Testing New World with my RTX 3090 FTW3... Surprising results!
Vložit
- čas přidán 21. 07. 2021
- Get a free pair of Wireless Bluetooth Headphones at Micro Center: micro.center/7c5fd6
Check out the Micro Center Custom PC Builder: micro.center/c144d8
Get your JayzTwoCents Merch Here! - www.jayztwocents.com
Worldwide PC Parts Giveaway including RTX 3090!! - • Ultimate Gaming PC Upg...
○○○○○○ Items featured in this video available at Amazon ○○○○○○
► Amazon US - bit.ly/1meybOF
► Amazon UK - amzn.to/Zx813L
► Amazon Canada - amzn.to/1tl6vc6
••• Follow me on your favorite Social Media! •••
Facebook: / jayztwocents
Twitter: / jayztwocents
Instagram: / jayztwocents
SUBSCRIBE! bit.ly/sub2JayzTwoCents - Věda a technologie
As an electronics engineer, these weird noises from the EVGA card really sound like inductors saturating under overload. This phenomenon creates massive (and very damaging) current spikes in the power stages. When an inductor reaches its maximum current, it suddenly stops being an inductor and allows the current to run away and increase almost without bounds within nanoseconds. Something about the load that New World puts on the VRM might make the phase currents get out of balance (for example, huge swings in power consumption that the VRM isn't fast enough to follow), which then causes inductor saturation and card death.
I'd love to see oscilloscope shots of the inductor current and power stage PWM signals, but I guess Jay isn't the right guy for that. Buildzoid...?
I totally agree Jonathan! But the diagnostic analysis from EVGA's RMA's would be nice! But one thing 'missed' by JayZ is the fact that as 'gamers play' they'd be pushing those limits for hours on end and wouldn't be watching the power draw; repeatedly stressing any components. Another 'tell all' would be if someone - some how was able to analyze the patch from 'Amazon' to see what they did. So says Gray Kitty from Gamers Nexus.
Electroboom is your guy then lol
Sounds like a lot of failures in QA testing. The firmware and drivers should have prevented it from ever getting into that state. Yes, I am an embedded SE who writes drivers. An application should not be able to break hardware, that privilege should be limited to firmware and maybe drivers.
Hey.Wanted to ask you something that is not really related to this 3090,only because u seem to know what you are talking about.
How can a gpus lifespan be extended or what are some thing to not do with your gpu to last longer?In this current gpu situation a lot of us are hoping our gpus dont die,so i was wondering if you have any advice on that.
Buildzoid did do a video on it. The first pop is OCP kicking in that's why users can restart the first time. The second time the power phases are already hot and get out of hand causing it to heat upway too much(in some cases melting the copper under the phases). But in reality a card should not be able to push itself like this. No user or program stock should just "fry itself ".
1st test: thermometer
2nd test: Nick says it's hot lol
Jayz gets all his information from word of mouth. "I got some information from some people who wrote me" means "I read in the comments" Really trustworthy sources LEL
@@tron-8140 Because most other "Tech guys" are clairvoyant or lucky enough to experience all issues first hand! Or have a lab full of people testing things so any issue they can write a science paper on it!
Every issue on the internet gets known by users commenting on it, a hacker who found something or a leaker who will remain anonymous either way so... What are "trustworthy sources"? A guy like him writing a blog on it with the same info? Is it somehow more trustworthy if it is written?
@@tron-8140 I mean isn't that why he's testing what the people said? So the hypotheses can get proof if any?
@@gerardopadilla2666 Moore's Law is Dead gets info from his mom and dog.
It's not hot until Nick says it's hot.
12:20 - "Absolutely nothing out of the ordinary right now" *power % jumps from 65% to 128% as soon as he turns his back*
yeah, not sure why he didn't point it out especially after watching it back in editing. plus the whole already shipping replacements response i would not believe unless given proof by those customers, not hard to say "yeah we're replacing them" meanwhile everyone is just stuck with a broken gpu & zero response from customer support.
@@agentmay0 yup.. kinda cheap and sketchy. he spoke with EVGA during launch and they "apparently" had free stock of 3090s lying around and already started sending replacement cards. lmao such bullshit😂😂
120% mate... not 128%
@@StatsGam4er My mistake. At the resolution I was watching the video, the 0 with a line through it looked like an 8.
@@HaniiPuppy I thought so too, until I watched in full screen mode. Either way, good spot and shame Jaystwocents missed that one
Oh, that grinding noise.. Fun to hear the VRM's crunch themselves to death.
mine make a nearly same sound since i had my card new. so if you buy newer highend cards, there is a big chance your will get a card with this vrm/coil wine
I feel like there’s one factor you didn’t properly rule out. Heat, most people who experienced the problem probably didn’t have open air test bench. How much hotter would card have gotten had it been in a pc case.
He also didn't have it in a small form factor case attached to a riser cable.
Interesting...
Especially for those in hotter climates with massive ambient temps. Not to mention all the historic heatwaves occurring in even the more temperate regions...
but heat shouldn't kill a card. Or: if the card is designed so bad, that it won't shut down if it gets to hot, I will break in benchmarking and not only in one game
After watching for 5 minutes I thought the same thing. No one is having issues playing the game on an open test bench. He didn’t properly duplicate the gamers closed case setup so of course there isn’t going to be the same results. Plus were the gamers who had problems using water cooling or air cooling rigs?
Sparkly VRM's, that takes me back to a decade ago when GPUs sung their high pitched anthems and groaned running Crysis.
Ahh I remember.....
Me: *loads up crysis*
My 8800gtx: let me sing for you, the song of my people
Even the gtx285 sung me a song
@@JH-es8qr I'd hate to say it, but that made me laugh.
New World engine comes from Crysis engine
I’m pretty sure you could win Eurovision just by enabling RTX in Cyberpunk 2077🤪
Great to hear the bricked cards are being replaced but looks like the behaviour might not yet be resolved. Given the difficult GPU market I can only imagine the tears my eyes would shed if my 3090 died.
Is it still alive? Your 3090?
You always do great work with your videos. Love watching you for tips and for the content you put out.
Your card actually hit 125% power at a point, 18% over the set limit, thats a lot of uncontrolled powerdraw
it hits 131 closer to the end, too
@Everyday Good Vibes max was 107%
I'm really interested to see what evga has to say once they get the rma cards back
Yeah, I hope they do an Honest Press Release saying what actually 'Popped' or behavior killed the cards
They addressed it first thing right off the bat during their Thursday lifestream. I agree with Jay. Great company and they are standing behind their product. I’m sure it can be found or linked from twitch.
No doubt the cards will need to go through the production line testing to establish defect location.
Given the ranges of cards affected it could be something deep in the hardware / software that needs to be fixed
Seeing what has failed will result in the identification of a number of possible causes.
It could be as 'simple' as component tolerance creating an OCP bias, Temperature drifts affecting same, software not polling for hardware errors quickly enough.... the list goes on and on.
I too would love to see the results, though not holding my breath.
@@shaneeslick Yes, exactly. If you're reading this Evga, learn from NZXT's colossal FUCKUP: If there's a problem, fix it right and don't bullshit people with the kind of condescending lawyerspeak garbage Amazon just put out. If you done fucked up, own it and take care of your people.
I feel like every manufacturer makes their cards to break. They know the limits of the capacitors and everything on the gpu and they know with the amount of heat and electricity the gpu can take and when it will eventually break after the warranty date. It so happens new world is so intensive it’s breaking gpus. Gpus are made to break but they didn’t expect a game like new world to break them so fast.
Oh god, those end/out-takes. Please keep doing those!
Nick: "Stop touching it!" That was hilarious.
That Sidious impression was really funny, and I love the editing at the end haha
Just what I've been waiting for. This and whenever Steve gets a hold on it.
Buildzoid is the guy I want looking at it. No offense to steve but this sort of thing is way more in his wheelhouse.
@@mgkleym he already did a preliminary video. havent watched it but its called evga 3090 ramblings or something like that
Which Steve? ;)
Yeah, quite a few tech tubers will just keep speculating while Jay, GN Steve, HU Steve & Tim will actually do some tests on it.
@@Kaoristillfox Yeah it's all theory.
"Stop touching it!"'
How many times has Jay heard that in his life?
Apparently, not enough times. lol
"Ouch this electrocutes me." - "Stop touching it?!" - "See, if I roll over the carpet with the gaming chair like this and then to...OUCH!" - "..."
Also these guys bought a Fluke thermometer ($400 on amazon now) but not a single thermal camera.
For fox sake you don't need the $20K Linus is swinging around, not even a $1k FLIR. Just get an HT-19 (for handheld) or an HT-301 (phone mount) like I did.
Always curious what content your gunna make that i cant help but to watch. Keep it the good work jay 👏
"and the power draw is really low"
it's using more power on the title screen than the max TDP of my card
Its amazing that refrigerators and AC nowadays draw less power and becoming more efficient, while Nvidia getting more power hungry over time. Is that technology advancement?
@@fynkozari9271 they dont play games either they just cool....so no sound about right.
@@fynkozari9271 Yes. That's one of the phases of GPU development. TYPICALLY you'll see a big jump in performance (or, in this case, feature set, i.e. raytracing) accompanied by a l;arge jump in power draw, heat, size, etc.. Then the next generation or two will have a small performance increase but with much more efficient power usage, heat, etc.
So, it's safer to have a weaker card. Less power use - less risk to burn. 3060 - for the win.
Never stop doing those ending out-takes/edits/funny stuff. They're great.
i came here looking for this
Two thoughts: 1. Trying to ascertain tone from text is notoriously difficult. I read the same thing that Jay did and didn't get the same impression that he did. 2. Love EVGA. Had a 950GTX die on me while playing WoW and since it was under warranty they took it via RMA and sent me back an upgrade, a 1060GTX.
I had a GTX760SC. One fan failed. Got a GTX950FTW. Deffo a win.
gtx comes before numer
you shouldnt even need to RMA at all doesnt it make you a little sketched at all that they sent you an entirely diffferent and BETTER card? They have been doing cheapo shit in power delivery for years guys dont be a sucker go MSI ASUS or any other brand. Ive ONLY ever had to RMA EVGA Cards and its stupid because its entirely on their slapstick engineering
@@MasterZixxLPs no wonder they named cards FTW. EVGA definitely goes FTW.
Note to self: Don't play new games on expensive PC , wait a few weeks. What a time to be alive
Some years ago I was like "Don't play new games on old pc, it will explode my old boi" now "Don't play new games on expensive pc, it will explode my young boi"
Exactly...and definitely not a game from freaking Amazon . Wait at least a week ..
love love love micro center! and you Jay keep up the good work!
Any true engineer at EVGA would love to get failed cards back to learn what happened and how.
Im sure they have quite enough those, since a lot of people has RMA'd their card allready. Hopefully Evga and other companies check the problem with this, and fix it. I still blame 100% Evga and other companies. There are people who design protections so card's won't blow up (literally read Evga forums), and they have failed. Blaming Amazon Games is just silly.
@@yesnoyeswait4306 I think it's more NVidia than EVGA. Ampere has always been a dumpster fire when it comes to power draw and heat generation. Seriously, it's a bad design(and it's not the first time NVidia has screwed this up)
@@Grimmwoldds evga has had a long running issue with the 3090 ftw and it drawing extra power out of the pcie slots and going out of spec of the 75watt limit. they have slowly been replacing peoples cards as the fail but its kinda lack luster that they didnt just recall all the affected 3090s when they discovered it.
I absolutely love EVGA! I was able to get a 3080 FTW3 Ultra in January because of their waitlist, which is a beautifully customer centric program. And after having to settle for an Aorus Master 3080 ti from the Newegg shuffle last month, I really wish I had been chosen for the FTW3 version that day. At least my son finally got my FTW3 he's been waiting for since October 6th last year.
@@clydemasterson1129 ur either a bot or a paid shill 😂
The power caps, there’s something I haven’t heard in a bit 😆 couldn’t escape the topic at launch
Yeah, and the outcome of nvidia's bad drivers trying to squeeze every bit of performance out of these cards for day one reviews being the real culprit.
Jay: "ala Mythbusters style..."
Me: "When in doubt, C4?"
_"Quack damn you"_
@@Outland9000 Jaime wants big boom.
"This behaved exactly as expected" - said about eight times.
What should we title the video??
"Surprising results"
"Surprise - we didn't have jaw-droppingly-shocking results" :D
This video should be called "I digress and nothing happens" 😂
To be fair, I found the fact that EVGA had 3090's in stock and was cross shipping for RMA surprising ;)
Yes, it was perfect to smoke a joint while watching lol
Came to see a 3090 fry, liked for the ending edit.
that shit was GREAT! lol Can I get that as a ringtone? 😆
EVGA's response to this alone is why I'm probably gonna swing more towards them as a brand when I build my new PC in a month or 2. Big ups EVGA!
Sick video idea Jay!
8:42 That sound is definitly a inductor going below it's intended Frequency range and getting saturated
7:25 "That card is hotter than I ever remember it getting" On screen: "Power Draw: 424W"
Hahaha
The Killawatt was showing 600W draw from the wall
Here I thought my 2080 drawing 250w was high lol .
My 3080 draws 430w.
To me, their statement sounded like this: "There's no way in hell we'd accept responsibility for GPUs blowing up. But uncapped frame rate on the title screen is a bit silly so we'll cap it."
basically what blizzard said about the 60fps menu cap in overwatch
@@danielegalizzi8562 Correct, wonder how much these cards have slowed down from their initial day one reviews due to having to be clocked back in bios updates. Nobody wants to test that huh😏
@@danielegalizzi8562 Even if it happened on all GPUs it still wouldn't be their fault. That's the whole point of having an API, there's a standard for which software can properly interact with hardware. The idea of APIs is that 2 different systems can interact with each other if those systems correctly deal with the API. AmazonGames is not doing anything wrong here, they're making standard API calls, if the hardware fails because of this then it's 100% the hardware fault and 0% amazons fault.
Well. They are right lol. You bought a malformed gpu that will explode a game which a 1050ti can run fine. Guess the manafacturer sold you the ultimate snake oil.
@@rickroll9705 no, no, no, mustn't say such things. Let's instead focus on putting evga on a pedestal for admitting failed hardware designs and replacing the cards from their mining racks..... I mean test benches.
At the end, saying "Nothing happened" - as the power threshold is hovering at 130% of tolerance in the background.
I would rephrase that as; "Something has happened - it just hasn't destroyed my card yet"
Would be interesting to see the test with the previous game build pre-patch.
Highly doubt amazon is gonna share the source code of the previous build to be investigated
@@Hikari_Mavio Wouldn't need the source code to do this test. Simply someone who has yet to update their build with the menu "fix" willing to share it would work fine.
@@SonoftheDragon I'd wager New World won't launch without an active internet connection and being on latest patch given it's online only, and I doubt someone would take the time to make a custom crack for it just to see if they can brick a GPU with it.
I was going to suggest to see if someone had archive built the game, but if the report of the game needing a permanent internet connection is true then that would not be possible
@@DankerTVTimeCapsule New World is an MMORPG, so yeah, it's online-only.
20:41 when the mechanic asks you to describe the noises your engine was making
Though it doesn't affect me I am glad to see Jay is doing due diligence for viewers. Kudos to Phil for the Jay sounds mashup at the end, hilarious!
Crazy the power usage considering your capping it in Afterburner. I Have never seen that either
Jay: "safeguards are not designed to be hit and used over and over and over"
Buildzoid: I beg to differ...
yep, lol.
it's a different world when you go negative and live life bouncing off V-max and use the power limiters as the clock control, lol. realy sure rockets also live by the same "on the edge of failure" methods. B)
400 W power draw? My whole system draws that much lmao
Mine does 550 at the wall with OC... Damn, those Ampere cards are inefficient as hell...
@@Donnerwamp Not really just Ampere, in fact Ampere is more efficient (slightly so) than Turing for the performance.
Any card is going to be inefficient right at the edge of what the architecture & hardware in it can do.
@@Donnerwamp its 3090 u know xD 3070 is fine around 220w 3060 under 200
@@markjacobs1086 A 3070 is as fast as a 2080Ti (except RT) and both are 250W cards, so the cards are pretty much the same in efficiency (aside from RT). Then, Turing is made in a bigger node than Ampere, so this tells me that Turing should be more efficient than Ampere when comparing architecturento architecture. Adding the fact that the high end 3000 cards run far from the sweet spot makes me think that they are inefficient cards. Note, this doesn't mean they are bad, they are just inefficient.
@@Donnerwamp Sure, they're definitely less "efficient" in that way. However it's not like AMD handled that any better, both basically increased performance & power draw significantly.
Just keep in mind that with either GPU manufacturer you can actually power limit them this gen and get the majority of the performance with much less power usage.
I want to see Jay, Linus, and Steve play this game live and bet who first gets the DSOD.
Been through the EVGA customer service/ RMA process a couple time and I do agree it's a very high quality experience. Love EVGA for this.
"Software can't break a component"... no, software _shouldn't be able_ to break a component.
Yes, and that's on the software's side of it.
but the hardware just does what it's told to do.
That reminds me of osu! the famous 2D circle game pushing my cpu (Ryzen 5 3550H) over 92°C in average when Minecraft push it only on 80-84°C at most on my laptop (up to 96-98°C average with PTGI shaders, basically RayTracing, which surprisingly doesn't put my cpu in thermal throttle state).
Even a more detailed game like PSO2:NGS can't push my cpu over 94°C on High settings preset with a GTX1650 Mobile OC with an average 72fps.
@@jetah50 That's what hardware protections, like fuses, are for. They protect the hardware REGARDLESS of what the software asks it to do. They are of course more expensive to implement than software-based protections (firmware/bios/driver) and are not customizeable.
@@jetah50
The hardware can say no, though. There are power shunters, overclock protections, and all sorts of other measures on the board that if well designed, would prevent this behaviour.
In fact, the hardware SHOULD be saying no, in part because trusting the software to not give poor instructions is a security risk.
Imagine someone wanted to prolong the silicon shortage? easiest way is now to send software to nuke miner hardware en masse.
Evga/Nvidia needs to fix this, Stat. Regardless of what new world is doing.
"I think I need to try ANOTHER 3090" my mortal ears were unprepared for such words.
The jokes over. Influencers get cards. Simple. Its hard for you to get a 3090. Not him.
I had to RMA a 1080ftw hybrid due to pump failure and I had a new one in my hands within 4 days. Their warranty is amazing and they stand behind it for sure.
Around the 8:00 minute mark in the video, you discuss the "rumbly" sound. I noted this issue on my PNY 3070 on day one of beta. It was so loud, that it made my pure wood, formica Steelcase desk rumble and hum a low pitch. When I checked my temps, I was slamming about 78C. But, this went away around Day 3. I then checked and saw there was a bios update for my card for the new resizable bar update. I added this, and believe it or not, my temps dropped...but, only down to about 73C max. All I do know is that weird rumble, was there for my 3070...but, is now gone.
This is normal lol
ALWAYS when you want to test a bug, you "miraculously" don't get it xD
The ultimate proof we live in a simulation.
Jay should put it in a build next time, if a thermometer is melting from the heat, imagine what a closed environment like a random ATX case with limited ventilation will do to the back of the card
I will be curious to see what Zoid autopsies when he gets one.
As always you doing great review & benchmark
I love this. This is how you serve the community "I want to try a different 3090." We can't do this but you can and you did.
me reacting to the last 12 seconds of video: lmao "goddamit phil!"
Been waiting for this one
I really appreciate this kind of videos
EVGA customer service is amazing. Thanks for this video. I hope either it's already been fixed or they figured it out.
Really curious too see the last new world update data mined. There could be a lot more than the menu fix bundled in.
they forsure did something lol
lmao no. They just set a fps limit on the menu. How delusional do you have to be to think a regular optimized game somehow "breaks graphics cards"? The cards weren't properly built, and therefore broke under high loads.
@@muuubiee Or perhaps you simply do not understand basic PCB/VRM design. Uncapped FPS in a menu is nothing compared to the bench marks that have been run on 3090s. By thousands of overclockers for the last eight months(majority of the cards being EVGA 3090). But a horrible piece of software with a few to many zeros in the code in just the wrong places can always destroy hardware. Why do you think New world had a clause in their user agreement releasing them from any and all liability for user hardware damage?? Try a bit of humility and actual research before making grandiose statements.
@@alt5494 Faulty hardware is the hardware designers fault, not an unrelated game. They've clearly not built the card to manage it's maximum load.
"few too many zeros", what are you even talking about? Nops? I don't think there are any nops in the code, and nops should not generate more than a fracion of what calculations do.
Jay: “Watch and it’ll probably go back to 100”
Also jay: “103, 104, 105”😂
Noice
> Probably.
I like that Phil had fun with the outro
ok, the ending got me up cranked a bit, hahaha
Conspiracy time: the extra power is being used to mine Bitcoin....
When you burned yourself on the VRM, it was like watching Arrested Development where everyone knows the cornballer is exceptionally hot and exceptionally dangerous, yet they all manage to get burned at least once, and continue to use it.
I have to confirm evga rma service. They troubleshoot the problem and sent new card right away and had me put my bad card right into the box the new one came in to send back to them. No charge. It was 2 years ago and glad its continuing! Great job evga.
Jay is highlighting a problem with the 3090 ftw ultra that causes problems with the boost clock and power, there is a 400 something page thread on evgas forums about it and they have a specific rma process for that issue and will give a revised pcb with different components on it.
Jay: All the RMAs, replacement 3090's are shipping today.
Me: My non-existant EVGA 3090 FTW3 needs RMA...
Exactly what I was thinking lol
It’s been my experience with EVGA Ampere cards that when you’re seeing power percentages that don’t make sense against the board power limit, it’s because it’s actually showing the PCIe slot power percentage. Check that.
I'm currently daisychaining/running pigtails for 1 of the 3 8-pin plugs in my RTX 3080, simply because my power supply only has 3 dedicated 8-pin outs, and I use one of them (also daisychained) for the CPU power. Should I look into getting a PSU with at least 5 8-pin outs? (My current PSU is a Corsair RM750x, 80+ Gold. I'm looking at a Seasonic Focus Plus 850W, 80+ Platinum)
Great Job Jay, very good sir. I like this investigation. Kudos for EVGA. #EVGA
I RMA'd my card about 2 weeks ago just got it back I was not cross shipped and it took quite a while to hear back on each step
Did you get a working gpu back at least?
@@yourmother8979 so far so good
I think this is more then just the game… or beyond that anyways. The game exposed something that was already wrong with the cards.
yeah, that's so fcking odd
Nah it’s the game
@@static7415 ? It’s a card issue if it allows itself to overheat/accept a higher wattage than it can handle. No game causes card problems
@@calebharris4127 its the shit game lmaoo
We do know that 3090 and 3080ti definitely have VRM thermal issues that are caused by the physical design of the card layout so you might be right.
you shouldve include that (Turkey!!! ) at the end XD,, that was funny af XD
The hero we need thanks Jay
I remember back with my GTX 1080, in Civilization 6, if I zoomed in on a full sized hurricane (lots of particle and other effects) my GPU temp went up almost 10C higher than any other game load I could put on it, benchmarks, burn-in tests, nothing. Seemed like some kind of power bug with the shaders used.
Least the good thing about the RMA process is they can put it in their lab and try and root cause the failure. Least I hope they would do that.
That last 20 seconds was the best part of the video lol
Imagine playing for "about three hours" but only being level two. :)
Gotta trust jay with the unbiased straight to the point rootin Tipton information. Cheers mate
Haha this is the beat joke I have seen 😂😂
This guys as biased as they come he spent the entire video convincing you and himself it's not EVGAs fault.
Jay ranting behind the computer, meanwhile power limit reaches 132% 👀
which is pretty normal, considering he turned off it voltage limit and is just letting it suck up power.
Great Video as usual.
8:29 Those sounds like arcs. Like an arc happening over air, multiple times a second
That sound reminds me of cars being pushed over their destructive limit honestly
10:19 "you guys couldn't really see"
7:43 you meant this, or that was from only the cable?
Intel: "you can literally see it"
LOL
That edit at the end was glorious
much thanks jay! i do a twitch stream and watch other stream as well and quite a few people in chat are complaining about it. i sent this video to quite a few watchers.
Jay: This isn't EVGA fan fanboyism.
Jay, later in the video: I'm such an EVGA fanboy.
It's not possible to love a product and still be able to admit a flaw with it. So we'll just dismiss anything said.
@@ntomnia585 Yeah it is, I also prefer EVGA
@@Mikemk_ I was being sarcastic. Statements like the one you made put on the appearance that you don't think Jay has the ability to be critical of a product he really likes. It just seems shallow and dismissive. It's a common theme for some reason though.
@@ntomnia585 woooosh
@@ntomnia585 he is crying and drying his tears with the check that EVGA sent him.
8:28 is straight gold! 🤣😂
I'm definitely buying one this winter Cheers.
I have a MSI RTX 2080 SUPER and New World is the first PC game I’ve ever played (been pc gaming since C64 days) that makes my PC completely shut off. CPU temps would always hit 90-100+ degrees Celsius “IN TOWNS” only with crazy temp spikes really fast. So if I stayed in towns with a lot of players like Everfall or Windsward too long my pc would totally shut itself off all the time. It was so bad it killed my windows 10 and I had to reinstall windows. GPU temps never an issue ever. So every time I was in town I had to change to low settings and 30fps to not get insane temp spikes. I searched the interwebs and CZcams for days for a fix. The one fix I finally found that worked (for the most part) is having to create a customer power profile in windows for new world.
Sonofatech broke his gpu after lowering the settings to high
THIS
EVGA'S actions on this are great to hear. Can't ask for more than that.
Agreed. Many card owners even got new card inside 72 hours, their actions after problems are good, cant deny that. But i hope their engineers look at this problem, because any software should never be able to blow card up.
@@yesnoyeswait4306 I'd bet they screwed up somewhere. But just the cost of these replacements will be plenty of motivation to fix what led to the issue I would hope.
Would this runaway power reaction still occur if you hard-capped framerate by using G-Sync with a G-Sync compatible monitor?
I have a weird off topic question, do you do reviews of gaming lab tops in the same way you do reviews or pc components or setups?
Just curious as to whether or not Jay ever considered the reason he wasn't able to recreate the problem was because he was using an open testbench instead of a pc case. I mean, my guess is that most of the PCs that people build in are in tower cases right?
@Misha Kyouko Yuuki cases are hotter than open test benches, notoriously hotter.
@Misha Kyouko Yuuki he has a very strong AC in his lab.
He can't recreate it because it's been patched.
@@jamesjx again, no proof thats even the issue from the game...
@@h1tzzYT the game was definitely doing something.
I know it made my 1660 Super freak out, and one of my customers brought in their ftw 3090 after playing the beta.
It runs fine now with the new patch.
I do feel it's some sort of issue with the GPU, that the game just accelerated whatever the issue was.
Chances are it would have eventually happened with any heavy game.
You failed to consider the possibility that the Evga cards are especially dodgy. The EVGA hardware itself may be a problem.
hmm so you mean that if you overclock a card from msi and evga, you are gonna get the same results?
software code can cause the limiter or throttle to stop working properly and when this happens, it will have hardware damage.
each hardware is made differently but software code should not override or cause incorrect load to the hardware causing issue.
kaspersky Internet security software has cause hardware damage before to a number of laptop motherboard due to the 100% CPU load overriding the cpu throttle causing the component that is sharing the same heat pipe to be damage. Doesn't that sound familiar to this case? and guys remember the clause about overclocking.
@@Hikari_Mavio Hey you may be right but I may also be right. Problem is that there are lots of possibilities and we are speculating what could have happened. My problem is that Jay failed to speak of the possibility of it being a hardware problem.
It doesn't necessarily mean that it is evga's fault. They may have recieved faulty components in the cards they put together. The game is just presenting those faulty parts. The parts may not even be faulty, just not meant for the amount of load the games is putting on them.
Seems like people have already forgotten about the EVGA GTX 1080 FTW3 incident... Tbh EVGA is really bad at engineering stuff, but at least they have a decent after sale support.
Same issue happened to me in MSFS 2020. The loud noise scared me to death. Shut off computer immediately and waited. Had just downloaded the latest NVidia drivers for the EVGA 3090 couple day before. It happened as the game was loading up a flight.
Dang I wish we had a micro center here in Oregon. We have to deal with Fryz 😩. Love the content Jay, keep doing to lords work. I would love it if one day I had enough money to have you build a custom PC for me.
Funny how they’re getting the replacement GPU’s within 1 day. I still can’t get one after 10 months
they got a few spares so the people dont need to wait weeks for RMA
RMA inventory is a thing
@jack mehoff why would a company that didnt give a shit make sure people are getting replacement products unexpectedly quickly?
how do you see a company making sure people with a defective product get a replacement swiftly and say "its cuz they dont give a fuck about the average customer"?
@jack mehoff ur whole comment is lol
RMA stock and new stock is different. Unless you want Evga to get sued for selling used products at full prices.
OMG, the end killed me. So good
I love your editor @20:42
I had an EVGA PSU that developed a problem after maybe a year of use. They replaced it with ZERO hassle and, if I recall, they shipped it out like two day something. It was super fast and no hassle.
Even though we still are in a shortage I want to give kudos to EVGA for stockpiling 3090s to be able to satisfy RMAs instead of selling them all at inflated prices / abnormal win rates.
It's literally law that they have to do that
that's only 30 cards btw
@@FG-td4vs They have to. But they could still choose not to if able to make a quick buck by selling the stuff at inflated prices in a shortage. Sorry for writing this short answer. CZcams deleted my previous anwer and I don't want to compose it all new.
That backplate do get realy hot when under load (at least in my experience with my 3090 FTW3)
Hmm I wouldn't know seeing as how I can't even afford a 3080 at these prices
I don't think that's limited to the FTW3, I think it's just a function of a power hungry card and having VRAM on the back. Unless you have a Water cooler that is getting bother the front and the back with direct contact (grumping at you Gigabyte), you're backplate's going to be fairly toasty. I actually mounted an additional case fan just to move more are over the back even though I'm using an AIO cooled card.
I find backplates tend to get very hot no matter the card. Even with my Strix 3090 OC, the backplate can get uncomfortably hot to touch while the VRAM temps and GPU temps are well below the throttle limit and that's at stock. My old 2080 Ti FE backplate was also was extremely hot and that didn't have extra VRAM chips on the rear.
Would an intake fan or case fan help with this? I don't think I've owned a card with a backplate.
@@chillnspace777 Should help if it's blowing a decent amount of air over it. I use an Evolv X case with a custom made mesh front/top panel so I can let more cool air in and out as its not great a great case for airflow at stock. Even with 3x front 120mm Corsair ML fans blowing air at 100% the backplate still gets uncomfortably warm.
The GPU core temp is usually at 68c max (GPU fans at roughly 75%) and the VRAM junction temp peaks at about 80c to 84c (GDDR6X just runs toasty but well within spec) depending on the game, so temps are great. As long as the card is running with good temps though I wouldn't worry, just means the backplate is doing a decent job trying to cool passively.
I would like to see you test some specific game conditions, , 50vs50 wars, the level 50 area Edensgrove, and the most busy town you can find… also, Corruptions, while in massive groups of people.
This game runs fine while standing in a forest smacking a turkey.
EVGA always and forever. Most amazing customer support in the business hands down 🙌🏼