Rtx 3080 lower quality capacitor Issue

SniperSmithy · 1 Oct 2020 at 17:04

chrcoluk said:
His excuses for not disclosing would have some leeway if he only had 1 GPU, but he had multiple, if they were all crashing, ok could be driver, could be his pc, but only one was crashing which meant in my opinion he should have disclosed. Although it was a driver fix, I consider it a manufacturing issue that happened to be fixable by a driver rather than a driver bug.

Hardware unboxed said in this very video that he had crashes on two different cards...

SeeNoWeevil · 1 Oct 2020 at 17:56

That J2C follow-up video was an absolute joke. I'm probably in the minority here but while I have a lot of respect for the likes of Igor. I feel looking at images of PCBs and making speculation about the causes of the stability issues was a bit silly and what kicked off all this nonsense.

SGz_Eliminated · 1 Oct 2020 at 18:45

To quote my comment on Hardware Unboxed's new video

"You completely glossed over the details of why the capacitor thing blew up (instead just blaming Jay), fact that partners, especially EVGA, made announcements that SP-CAPs weren't good enough / failed testing. And many partners had changed the configuration of capacitors very close to release. A few brands sent pre-release cards to reviewers which they are now replacing and many of the images found online of the capacitor configurations differed on the actual cards (because they were changed). Its hardly surprising people were blaming the capacitors when partners were outright saying they were making/showing they were replacing or changing last minute the SP-CAPs to MLCCs. Of course people are going to think its the capacitors.

Although there was plenty of evidence showing various configurations were affected, the Zotac trinity (with stock spec SP-CAPs) were one of the worst cards affected by this issue. Although I agree there wasn't enough level-headed thought and people were too quick to grab their pitch forks (especially those cancelling their order) there was enough factual information present that suggested capacitors were an issue."

As for Jays update video, he reiterates in the new one how its only part of the system and was just theory and speculation, although he does it a bit too passive aggressively. Clearly thinks his viewers just misunderstood and not so much he did a **** job of making it clear it wasn't fact even if that was his intention originally. I don't deny Jay added fuel to the fire but he wasn't the only one, partners didn't help with the whole changing configurations business.

Shiari · 1 Oct 2020 at 19:49

SGz_Eliminated said:
To quote my comment on Hardware Unboxed's new video

"You completely glossed over the details of why the capacitor thing blew up (instead just blaming Jay), fact that partners, especially EVGA, made announcements that SP-CAPs weren't good enough / failed testing. And many partners had changed the configuration of capacitors very close to release. A few brands sent pre-release cards to reviewers which they are now replacing and many of the images found online of the capacitor configurations differed on the actual cards (because they were changed). Its hardly surprising people were blaming the capacitors when partners were outright saying they were making/showing they were replacing or changing last minute the SP-CAPs to MLCCs. Of course people are going to think its the capacitors.

Although there was plenty of evidence showing various configurations were affected, the Zotac trinity (with stock spec SP-CAPs) were one of the worst cards affected by this issue. Although I agree there wasn't enough level-headed thought and people were too quick to grab their pitch forks (especially those cancelling their order) there was enough factual information present that suggested capacitors were an issue."

As for Jays update video, he reiterates in the new one how its only part of the system and was just theory and speculation, although he does it a bit too passive aggressively. Clearly thinks his viewers just misunderstood and not so much he did a **** job of making it clear it wasn't fact even if that was his intention originally. I don't deny Jay added fuel to the fire but he wasn't the only one, partners didn't help with the whole changing configurations business.

I believe you've captured the essence of it yeah. The incomplete and misunderstood information and kneejerk reactions going around haven't helped anyone really.

Rroff · 1 Oct 2020 at 19:57

There was a lot of people with an agenda who ran with it as ammunition, with no interest in the facts, to fling at nVidia never mind what Jay did/said.

SniperSmithy · 1 Oct 2020 at 20:34

Rroff said:
There was a lot of people with an agenda who ran with it as ammunition, with no interest in the facts, to fling at nVidia never mind what Jay did/said.

I'm sure that there must have been at least a hint of AMDs marketing department involved in this some of this sensationalist hype. If there wasn't then frankly they weren't doing a very good job.

SnR3 · 1 Oct 2020 at 21:02

Locky said:
About 270 on average @1440p but thats relative to my graphics settings in game..

Yeah I totally get it. Can't play competitive and have max settings it don't go in pair. That's really cool was hoping for 240+ so I can take full advantage of my g7

K1LaW8tt5 · 1 Oct 2020 at 21:34

of course everyone blamed the caps, Everyone knows Nvidia doesn't have driver issues.

Locky · 1 Oct 2020 at 23:48

SnR3 said:
Yeah I totally get it. Can't play competitive and have max settings it don't go in pair. That's really cool was hoping for 240+ so I can take full advantage of my g7

I was getting about 150 on my 1080ti. This card monsters that.

mmj_uk · 2 Oct 2020 at 00:03

weldon855 said:
of course everyone blamed the caps, Everyone knows Nvidia doesn't have driver issues.

EVGA made a statement saying they had discovered the issue themselves and delayed their FTW card as a result? if the boost clocks are being lowered as a fix then it points to either NVidia being too aggressive with boost clocks to begin with or due to confusion over caps some cards were unable to meet specification and so the specification has been lowered. NVidia cards boost several hundred megahertz above their advertised speeds anyway so it's not like they can be done for false advertising.

Rroff · 2 Oct 2020 at 00:05

EVGA's statement I think was just PR people trying to leverage the situation - there is no way they ran the full gamut past engineering to get a full take on the situation or they'd have had a very different statement.

opethdisciple · 2 Oct 2020 at 00:51

benjii said:
New features, performance improvements for new game releases and miscellaneous bug fixes.

A game performance driver can turn a game from a stuttering mess to playable. There's also a bunch of new features that have been added over the past year, like adaptive power, fast sync, sharpening etc. Plus I've had a number of long persisting bugs resolved over the last few months.

Don't forget you have security updates as well these days also.

SniperSmithy · 2 Oct 2020 at 01:37

mmj_uk said:
EVGA made a statement saying they had discovered the issue themselves and delayed their FTW card as a result? if the boost clocks are being lowered as a fix then it points to either NVidia being too aggressive with boost clocks to begin with or due to confusion over caps some cards were unable to meet specification and so the specification has been lowered. NVidia cards boost several hundred megahertz above their advertised speeds anyway so it's not like they can be done for false advertising.

EVGA saw a PR spin opportunity and took it. You think every corporate statement is fact despite all the evidence to the contrary? It's almost as if they want people to switch brands or something...

Rroff · 2 Oct 2020 at 03:57

At least he is honest

https://youtu.be/y_v7p3xEPDM?t=477 had to laugh when it was in my YT recommendation today.

hyperseven · 2 Oct 2020 at 04:05

Rroff said:
At least he is honest https://youtu.be/y_v7p3xEPDM?t=477 had to laugh when it was in my YT recommendation today.

Was he taking the **** in the vid or is that how he thinks you solder :confused:

Devious · 2 Oct 2020 at 08:16

Ive not been experiencing crashing unless I try to overlclock but have installed the 456.55 driver out of curiosity.

My default boost curve was quite high before but it has gone up another 15mhz. Comparing GPU-Z after a timespy run its drawing more board power and has drawn has a higher voltage.

So, not had my boosts knocked back by like others are reporting.

456.38

456.55

Locky · 2 Oct 2020 at 09:19

manged to push this card a bit more https://www.3dmark.com/3dm/51165910 peaked at 2205mhz

Devious · 2 Oct 2020 at 10:53

Locky said:
manged to push this card a bit more https://www.3dmark.com/3dm/51165910 peaked at 2205mhz

Nice, has the new driver improved overclocking at all?

Chris Ecob · 2 Oct 2020 at 11:43

Locky said:
manged to push this card a bit more https://www.3dmark.com/3dm/51165910 peaked at 2205mhz

Nice results. What card is it? Is it on water?

Locky · 2 Oct 2020 at 19:16

Chris Ecob said:
Nice results. What card is it? Is it on water?

Its a gigabyte eagle OC just on air.

Devious said:
Nice, has the new driver improved overclocking at all?

Nah its the same to be honest.

I did another run and it maintained an average clock of 2025mhz