Rtx 3080 lower quality capacitor Issue

Zeed · 28 Sep 2020 at 17:05

TrM said:
next big thing on the internet omg nvidia nerfed the clock speed on my card to make up for bad caps and bad board design how dare you nvidia! Change.org petition income for sure.

Can i call my AIB's MANAGER is next stage

Zeed · 28 Sep 2020 at 17:06

Gamer_3983 · 28 Sep 2020 at 17:11

Well Amd says that there gpu will have 50% more performance with less power and might beat ampere and probably use less power compared to ampere. Hope this is true. If so, Nvidia will fail badly and will get a big slap on the face by Amd.

Zeed · 28 Sep 2020 at 17:15

Gamer_3983 said:
Well Amd says that there gpu will have 50% more performance with less power and might beat ampere and probably use less power compared to ampere. Hope this is true. If so, Nvidia will fail badly and will get a big slap on the face by Amd.

AMD says allot then REALITY kicks in

p1on · 28 Sep 2020 at 17:16

Zeed said:
AMD says allot then REALITY kicks in

people who want your money can say pretty much anything

ALXAndy · 28 Sep 2020 at 17:24

Gamer_3983 said:
Well Amd says that there gpu will have 50% more performance with less power and might beat ampere and probably use less power compared to ampere. Hope this is true. If so, Nvidia will fail badly and will get a big slap on the face by Amd.

You have misconstrued that.

AMD claim 50% performance per watt over existing Navi. IE RDNA. RDNA2 is supposed to have 50% gain in performance per watt. At no point does that say they are going to be 50% faster than Nvidia per watt or anything else.

Zeed · 28 Sep 2020 at 17:30

p1on said:
people who want your money can say pretty much anything

Nv gave AMD free "card" to get some people cause cant buy em 3xxx cards anyway

SeeNoWeevil · 28 Sep 2020 at 17:31

https://www.youtube.com/watch?v=2QVHX1mPh1o

Zotac Trinity running comfortably at 2100Mhz after the new driver

FlyingFish · 28 Sep 2020 at 17:39

mat690 said:
Based on pictures from ASUS's website. The regular TUF card has the problematic capacitors and the OC version doesn't.

I've got the TUF non OC, and it's got a complete set of MLCC caps.

Gamer_3983 · 28 Sep 2020 at 17:40

Why cant they just raise the power limit. Increasing the power limit will reduce the fluctuation by a lot since thats the problem.

ALXAndy · 28 Sep 2020 at 17:42

Gamer_3983 said:
Why cant they just raise the power limit. Increasing the power limit will reduce the fluctuation by a lot since thats the problem.

They have and it has raised the TGP by another 10w. Whether you care about that or not is up to you.

Gamer_3983 · 28 Sep 2020 at 17:42

The caps isnt the culprit I guess

Gamer_3983 · 28 Sep 2020 at 17:43

ALXAndy said:
They have and it has raised the TGP by another 10w. Whether you care about that or not is up to you.

I never knew they did

SeeNoWeevil · 28 Sep 2020 at 17:49

EVGA are going to look like clowns after trying to dunk on everyone else thinking they 'solved' the issue with a different cap configuration.

DFroN · 28 Sep 2020 at 17:55

Very early to say but this driver does indeed seem to have altered the cards behaviour. I had a crashing card and I was testing undervolting earlier today in COD. In the same place on the same map the new driver is glued to 2025Mhz and 1.068v once up to temperature whereas earlier the boost and voltage was fluctuating massively (over 200Mhz).

I’ve had the game running at 2025Mhz for over half an hour now and before the update I had to limit boost to 1995Mhz max, which meant it would settle around 1965Mhz.

Needs others to validate but I’m pretty confident something has been changed/fixed.

SniperSmithy · 28 Sep 2020 at 17:57

SeeNoWeevil said:
EVGA are going to look like clowns after trying to dunk on everyone else thinking they 'solved' the issue with a different cap configuration.

Yep, it was a weak attempt to spin their lack of stock into something good I think.

p1on · 28 Sep 2020 at 18:10

so nothing to see here?

KompuKare · 28 Sep 2020 at 18:12

Curlyriff said:
If you are getting an OC over the boost of 1710Mhz that what the box says then surely your return varies as per all over launhces were some cards are better than others?

Well except for one small detail:
The 3080 FE cards boosted to close to 2000MHz. Not the 'spec' 1710MHz.
TPU had them around 1900MHz mostly:

https://www.techpowerup.com/review/nvidia-geforce-rtx-3080-founders-edition/32.html
CB also had them at up to 1965MHz:
https://www.computerbase.de/2020-09/geforce-rtx-3080-test/2/#abschnitt_die_taktraten_unter_last
If a driver would have 'fixed' this by limiting boost to 1710MHz, then all the reviews would have become invalid.
And then a 3080 limited to 1710MHz versus an overclocked 2080Ti... well, that wouldn't look good at all.

Zeed · 28 Sep 2020 at 18:28

KompuKare said:
Well except for one small detail:
The 3080 FE cards boosted to close to 2000MHz. Not the 'spec' 1710MHz.
TPU had them around 1900MHz mostly:

https://www.techpowerup.com/review/nvidia-geforce-rtx-3080-founders-edition/32.html
CB also had them at up to 1965MHz:
https://www.computerbase.de/2020-09/geforce-rtx-3080-test/2/#abschnitt_die_taktraten_unter_last
If a driver would have 'fixed' this by limiting boost to 1710MHz, then all the reviews would have become invalid.
And then a 3080 limited to 1710MHz versus an overclocked 2080Ti... well, that wouldn't look good at all.

Small DETAIL was said that NV is keeping BEST SILICONE for FE cards...

DFroN · 28 Sep 2020 at 18:34

Yep the new driver has fixed it, at least for me. Before the card couldn’t go over 2GHz, now it never drops below 2GHz. The difference seems to be that the clock and voltage are much more stable whereas before it would jump about the boost table aggressively. Either that or Nvidia downloaded new capacitors into the drivers.

Edit: The wording above makes it sound like the driver is boosting the card higher, that’s not the case - I meant that the card was only stable under 2GHz with the old driver, now I can overclock it so that it never drops below 2GHz and it remains stable.