Discussion in 'Videocards - NVIDIA GeForce' started by Witcher29, Oct 11, 2022.
Might be waiting a while bud.
Off you go to the ignore list.
Again, I did not do any kind of hate speech, neither did the others.
You are basically the only one talking about hatred.
What you say doesn't make any sense in this thread.
This is starting to become a nuisance. I kindly ask you to stop please.
A little update on this entangled aio mess .
I'm not sure if the card got mad at me or something for saying its a pos but
it just went 180 degrees and proved my ass wrong. I decided to give it one last
shot and flash the bios with some random from techpowerup for this model
before RMA'ing it and low and behold. It's been what, a week maybe and it the
pump turns on every single time. Not only that but, and i don't even know how
this is actually possible but the coil whine... it reduced itself by 50%.. It's no longer
audible enough to drive me nuts.
The card does 66c average with the core doing 3000 mhz stable at 1,075v and
the memory is at 11750mhz. Card basically runs like a dream now, going to be
a blast shooting commercials on it. For anyone wondering, 2050 euro was the pricetag.
I'm from the EU so that's that.
Hold it there Sir, can you prove that? Do you have before and after recordings perhaps?
This seems very unlikely to me, that a bios update being capable to reduce coil whine is something I hear the first time.
Could you please elaborate what did you do exactly? Which bios version did you pick and did it reduce the coil whine instantly under load? Anything else you might have done in addition, perhaps hardware mods or something similar? Is the reduced coil whine temporary, or did you observe it to be permanent ever since the bios update? Also, are you sure it's not just placebo (sorry for the dumb question lol)?
I would appreciate your feedback.
I'd never waste time recording coil whine when I was set on RMA'ing the card. The coil whine was not my main issue, it was the card hitting 90c during boot due to the pump
not turning itself on. All I did was flash this bios VGA Bios Collection: MSI RTX 4090 24 GB | TechPowerUp. I don't remember what my old bios number was, all I do remember
is that at the end it had "0C", there's one at techpowerup that had "4C" at the end so i thought I wouldn't even bother flashing something similar to mine so i went with a bios
that had its numbers way different. The moment I flashed it and shut the pc down, turning it back on made the gpu's pump work straight away(it usually never works on a cold boot)
and It's been working from cold booting ever since. About the coil whine however, I'm not exactly sure what the logic was there considering the bios has literally nothing to do with it
but I'm not deaf. I shoot commercials inside of GTA V and that game's menu is capped at 188FPS, this card at that fps makes it sound like a tiny violin, or at least it used to.. now after
the bios I can't really hear it anymore. I don't game a lot but I'm pretty sure about it being pretty audible inside of GTA.
Hhmm, this is interesting. Thanks for sharing this information.
Could you perhaps do another check for me please?
Can you try undervolting your card once with the new bios version with 81% power Target and +100MHz overclock boost clock and check about coil whine in these settings?
It would also be interesting if the coil whine issue changes over time, so if you could observe it for a while and report back in let's say a week or so (with your normal clock settings that is).
When I had my Liquid X there were no issues with the pump but coil whine was so ridiculous that it was impossible to keep the card as is.
Nothing changed with these settings, did 0.950v on the core. Whining sounds the same but its not even near annoying levels.
Can't keep the card undervolted however because I loose performance due to the niche scenario where the card runs at ~1500mhz
due to inconsistent load during exporting footage from the game. However I do remember the coil whine changing as soon as the
bios was swapped. It didn't take time for it to fade away, as soon as the new bios was up its like the card was not the one I initially got.
Hi Bud, I know it's about a year to late but have you ever flashed the zotac with the 600w bios? If so how did it perform? Was it stable?
Hi, no. Not sure the zotac cooling would be the best for that or at least the fans would be maxed out and be very loud. My card can do 2950mhz @ 450w which is enough for me.
Have the memory failing on my MSI Gaming X Trio... I've never had any issues with GPUs until the EVGA 3090 and now this one.
Super annoying black flashing on top half of the screen during games that run 800MHz memory unless I force it to run at maximum performance mode, Apex occasionally runs at what looks like 20fps while afterburner reports 160~ FPS (get a solid 238 normally).
That's unfortunate bud, I've only ever had one card die on me and that was also a VRam issue and black squares on screen on my 280gtx, I think it was an EVGA and they replaced it with a 285gtx in about two weeks. Hope it works out for you.
Very odd. Do you have ECC enabled, perhaps? If so, try to turn it off.
Have you been overclocking the memory or running it at stock all the time?
I was running at about 11000MHz ECC off before, never seemed to go very high compared to lots of other results tbh back when I got it at launch, might have been doomed to fail. Was sort of hoping someone would go "oh that happened with this batch of blah blah" so I could squeeze out an advance RMA, but I guess I was just unlucky lmao.
If I DDU my drivers they repeatedly fail to install over and over until I get lucky too... It's only a matter of time but I sort of need this to be working and can't afford downtime...
Rather than start a new thread about this, I thought I'd pop in here and ask about the TDR bug plaguing some (all?) 4090s?
Ever since early May (this may have surfaced due to a Windows 10 update) I've been plagued by intermittent TDRs on the display. Either a game will crash, or the screen will go black.
It happens now and then, funnily enough, 95% in my partner's account, rarely in mine (same PC).
I've learned this is a historic bug, and turning on power management to high performance has massively helped with it, but had another one tonight, the first in a month.
The question here is how many others experiencing this issue with this card? Mine's an MSI Gaming Trio, and the system is otherwise completely stable. Almost every crash re: this is Kernel 141 in reliability history.
I have now turned off GPU scheduling to hopefully totally eliminate it, but boy is this frustrating.
Curious thing is this happened with my old 3090 as well, on a completely different system. What is Nvidia playing at these days?
Upon reading that "(all?)" speculation on your end, I'd like to remind you that when reading
up bug reports on the web, you always have to keep in mind that you're reading like 1 to 5%
of all owners of the specific product. You don't see people going around forums saying
"yaay my 4090 is a beast, bugfree mf" if everything is fine and dandy. Those people just never
post anything on forums because.. if there's nothing to complain about then its pretty much
human nature to just enjoy your product.
And.. i did read up that TDR thing you mentioned, for all these years ive never seen it on my end
and id hardly call it a plague. One thing I did learn is motherboards sometimes don't play nice
with specific gpu models. You've changed twice and say the issue persists. Maybe.. slap your gpu
in a system a friend of yours has and can support. See if you can replicate it there.
Nobody makes a post on Reddit and says "my rig is stable, that is all"
Some guy on overclockers re-hacked nvflash to support 4 series. You 4 series owners rejoice.
Omg me lmao. See my post above yours. I get this sometimes too LOL.
People suggested it was caused by Cablemod cables - are you using any? I got mine replaced and it's all good now (apart from previously mentioned issues...)
Just thought I give my impressions on my 4090 I bought recently. The performance is simply amazing, literally, 4K 120fps in all but 1 game I have played so far (I have tested many old and new games). The game in question was the last of us, the last level just didn't play smooth for some reason. It's probably the best card I have used when comparing performance over previous card upgrades.
Regarding the power connector, my 4090 strix had a manufacture date of March, the power connector felt chunky and made a satisfying click when I plugged it in. I do not know if they revised the power cable since the melting issues but there is no wiggle, it's a really firm connection. I currently use 3 out of 4 pci e to power it up on a Seasonic 1000w titanium psu. I have had 0 crashes or black screens since I have had it, used it in long sessions (8+ hours).