Pilot 53

R3L3NTL3SS

SprayingMango
This is cute and all but ASUS is shipping cards with 6 MLCC, which seems like the best solution. Why isn’t EVGA doing the same?

Why are you all stuck on this when multiple people have said 6 MLCC are actually not ideal?

I haven't heard this yet, do you know what the reason is?  Is it less overall capacitance? 

No clue. But people with much more knowledge on the subject than me have stated it across multiple groups and forums I frequent now.

Sanctuary

R3L3NTL3SS

Maxidrom2002

savi01
I'm glad this was caught before cards were shipped out to everyone.

I don't understand why some people are saying this doesn't build trust? If the cards with the potentially bad setup were shipped out and EVGA didn't say anything, then it's fair to be upset. They fixed the issue BEFORE shipping them out. Were they supposed to apologize for cards they didn't send out???

Also, I think this is catching a bit too much fire/attention. Igor did some fantastic work in finding a potential cause of issues, but I think Jay's video is pouring gasoline on this whole crazy release and he's enjoying the attention. 

Hey, plz become educated. They/Jacob only mentions FTW3 card...how it has 2MLCCs...which is same as FE version...

At the end he is like...oh, and BTW, XC3s have 1 MLCC...but thats no issue...they will run just fine...

Um...no, there will be left performance on the table...because of early on decisions.

Buy a higher SKU card if you want the best with the most OC potential. Thats literally nothing new. Lol

The 3080s have so far all been shown to have a notoriously low headroom for overclocking, regardless of the brand.  So far the normal FTW was shown to be the "best" in that regard, but the "best" by a paltry 2fps.  If this is a hardware limitation, and even EVGA is not safe from the 4+20 crash, then the FTW3 Ultra doesn't seem like it's actually going to do anything that the base FTW can't do since they will both be limited by the same boost clock.

Yes! There is not reason for OC cards because these rtx cards will boost higher than spec (1740 or whatever spec its set) automatically based on thermal performance/limitations.

EVGA for the win

rain2_usa

Maxidrom2002

rain2_usa

Maxidrom2002

rain2_usa

pbmpharmacist

Maxidrom2002
U guys knew about this issue weeks, if not months ago...and you are all coming out of the closet now...ONLY BECAUSE OF INDEPENDENT REPORTING. This is not how u build trust...

Agree

Because none of the EVGA retail cards have issues...absolutely no reason whatsoever to make a statement on an unreleased product where the issue was discovered and fixed before release.  

Thats a bold statement cotton, lets see if thats gonna work out for ya...

The "bold" statement is what an EVGA brand director stated in this thread about the retail release.  Review cards may have them, which are being swapped bout, but NO retail cards.  EVGA has made that statement. If they lied about that, there will be repercussions on their brand, which I don't think anyone in the company would risk.  

U highly doubt anyone would straight up lie
..thats too damaging.

Its marketing deception. 2 points:
1. I wish they would explain early on WHY they didn't have card availability...and tell u the truth still don't.
2. They only address FTW...while XC3 gets left out for the most part. Jay already mentioned cards having problems with just one MLCC and will most likely also need to update v-bios to downclock.

1. No...they don't.  If It's delayed, it's delayed.  They don't have any obligation to specifically state so and so reason for delaying the card.  It's also a very custom card compared to the XC3.
2. No, Jay didn't state 5+10 have problems.  You'll need to backup that reference because even Jay had to post a follow up video on Twitter that's not what he was saying about the POSCAPs and MLCC / power delivery.  And 5+10 is NVIDIA Reference design; not a deviation.  You obviously are trying to cherry pick pieces of information to back up your claim.  

Let them think what they want.  The more that believe this and decide to buy a competitors product, the easier it will be for the rest of us to our EVGA's.  ;)

Maxidrom2002

savi01
I'm glad this was caught before cards were shipped out to everyone.
 
I don't understand why some people are saying this doesn't build trust? If the cards with the potentially bad setup were shipped out and EVGA didn't say anything, then it's fair to be upset. They fixed the issue BEFORE shipping them out. Were they supposed to apologize for cards they didn't send out???
 
Also, I think this is catching a bit too much fire/attention. Igor did some fantastic work in finding a potential cause of issues, but I think Jay's video is pouring gasoline on this whole crazy release and he's enjoying the attention. 
 
 

Hey, plz become educated. They/Jacob only mentions FTW3 card...how it has 2MLCCs...which is same as FE version...

At the end he is like...oh, and BTW, XC3s have 1 MLCC...but thats no issue...they will run just fine...

Um...no, there will be left performance on the table...because of early on decisions.

Ok, "plz" educate me then. I'm 100% sure you're just going off of a youtube video and have no idea about any of the components that are actually being discussed. 

Educate me on how much performance you know for a fact you're going to lose because of this. Educate me on what the best solution is. Educate me on how to a company should handle this situation.
 
Do you even have a new card in hand????

EVGA_JacobF

There were no 6 POSCAP production EVGA GeForce RTX 3080 FTW3 boards shipped.

 
...And there were no EVGA Geforce RTX 3080 FTW3 bought from me because... bots and scarcity

Pilot 53

R3L3NTL3SS

SprayingMango
This is cute and all but ASUS is shipping cards with 6 MLCC, which seems like the best solution. Why isn’t EVGA doing the same?

Why are you all stuck on this when multiple people have said 6 MLCC are actually not ideal?

I haven't heard this yet, do you know what the reason is?  Is it less overall capacitance? 

Yes.  The MLCCs are low capacity, but good at filtering high frequency transients.  The SP-CAP have much better bulk capacitance to keep the rail up during load transition (while the VRM catches up).  Most decoupling networks involve 2 caps for the bulk+filter approach.

Imo the options are FE or Asus Tuf at MSRP, at high end EVGA FTW3 seems to be all right, Strix seems to be really really well built though

Thanks for the info Jacob!

Maxidrom2002
U guys knew about this issue weeks, if not months ago...and you are all coming out of the closet now...ONLY BECAUSE OF INDEPENDENT REPORTING. This is not how u build trust...

Uhh facts are that technically none shipped with the defect. Since none shipped it technically didn’t exist so there really isn’t a story here. They are simply being upfront with their QA process. That is all this is.