• 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.

Rtx 3080 lower quality capacitor Issue

killari said:
No one has promised that the GPU should run at ~2000+mhz clock.
Click to expand...
No, indeed they have not.
However, if all the reviews were done with cards which boost to around 2 GHz and people have bought them based on their performance in those reviews, then if a driver fixes the issue by down-clocking...
... We'll I would return the card.
Similarly, if CPU manufacturers sent out only golden samples to reviewers whose distribution in normal samples is 1 in 1,000 then normal consumers wouldn't be impressed and any half-way honest reviewer would call them out.
 
killari said:
No one has promised that the GPU should run at ~2000+mhz clock.
Click to expand...

My issue (and its not one that will get me to cancel my order or anything, is that I would be fine if it boosted to 1740 (OC) and that was it, but this is a built in card/driver feature that will attempt to boost more and crash to desktop.
A bios update sounds like it will solve it by setting conditions, but of course this all may end up being some driver issue and all a red herring, or literally not impact me at all!
 
https://www.reddit.com/r/gigabytega...ent_the_cheap_way_with_the/g6lggto/?context=3
This seems to be also interesting. POSCAP capacitors are not the same, and manufacturers are building different capacities on the cards. There are so many constellations of POSCAP only and POSCAP+MLCC. Furthermore also different qualities on MLCC.

I don't think that anybody, as long s/he is not an hardware engineer, knows about power delivery networks and additionally about the 30x0, can properly comment on this topic and say what the best constellation is to avoid the problems we see.

But it's a fact that there is a problem. Who knows how they are going to solve this. Most likely with firmware and driver updates which would be the simplest solution.

I think this will really end up in a marketing disaster although the solution might be even simple.
 
ubersonic said:
Reviews are started weeks before launches using pre-release drivers, cards crashing to desktop during testing is actually normal hence why none of them would have thought to mention it.
Click to expand...

They should do the same as smartphone reviewers, they disclose it but will usually add a comment as to what they think.

If I was a reviewer I would certainly mention it, if it got me struck of the review list so be it, better reviews should be from retail samples anyway.
 
I'm not sure why some people seem to think that if they don't manually overclock they won't have this issue. That's obviously not the case.
 
whole lot of people pretending to know about computer parts at the moment. Think I'll wait until somone smart says something personally.
 
Devrij said:
From buildzoid video, there's basically nothing that says "spcaps are bad". They have higher impedance and their peak performance is at a lower frequency, making them a little less responsive at higher frequencies compared to MLCCs. MLCCs on the other hand have worse longevity due to the piezo electric effect essentially wiggling the hell out of them all the time they're in use (hence the talk of cracking).

They're both perfectly good components, but the question is how they interact with all the other power regulation on the PCB which is way more complex than most of us are able to figure out even if we had the whole pcb stack diagram in front of us.

So yeah, until someone who actually has all the info says something, I'm withholding judgement.
Click to expand...

Thank god someone on here talking sense.
These capacitors are good quality caps, it is more about choosing the correct characteristics of components for the design, voltage, tolerance frequency range, temperature etc. They went with the SPCAPS as in mass production time is money, placing one big component in the place of 8 caps saves time on the pick and place machines at the factory. My job is a PCB design engineer and I design the loadboards for bga, fpga and other types of chips for companys.

I know from experience the customer will produce design guidelines to follow. What I have seen it seems they should have had 5 SPCAPS and 8 MLCC caps minimum.
 
Last edited:
Its probably just the turbo boots getting out of hand on immature drivers.

Either more stable drivers or a more limited turbo will solve the issue.

No card promises 2000mhz boost anyway most cards are usually in the mid 1900's most of the time.
 
https://www.reddit.com/r/hardware/comments/j09yj5/poscap_vs_mlcc_what_you_need_to_know/
Nice read, also further down the comments. I think the capacitor discussions are too superficial. What might solve the issue on one manufacturer (Zotac), might not be the solution for others.
Root-causes are still unknown. At least for us. And perhaps it's also better as that are deep topics.

Solutions and the impact of them are of interest. Let's cross the fingers.
 
MagicBoy said:
That's false. The 970 had 4GB, however the last 512MB was slower to access due to the cut down number of shaders and the memory controller design compared to the 980.
Click to expand...

The courts in the US disagreed and ordered Nvidia to compensate 970 buyers as it was miss-sold.

The last 512mb was useless as it completely tanked performance once you hit it.
 
Has anyone ever considered it may be silicon issue ie the GPU units themselves. The mass production may not have yielded as many chips that are capable of 2GHz+ boost. Where the reviewer samples are likely early production which are binned to somewhat?

anyway buildzoid video did dispelled a few incorrectness POSCAP and SPCAP naming etc. But without knowing the specifics of power delivery it is not possible to say these variation is definitely the cause. It “may” be. But again there are so many things going on in a GPU any one of them or a combination of things can be contributing to crashes. Plus AIB with different filter cap design is fairly standard as buildzoid pointed out.

tho the buildzoid video is unnecessarily long.

wait and see, most likely situation is that a firmware update to limit the boost clocks to no more than 2GHz so no matter what happens your card will not result in crashes. And if you want to risk it then manual overclock to see what happens.
 
Maybe a driver update will reduce boost clocks on all non FE cards

Really unless an OC or FE varient shouldn't all 3080s be effectively the same ?

Else all the reviews should be redone ?

Also a benchmark score isn't really valid if it is not stable ? I mean I could run my GPU or CPU at stupid clocks with chance I might get through the level being benched or the benchmark but if crashes in games if actually played for a time they are not valid benchmarks are they ?
 
Last edited:
CCC said:
I wonder how none of the all the super experienced hardware reviewers found this issue?
Click to expand...

I think there is more to it than just capacitor selection - otherwise the reviewers would have wondered at all the crashes in their testing. It is possible it is a factor in cores that are already marginal for the binning.

pc-guy said:
Has anyone ever considered it may be silicon issue ie the GPU units themselves.
Click to expand...

Personally I think it is quite likely a major factor is that the binning isn't good enough and there are some cores of marginal stability in there.

EDIT: Also possible they just screwed up on the boost algorithm and can fix it in the drivers.
 
Last edited:
CCC said:
I wonder how none of the all the super experienced hardware reviewers found this issue?
Click to expand...

because they have press drivers that are not the final release drivers. I seen a few reviewers state they generally have issues during testing but amd/Nvidia/intel iron out the problems before release so it’s no reported. Also they all wouldn’t have any the issue, I know 3 people with 3080’s already and none have had any issues.
 
You must log in or register to reply here.
Back
Top Bottom