According to the reditt post by some engineers POScap cost more than MLCC.
You've linked to a good post, but then oversimplified it. It's too early to draw definitive conclusions on this topic.
Most people don't understand that more expensive components does not necessarily mean better. It means that it is more expensive to produce a component that has the characteristics that it has. If you want those characteristics then yes, more expensive = better. If you want other characteristics then more expensive = worse.
The problem is supposedly triggered by a POSCAP design + Greater than 2 GHz clocks.
Anecdotally:
The Zotac has the POSCAP design but low clock rates. Problematic card.
The MSI has a mixed design but high clock rates. Problematic card.
The EVGA has the POSCAP design but high clock rates. Problematic card.
The ASUS has the MLCC design and high clock rates. No real reports of problems.
The Gigabyte has the POSCAP design but high clock rates. No real reports of problems.
Doesn't quite align with the theory.
That could very well be the case. It is worth noting GN were using a POSCAP RTX 3080 FTW3 when they were running their LN2 live overclocking stream. Granted though they were doing it in short bursts.
https://youtu.be/fMYZJvP8Sp0?t=419
Yea that's interesting, seems Igorslab (Jayz source) are perhaps being overly simplistic, and there may be other PCB design factors to take into account here.
And this ladies and gentlemen is why you never ever early adopt any tech, always wait for the reviews/benchmarks and give it some time and then make your decision.
So glad I never got a hold of one of these and will also pass on the 3070 until we know more about is going on here.
I'm leaning toward driver issues because the majority of times when dealing with a brand new architect, GPU's etc it's more likely driver issues.
And this ladies and gentlemen is why you never ever early adopt any tech, always wait for the reviews/benchmarks and give it some time and then make your decision.
95% of people are still waiting for cards
95% of people are still waiting for cards
And now they can cancel with any luck and wait this out. This launch is now officially the worst GPU launch of all time. I honestly cannot remember one as bad as this shit show.
I am also so glad that I didn't get one of these cards and when I get one I'll be so happy that I got one. Because that's how it works with some people :p
I'm leaning toward driver issues because the majority of times when dealing with a brand new architect, GPU's etc it's more likely driver issues.
Yea, my best "guess" is some of the AIB's designs are boosting the frequency just too high for the quality of the specific silicon sample in the card, and the "correct" fix is simply adjust the max boost down a fraction with firmware. This would still mean anyone can OC back up if they feel their sample is capable of better, so no drama there.
With EVGA going with the all POSCAP (aka the supposedly cheap option) design on premium FTW3 model but not on the cheaper XC3, its "literally" the opposite of what we would expect based on "igorslab" theories, which is what people need to remember its just one theory atm.
This launch is now officially the worst GPU launch of all time. I honestly cannot remember one as bad as this shit show.
Correct me if I am wrong but the 2080/2080Ti launch was also pretty bad. I seem to remember stock shortages, overheating issues, and driver instability in many games.
With EVGA going with the all POSCAP (aka the supposedly cheap option) design on premium FTW3 model but not on the cheaper XC3, its "literally" the opposite of what we would expect based on "igorslab" theories, which is what people need to remember its just one theory atm.
Definitely!
There haven't been any widespread reports of any issues with any Gigabyte models, which is using the POSCAPS. I think the issue is likely way more complicated than is being reported.
These outlets can only speculate based off of anecdotal evidence without proper testing, but they have to generate clicks and content to attract viewers.
Here is a reddit post clarifying some misinformation about the whole capacitor thing: https://www.reddit.com/r/nvid
From EVGA: 3 minutes ago:
https://forums.evga.com/Mess
Message about EVGA GeForce RTX 3080 POSCAPs
Yea, my best "guess" is some of the AIB's designs are boosting the frequency just too high for the quality of the specific silicon sample in the card, and the "correct" fix is simply adjust the max boost down a fraction with firmware.
I think nVidia has set an unrealistic standard for AIBs with the price, material quality and design of their FE cards. It has squeezed AIBs into a corner, forcing the clocks up so they don't fall behind the base price product without the necessary additional design and validation required to do it reliably.
The first party cards are beyond reference quality being sold at reference price which is a bit unfair to the partners.
The backlash on reddit when Zotac customers saw they were getting a card that performed a few % worse than FE at a higher price is an example of this.
Check PCCG websites on the 3 Asus rtx 3080 , they all use MLCC.
This photo is from PCCG's website for the Asus 3080 Tuf 10GB. It shows poscaps
Message about EVGA GeForce RTX 3080 POSCAPs
Woah! Well there you go. I wonder if other AIB's will release a statement.
So what does that say about the EVGA cards that supposedly had the issue?
Is it just the FTW3 series that have the update?
From EVGA: 3 minutes ago:
https://forums.evga.com/Mes
Message about EVGA GeForce RTX 3080 POSCAPs
Oh boy, this is heating up quickly. I think EVGA just got a HUGE reputation boost from this announcement.
From EVGA: 3 minutes ago:
https://forums.evga.com/Mes
Message about EVGA GeForce RTX 3080 POSCAPs
The question still remains though why are people reporting the crashes, one of the boards in question is the EVGA GeForce RTX 3080 XC3 GAMING.
Maybe just an issue with the drivers or something else I don't know.
The question still remains though why are people reporting the crashes, one of the boards in question is the EVGA GeForce RTX 3080 XC3 GAMING.
There is going to be a lot of noise in the current data. It may turn out that EVGAs design in the wild still causes the occasional issue or it might be other factors in the reporters computer causing the crash.