became aware of this during testing and upgraded their cards that's why they're not available yet. I doubt other manufactures will I think they will do bios fixes limiting the cards.

Right now, not really.

• We know the review and early shipped units of the Trio has a 5-1 arrangement of the larger capcitors to 10 smaller capacitors.
• We know EVGA said they found that for their FTW3 card they needed a 4-2 arrangement, but that their XC3 was fine with 5-1.

• We don't know if the current Trios are shipping the same as the initial ones, or if MSI picked up on the issue and changed at some point (which might explain the nearly 0 numbers of them that seem to be out in the wild).
  
• We don't know what exact difference between the XC3 and FTW3 means one was fine and the other needed more, and how that translates to the Trio.
  
• We don't know that the capacitors are the only issue, since some people are reporting to have similar problems with suposidly "good" layout ASUS Cards and FE Cards

Its a developing issue, and while it seems fairly certain that the capacitors are part of what is causing it, it might be more than that.

Personally I suspect it'll turn out to either be the 30% "ok" Bin 0 chips or the 10% "really good" Bin 2 chips that are susceptible this issue (either because they can't stay stable at higher boosts, or can boost higher and draw more current which the power delivery can't handle). EVGA just said they found the problem in testing, not that it affected all the cards, and from what we've heard AIB's didn't have time to actually Bin the chips themselves before starting manufacturing.

 

[DFroN]

I think that the capacitors are an indicator of the problem rather than the cause. A card with high quality caps is going to have higher quality VRM and power delivery overall and with it a better chance for higher clock speeds. I think that the AIB’s have clocked the cards too high and didn’t have adequate time to validate them else they never would’ve released how they are.

There are reports of TUF cards crashing with all their ‘good’ caps but it’s all second hand forum posts, I haven’t seen a first hand post by somebody claiming to have one. And again, they have better VRM, not just better caps, so it shouldn’t be surprising that they can clock higher on average. I haven’t seen any FE owners claiming crashing but there’s not many of them in the wild yet and they don’t clock to 2Ghz anyway unless you overclock them, and then you’d just think you’ve reached your max OC and back it off without thinking about capacitors.

I have a 3080 Gaming Trio X and it crashes occasionally in some games. -30Mhz clock seems to have fixed it for me, the card boosts to 1980Mhz now and stays there. There’s no performance loss because it would only boost to 2010 stock momentarily and then come back down to 1950-1980 anyway, but it would suck if you have a bad chip that only does 1865 or something. But that’s the same silicon lottery for any card.

I’m sure there will be bios updates to limit boost but if you can find your maximum stable clock at say 1980 for me and the new bios limits it below that then you’ll be better off keeping your bios and manual underclock.

 

[Jlang]

Haven't read through the whole thread here, but this link at Reddit has a good set of tables and links to what cards have what POSCAP/MLCC's fitted - https://www.reddit.com/r/hardware/comments/izmi1k/ampere_poscapmlcc_counts/
My Palit (non OC ) is doing great I suspect it has the exact same board as the OC version - https://www.techpowerup.com/review/palit-geforce-rtx-3080-gaming-pro-oc/3.html which follows the "reference" design of 1 MLCC on the board.
Anyone having problems I'd do as Jay says in his video and use afterburner or similar to slightly underclock - this will hold you over till the AIB's get a new VBios out for you that should get around the problem without I'm reasonably sure losing any performance.
It's very, very small amounts of electrical "noise" behind this, so a small tweak here and there should solve it - nearest example I can think of is the tweaks Apple put in there OS that fine control voltage such that the same components give better performance under their os when compared with bootcamp'd windows on the same hardware.

 

[Jlang]

Does anyone know if the msi gaming trio is effected by this issue? Jayz2cents shared a tweet from EVGA who said they became aware of this during testing and upgraded their cards that's why they're not available yet. I doubt other manufactures will I think they will do bios fixes limiting the cards.

I'd be pretty sure the MSI gaming Trio will be fine it has 1 MLCC and 5 POSCAP - that's the reference design and should be fine - it's the cards that have gone all POSCAP that are "potentially" having a problem.

 

[SeeNoWeevil]

There are reports of TUF cards crashing with all their ‘good’ caps but it’s all second hand forum posts, I haven’t seen a first hand post by somebody claiming to have one. And again, they have better VRM, not just better caps, so it shouldn’t be surprising that they can clock higher on average. I haven’t seen any FE owners claiming crashing but there’s not many of them in the wild yet and they don’t clock to 2Ghz anyway unless you overclock them, and then you’d just think you’ve reached your max OC and back it off without thinking about capacitors.

I'd be pretty sure the MSI gaming Trio will be fine it has 1 MLCC and 5 POSCAP - that's the reference design and should be fine - it's the cards that have gone all POSCAP that are "potentially" having a problem.