Old RTX 2080 Ti FE Meets Replacement RTX 2080 Ti FE

DigitalGriffin said:
Lol. That's the link I posted a couple back
Click to expand...

Along the lines of this theory, I'd add in there the possibility of the Memory Controller being added into the mix (interconnects, silicon, including memory controller) - This could possibly explain high memory temps, yet the memory itself functioning fine at the higher temps (high temps result of problem with interconnects, silicon, and/or memory controller, but not the problem itself).
 
homernoy said:
I have a Gigabyte Windforce 2080ti. Unfortunately it has Micron v-ram. Hopefully my card doesn't take a shit. So far, no problems to report after extensive 4k gameplay, and benchmarking.
Click to expand...
Tell giga you want to rma for one with samsung.

phawkins633 said:
I have a Zotac 2080ti Amp that also has Micron Ram. As these problems are primarily on the FE cards, especially the "Space Invaders" effect, I to am crossing my fingers/toes/eyes.....but so far it's a shit-ripper....and playing BF5 at 3440 x 1440 ultra works as smooth as butter on my Predator G-Sync monitor....but I'm still following this pretty closely.....yikes!!
Click to expand...

Uh huh better head to your local usps store and rma that sucker for one with sammy.

Galvin said:
Whats the best way to test the memory? Got a 2080ti XC from evga. And it has micron ram. Played world of warcraft yesterday no issues.
Click to expand...
You better wise up and contact evga and demand an rma for samsung memory.

Man ya'll crazy.:ROFLMAO: if you have a 2080ti with micron you better suck it up, get over yourself and get that rma asap! $1200 dollars?:eek: shoot i'm rma'ing now. rat now.:D don't know what yall are waiting on. i'm guessing y'all just want to play space invaders 2018?:confused: i dunno let me get out of here.
 
Good to know they found a fix, I have the Ti ultra xc from evga, running at the crazy power target and the overclock the scan of precision X1 gave me. But it's the colder months now, we'll see how this thing handles in the summer.
 
I think a more interesting use of that FLIR camera would have been to video Kyle last week when FedEx drove off with his replacement card.

I'm betting he was hotter than any of that Vram.
 
After the verbal jerking off Jensen done about this cooler being "designed for overclocking" i have to wonder how much actual testing it got. Seems like they forgot about cooling other components and just focused on the gpu.
 
Trepidati0n said:
You said the voltage affected the resistance...and then you give me a link to how temperature affects resistance. *boggle* Again...now with signal reflection comment on top of it....seriously...just stop.
Click to expand...

What this actually proves is that, with the internet, you no longer need to spend four years earning a BSEE. The only value in doing so is that you'll know what you're talking about - but you still won't understand the internet guy, nor he you. So I'd give it up. Though you could print this thread and send it to your alma mater... they might give you a refund out of sympathy.
 
Last edited:
Formula.350 said:
Beings things what they are, computers and all I took it as being Hex, considering the 0E on the end. Though, I guess one could presume it's a "Beta Engineering" version heh

Now full disclosure, I am out of my depth when it comes to coding, hex, binary, etc... But my interpretation is that, for the Micron, 0 is "00", B is "12" and we can ignore the last pair since it's irrelevant at this point (due to versioning sequence already having been interrupted with the difference in the third pair). Where as on the Samsung, 1 is "01" and... well 7 is "07". So the Micron would indeed be older at "0012", compared to the Samsung's "0107"; however, as far as indicating it's a beta version I don't think that's the case since as far as I know, that would fall at the very end of the pairing and more importantly, be lower case.

It's that which is how I've seen things over the years, as it keeps from getting things confused in just this sort of instance. If it's inferring Hex then capital letters are used, and if they are lower case then it's a "string" (or whatever you want to call it, but not 'numbering' related).

The more interesting or perhaps telling info I'd suspect would lie in the Advanced tab of GPUz, in the BIOS adnd ASIC Quality drop-downs. (Though I believe it's been said that "ASIC Quality" is mainly just an AMD thing, so that may not yield anything meaningful, that's assuming it's even there to select for nVidia cards.)
Click to expand...
Very well could be, honestly, I do not pay much attention to GPU EFI.
 
Wonder if the Samsung ram are better overclockers?
On another note.. Got my second EVGA 2080ti, but it's a 2080ti disguised as a 2080.. They shipped me a 2080ti with a 2080 cooler and backplate :cautious:
I didn't even notice till weeks after.
The support guy said he never heard of that happening before, but both serial numbers match up on 2080ti box and the card, so what's going on at the EVGA factory?
The card is cool and much quieter than my first card, and I haven't had any problems with the card at all, but I think the resell value would be worse down the line.. Time for trying out my third card with Samsung memory (I hope)
 
To be honest, being a gamer from the past 20 years, it as always been told that squares on the screen equal faulty gpu and lines on screen mean faulty ram.
http://www.playtool.com/pages/artifacts/artifacts.html

That said i would be thankful if one add the sources from the industry pointing out that there is indded a Micron memory IC issue, because until now i still can't leverage who said what.

Then i have no issue considering the ram IC itself, but again from my past experience, ram do not bork straight a gpu, often it take some time to completely die.
So i would still point toward the gpu die, power management issues even if the industry say the opposite, could be the memory controller or bus issues.
Still these issues are regarding the gpu die core and not the ram chip itself in my opinion.

I'm still waiting some of these high profiled engineers release some X rays of the card or some decent scope analysis, as it has been already prosed here.
Also nobody tried to bake the card, often this give some indications about the issue, if the internal die solder bumps are themselves borked or if it is a pcb or solder balls issues.
Also note that one can clearly say that the pcb and card manufacturing cost were keept very low, the solder mask and some sketchy caps saw on the RTX series speak by themselves.
So again i'm not sure this is a ram chip issue only, Nvidia have not a long lasting tradition of good and solid reference pcb manufacturing.

A decent overview from peeps in the industry of the issue is greatly welcomed!
 
Looks like Kyle needs to contact an aquacomputer rep for one of them active backplates
 
Creig said:
Nvidia sure like to run the memory toasty warm on their cards.

View attachment 121580
Click to expand...
That is the Maxwell Titan X, which had 4Gb memory chips. Meaning, they had to double up to 24 (12 on each side) to get to 12GB capacity, creating a heat soak condition. For the record, though, I ran a custom 425W BIOS on my Maxwell Titan X for a little more than 2 years without having any issues. Core was overclocked to 1.4 GHz and memory was overclocked to 8200 MT/s. And this was still on air, using an EVGA ACX 2.0 addon cooler.
 
Well tested the 2080ti with micron ram. Did heaven for 9 hours, GPU never went over 71c. Guess I should run the evga artifact tester for 9 hours too. Since heaven benchmark isn't going to test all 11gigs of ram.
When I go to sell this gpu in a few years. People wont want to buy cards with micron ram in them
 
Welp, my "new" Samsung memory 2080 Ti just gave up the ghost. Will not longer even give me a DP signal.
 
Crystoff said:
What this actually proves is that, with the internet, you no longer need to spend four years earning a BSEE. The only value in doing so is that you'll know what you're talking about - but you still won't understand the internet guy, nor he you. So I'd give it up. Though you could print this thread and send it to your alma mater... they might give you a refund out of sympathy.
Click to expand...

Out of respect,

I'm not going to get into it. He's an engineer. But I don't know what his background is. But I'm an engineer that worked for NASA on space hardware. I also worked as an embedded control sensor engineer for someone who received the Nobel Prize in chemistry.

I imagine something is getting lost in translation over the internet. My specialty is as an astrospace engineering & computer science with experience and emphasis in advanced engineering specialty applications, IOT, AI, with some electrical engineering.
 
DigitalGriffin said:
........ But I'm an engineer that worked for NASA on space hardware. I also worked as an embedded control sensor engineer for someone who received the Nobel Prize in chemistry.

I imagine something is getting lost in translation over the internet. My specialty is as an astrospace engineering & computer science with experience and emphasis in advanced engineering specialty applications, IOT, AI, with some electrical engineering.
Click to expand...

OK, You have blown your cover sir.
Expect questions from several of us, when we need NASA help on things !!!

:ROFLMAO:
 
Nope !
I think we all need to sign up as patrons.I'm going to in 5 mins !

Kyle and [H]ardOCP had to pay that 4000$ because they did not sign the NDA.
I know I'm not buying 2080's right now, because of what I have read here, and that's saving me hassle, and probably money.

The caveat : we all voted for [H] to not sign that agreement.
The price : 'has gone waaaay up'.....and I think we should all pony up for that unbiased no BS review we all like.
 
Kyle,this thread needs updated title:

Old RTX 2080 Ti FE Meets Replacement RTX 2080 Ti FE - In the Afterlife!
 
nVidia - putting the 'halo' in 'halo product'!

I'm holding out for one that can reincarnate into a Quadro!;)
 
You must log in or register to reply here.
Back
Top