#graphics-cards
1 messages · Page 74 of 1
That hasn't happened. It should be up to the retailer or brand to pay for it the 2nd time imo
It's not your fault it happened twice, they failed to deliver
The typical norm is the consumer won't foot the 2nd bill in this situation
I’ll email them and see if they’d send me a return label and a new RMA number, my plan for the card initially was to RMA and test it so I could sell it to my cousin who does graphic design and try the 4090 if I could find a waterblocked version
You have a water loop?
For my cpu right now, I did have the 3080ti waterblocked for a while before it started acting weird. Turned out it was the mounting pressure for the water block that was causing issues but that’s besides the point
I’m just testing the card before I give it to him to try before buying it, I don’t want to give him a card that fails in the first months of owning it
Ok fair enough
Reason I say is waterblocking a 4090 does nothing
I went from suprim x air to an alphacool block
My temps went up 10C and my vram gained +200
Which is suggestible to be plain worse tbh
Card is quieter than my loop when stock lmao
Just some food for thought, the good 4090's are almost too good lol
Which is why I didn’t water cool the 7900xtx lol, that ac unit of a cooler didn’t let it break 65 degrees
Yeah, a lot of xtx's are like that
But yeah I would say, all other variables out, my only other guess would be your e cores might be unstable
Also btw
Go check this out
https://wccftech.com/intel-13th-14th-gen-instability-issues-buggy-microcode-etvb-fix-bios-fix-0x125/
I know what I said about evga bios', but a fix is a fix, and it's a big one. Worth the pain
We don't know when this will drop yet
Oh, that’s the microcode thing I was looking at on my bios last night, I’m on 0x125 microcode
Never mind, I double checked chat and it is NOT 0x125 but I’ll be looking for that update
If it doesn’t arrive for my board a new mobo is a lot cheaper than a cpu
The fact they hid this internally is a huge deal lol
Yeah, but the fact that i was informed about the power limits for my cpu is still a big help, at least until that patch is released
All good mate. I think at the very least we've made fair headway into this
It wouldn't surprise me if it went through rma without a confirmed fix
Happens a lot in the tech space sadly, very annoying when it turns up
while gaming my gpu is reading 76C, but the hotspot is 91. Is that fine or should I be looking into a repaste or something?
instability continues 
Hmm it's almost like the article that has a typo in the title wasn't reliable. Who could have seen that coming?! #cpus-mobo-and-memory message
What gpu
Sounds fine tho
Yeah you're fine
Ok thx
Yes they'll deny the claim as it leaves them legally exposed if it becomes the lynchpin factor that caused hundreds of thousands if not millions of chips to self-immolate
This was their own internal investigation lmfao
Them having to potentially replace or recall tons of these chips would chop their top off the business no doubt
Like the literal text excerpt is damning 💀
They're fumbling rn 
Would switching to ryzen be the better option until intel gets their stuff straight or is it better to wait for the microcode updates
If there's nothing wrong with the cpu rn, just use what's safe until further notice
Just have to wait and see when the gpu comes back
🧐
👀
Apparently the instability and so on has nothing to do with eTVB
At least according to any companies asked
Something something save face something corporate liability something something
I'll wait until the dust is settled before I believe anything. But an internal nda'd investigation is a big one
TBF it's Igor
So like I don't doubt he's telling the truth that Intel found the root
But I also don't think it's eTVB
Since Intel can't be dumb enough to release code that'd literally fry their top chip and ONLY their top chip
@odd siren also hold on didn't AMD have a big scandal about top-tier X670 boards literally detonating X3D chips?
Though there IIRC it turned out that mobo manufacturers went full stupid and had bad BIOS stuff
So like
They absolutely can be as they didn't realise they were pushing their cpus way above the design limit
It feels similar but AMD probably handled it better?
Overclockers knew about this problem BEFORE intel.did
Oh?
This degradation has been known since mid 13th gen
I mean yeah OC would reveal that kinda stuff
But many people didn't ever OC in the sense they just let the chips run to the numbers in bios
Long known running ICCmax at 255A saves the cpu from multicore death
Um yes they do
That's called the darwin effect good sir

So you have to TECHNICALLY overclock to get advertised performance so if something breaks you're SOL
There are people who benched at stock with cinebench, after 3 cinebench benchmarks, only 3, chip's degraded
30 minutes in cinebench is KNOWN to degrade 13th and 14th gen even at stock
I didn't think I can have more negative feelings towards Intel than I already did
That's effing awful
Yes why do you think I'm not making a huge fuss about it
I've known about this for a year now
I only knew that your rig can crash if you push it
But like degradation that bad that quickly is borderline scam on a like 800 dollar chip
AMD went through the same thing with static oc on the old zen 2/3 process nodes
I wouldn't hate if my PC was like "Yo you're really pushing the limit, I'm not happy I'd rather just shut it down to avoid anything bad happening"
But degradation at stock speed is probably scam
And it's not like LGA1700 was a new socket or 13th gen was that revolutionary new professor tech
Which is why they're fumbling horribly behind the scenes to fix this as being unable to run as advertised leaves them legally upheld for a number of products
They're under a massive risk exposure
Like I get the early teething issues with Ryzen 7000 with the new socket and all that (though X3D chips desoldering themselves is sucky)
But like it's an ESTABLISHED tech
That's why there's all this "They didn't actually find the issue" "No it's not that" bs
It's classic smoke and mirrors
That's the trillion dollar question
700 and 600 SKUs affected too?
14600 I believe yes
Cuz that'd make it from a disaster into a cesspit of lawsuits
900 SKUs already bad
But if it's more widespread then it's just gonna be a massive disaster
We don't have a defined clock and vcore definition for degradation
13700k?
I was eyeing that for my dad's upgrade
I'm probably not gonna get him that upgrade in that case
Wasn't it vcore voltage control failing with the X3D degradation thing too?
Somewhere around that under full load is known to degrade, but current also plays a factor hence ICCmax = 255A
Or for intel
ICCmax 307A
That was the vsoc rail. AMD yeeted it way beyond spec
If you touch voltage at all on x3d it's basically guaranteed death over 1.2v
Let alone pbo overclocking being mid, all core oc is basically death guaranteed
X3D is fundamentally very voltage sensitive
Isn't that why OC was meant to be locked on it?
To avoid someone accidentally killing their chip
Kek
bruh no mans sky makes my pc not bright lol
It has HDR mode
So if you have a good HDR monitor then enable that, but if it's only HDR400 or not certified at all, disable that or it'll screw with the colors.
The "good" HDR monitors start around $500 these days
what could be some reasons my gpu would spike randomly to 100% and freeze my pc for a few seconds? Its a 6700 xt and it usually runs 60-80% during games. My graphic drivers are up to date
?ddu
DDU (Display Driver Uninstaller) Guide by Pato
https://github.com/NeweggTechie/Graphics-cards/wiki/How-to-use-DDU
huh
If you got that for 550-600 it might be not bad tbh
Make sure you have a good psu for something like that
alr cheers mate
u7h anyone know why i just logged onto r6 and i lost evreything? its like a new account it says?
Seems like a common thing and the only thing I can say is contact Ubi
https://www.ubisoft.com/en-us/help/contact
No one else with the same issue ever reports a fix
Which is not to say it doesn't get fixed, just if it does get fixed it's by support and they don't report that
lol there's a whole page for this
https://www.tiktok.com/discover/r6-account-reset-glitch
new xfx gpus
The ones where the fans pop off?
White is gorgeous
There is a lack of white AMD GPUs right now so any new white model is welcome
yup
the white is only on the 7800xt by the looks of things
box art for the xtx is amazing though
would look good on a shelf
Hey guys, does anyone have the reference RX 7900 XT card? If yes, how is it?
Some say it's not great compared to custom models
Most of the time custom PCB cards are only a couple % faster than reference
I'd not worry that much
Relatively loud but relatively small vs other 7900XT's. That's about it for differences.
Thanks to both of you. Found a really good deal on the reference so I was curious about differences. Guess I'll go ahead and buy it!
Oh yeah, reference vbios
@fluid grotto
Forgot that last part but sorta comes with the reference cooler territory anyway
Rx6600 are they any good gpus?
If you can spend $235 for the 7600 it's a pretty noticeable jump in perf
Like 20%?
yeah around 26% with av1 support
you need to do av1 above 1080p tho
Nah 7600 far better
That's probably the 6800 XT/7800 XT you're thinking of, that's a lot closer.
Closer to 6700, no?
That is tpu; are there any other sites with a good comparison tool for gpus?
3dmark > compare graphics score with slight amd bias in mind
Didn't notice that it is actually hilarious 
Wdym by that compare score with slight bias in mind?
Amd gpus under score relative to Nvidia/intel you mean @odd siren ?
Radeon cards score better than their relative counterpart from nvidia in time spy
It's a small bias
Less than 10%?
But even now time spy extreme is still very accurate
Like let's say you see a 6800XT do 22K but a 3080 12gb do 20K
In games normalised it's more akin to very similar raster between the two
But if you remember that slight bias then it's easy
I wonder if some of the stuff timespy does fits in the Infinity cache
That would explain it
I'm not in the slightest aware why it does that but it does
Intel also gets high scores despite AMD's capable multithread with X3D
Is the Mercury always red or is that LEDs?
dGPU only mode enabled
hey waht should my mouse polling rate be at? i mainily play fps games?
on reddit it says should be at 500hz? instead of 1k
I'd say highest setting unless it's acting weird
idek how to move my polling rate or change it i mean
You'd do that in the mouse software
I agree with setting it to the highest polling rate
It's straight up a fact that finer controls and details are improved with high polling rates
If a game struggles with it, crank it down, etc
bro no shot
im having the SAME issues i was with the last pc lol
wtf
what are some programs i can download to make sure my ram is running fine and same with cpu
so whats weird if my graphics arent high and i put them on low my game feels like crap? so i assume my cpu might be messedup?
on high settings
my cpu temps
Might need to just set a fps cap
What 3d mark spytest should i run on my pc
Timespy
Normal time spy
I read that the pcb for my card and the reference pcb is nearly identical other than the missing screw in the power input in the top, would it work with a reference water block if I shave some acrylic
Seems wacky
is this towardsme or him
Both
The fps jumps around without temps or clocks jumping around
And drilling into the acrylic seems sketchy
bruh
talk about not smooth
how do i fix this
maybe my ram being at 6400?
maybe it isnt stable?
This looks completely normal, no?
nah look at my lows
Are actual games choppy?
This is 3dmark, no?
he asked what game not the monitoring software ig
I thought the app was 3dmark
first pictures it was but the pic i just sent is capframex
how many loops
oh, ight lol im confused on waht your asking xD
this says 5600mhz? but i got my ram set to 6400
5600 is the jedec
oh ight
Is the memory unstable?
Then put it to 6K @solid glacier
Yeah set it to 6000 then if it's stable
If it's stable at 6000 then that's a sign that UCLK was unstable
How can i check if i can put my gpu in and see if the psu can handle it?
what gpu and psu do you have
I have a asrock challenger pro 6700rx
And my pau is 550 wats
what is your cpu?
asrock reccomends a psu of at least 700watts for that gpu. I personally wouldn't go less than that.
but it depends on the rest of your specs.
AMD Ryzen 7 5700G 8-Core, 16-Thread Unlocked Desktop Processor with Radeon Graphic
Yeah I feel like you're at the high end of what that psu would be able to do. It would have risks using it imo.
you don't usually wanna go over 80% of what your psu can handle which would put your psu at about 440. The power draw on that gpu from what I can see is 230 watt at max, cpu is at 70, ram uses 5-10 watts so if you have 2 thats another 20, if you have hdds they could use up to 25 watts each, gaming mobos use 45-80 watts, m.2 would use another 10 watts max.
Everything except the HDD would already add up to around 410 at max and then you still got any LEDs and Fans and whatever else you may have in there. It's a bit to close for me, but it may work. Pushing a PSU to the limits could have major issues though and could fry some parts if you're not careful.
80% load is perfectly fine if it is a good unit
you'll fry em only if you load up bad psus
like the thermaltake smart
use it too much and you increase the risk of catastrophic exploion
but still
the gigabyte pg-m is an exception tho 
way too high protection trip point
Hey the protections work perfectly
Once
I'd worry about the 80% rule only for some PSUs. Usually those that don't like transients with transient heavy components (also technically bad PSUs, but bad PSUs don't get recommended anyways). Kinda just SFX ones I guess, or certain old-ish models (prolly from used market).
Or power efficiency. I think it's 70% is where the average PSU is most efficient? Can't quite remember now.
50-60% prob
Well I'd worry about that rule for anything B tier and below due to personal experience, and considering his was from a prebuilt, I'm not willing to trust that the manufactorer used a good psu lol. But I follow that rule as a general to be safe plus the manufactorer of the gpu reccomends 700+
Over time a PSU will lose some of its peak power capacity due to capacitor aging. Depending on the usage level and capacitors used, you can lose about 5-10% of your capacity every 5 years of use. It won't be an issue here if he's just looking at a 4070 ti for an old 1250W PSU.
Even assuming the worst, 20% loss, that still leaves PLENTY of headroom.
This a good build for the price ?
I would change some things
Part List - AMD Ryzen 7 7800X3D, GeForce RTX 4070 SUPER
Aio isn't needed but figured you liked the aesthetic
Part List - AMD Ryzen 5 7600, GeForce RTX 4070 Ti SUPER
7600 and 7700x are within a few % of eachother if just gaming
which is better
Ah my response was in response to fiery asking about using a asrock challenger pro 6700xt rx with a 550 watt PSU from a prebuilt
Also fair. Been a while since I thought of budget/prebuilt stuff do forgor that low end safe PSUs can be encountered. Just want to be careful for words, you kinda made it sound like >80% is scary for all PSUs.
It's typically a rule I try to follow because I went over that limit with a B tiered psu which was considered pretty safe, and it ended up frying my mobo and a couple other components when the psu gave out. Not saying it's common, just a better safe than sorry thing when most PCs won't really hit that limit with something 750+ and 750 watts aren't that bad of a price.
Depends on what you do
1440p gaming likely the first one
4k gaming or gpu focused workstation tasks, second
for 1080p max settings and also using giants editor
7800x3d one
Why are we doing a 1K+ pc for 1080p???
good point
Wwayyyyy overkill
@plush garden https://pcpartpicker.com/list/DKyrkJ
Part List - AMD Ryzen 7 7800X3D, Radeon RX 6750 XT
Max, max, maxed until horse is twitching
Would preferably do this honestly
Part List - AMD Ryzen 5 7600, Radeon RX 6750 XT
Any higher and you go just, way beyond the pale for 1080p situations
RX6800/7700XT at the absolute most for GPU
well because i want it to last for a long time and also i want to play warzone to and get 200 fps plus
i would suggest an upgrade to 1440p soon
^
out of yours which is the better one for 1440 with high fps
1080p is like using a ferrari in a go kart track in a gpu any faster than 6750xt
Part List - AMD Ryzen 7 7800X3D, Radeon RX 7900 GRE
same answer, 7800x3d 4070 super
thats just a cheaper option for same perf, and the cheapest not bad 1440p monitor id suggest
if you can swing another $40 upgrading to the g27q for monitor is worth it
so go for this ?
yeah
bet
give me one sec to check something out
ok

Even if it's warzone
Use newegg combo up savings to get the 7800x3d, mp44l, b650 eagle ax, the teamgroup 32gb white rgb kit, and be quiet pure power 12m 750
@plush garden
alright
$737 for all of the above I believe
bet
itll be less than 737 idk what newegg mobile is doing
i cant fint the deals
send link if ya dont care
this ?
youve gotta select the individual parts yourself
which is better ryzen 77800x3d or the i714700k?
7800x3d
what is a mp44l
ssd
this the the ram i should get?
ye that works
Much better than crucial pro omg
this better?
looks good, is psu in there too?
i would not get the p41 for $105
mp44l will do the same thing for gaming at much cheaper
Please don't get a p41 plat
VP4300 Lite will almost match it for way less
Use that $37 on other stuff
ok
hey would this channel be okay to ask for troubleshooting questions?
regarding graphics card?
Yeah go ahead
Does anyone own 7900xt but Asus TUF?
@sharp trellis I've owned and worked with a lot of rdna 3 if that helps
Ever gamed on it?
at least.
Ofc
Hm
Nonstop crashes
Ok
Ok
I've tried every possible fix, and other ways..
Let's keep it chill my friend I'm here to help
nothing works so I rmaed it.
Yes sorry I didnt mean to sound like a record going rough i just was frustrated.
I can tell
its fine, it was many months ago.
i ended up rmaed it and sold it out of anger.
apparantly buyer said its working well
on games i even tried to play,
so now i just ordered 7900xt Asus tuf oc its coming tmw morning
i hope it wont crash on me again.
I will be doing DDU + new driver and Internet disconnected
1st off did you try the rma card yourself
no. 💀
Ok
I'm at 2 conclusions at this point tentatively
My first, being that the card was indeed messed up, came back, it was fine, now it works
My second
The card has some strong incompatibility with your system and refuses to play nice. Very rare symptom but happens
it wasnt same gpu when it came back, it was a whole new card.
huh
I see.
I wonder what it is
but that has changed, I am on Windows 11 with AltasOS optimizations.
Very rare. Sometimes you get a piece of hardware that just hates the entire thing, refuses to work and has terrible issues
Ram, cpu, gpu, psu
Ssd
All of them can get this
And it's perfectly fine in a different system
I've seen it once myself, there's an example I remember from jayz2cents from a really long time ago, he had a 10th gen system that refused to use a kit of crucial ram
I myself had this with my 6900xt
Okay, my pc has underwent new changes.
new cpu,
maybe ram i cant remember if i swapped out to 2x16gb sticks coming from 4x8gb sticks
Same old psu
It's a very rare issue
I mean it's a wild guess if this may have been a factor
Often times it happens it results in an rma of that new part
And people usually get a working system later
I've seen others with this particularly on ryzen systems
Cpus that just refuse to work
Apparently common for ryzen
So uh yeah. When you get it, ddu via safe mode. Give it a whirl, see how you go
If you don't have issues with it rn then odds are you're fine
Make sure you land on a good bios and stick to it basically
I see
so i hope this time its gonna work beautifully
because i want the 20gb vram
and raw power for my alienware DWF monitor
Alr
Okay
Tech is either happy or completely mad
Well there's that but I mean the pcb
Ah okay.
Asus use a very unique pcb layout to meet rdna 3 board spec
It's actually very efficient
One of the few that aren't reference design
it pulls 310+ watts
oof
Anyway, give us a shout if you run into any issues, just remember to ddu and shutdown via safe mode first before you swap it over
Odds are whatever it is there will be an answer
not my first rodeo, if my own fixes that i know of and dont help me
ill come back here
apperciate you telling me its a rare issue
I mean I speak from the 1st person about most things
part of me wished i tried the rmaed card but well, that money from selling it got me a nice DWF monitor so win-win?
i liked the amd gpu software
esp recording
but formats lol i had to convert it to an MP4 file
do you know a way to have it send as mp4
or whatever most compatiable with Discord
Game bar works well lol
you got a point
Especially with hdr involved
Indeed
then i forgot my d41 case is not a real free estate pc lol so I went with it
let me check my total
amazon lists it, and direct from amazon
$800
i paid $763
amazon employee discount came in clutch
Before or after tax
after tax
Not bad
im looking forward to seeing dwf freesync premium working in bias with the 79xt
I wanna be sure of all possibilities before goin for it.
So, playing a one game often in 4k max settings, on a 3080ti FE since 2020.
overall GPU temp is 72 C and VRM 90-94 C this summer. Was goin for 12 hours straight this day.
Gpu continues to blue screen all day. I tried updating drivers, re-seating, going to the "fail safe" driver and lowering settings to 1080p low and it still crashes.
Is my GPU dead?
What are the blue screen error messages?
Couldn't see, it flashed for a second then auto reboot
Never recorded the log either to storage
Ok, I can analyze them and see what's up.
Copy the contents of C:\Windows\Minidump and zip it up. Then DM it to me. I can help narrow things down with the codes.
it isn't up to date, last dump it has is from january
That would indicate either the dumps are disabled or it's definitely not the GPU
how do i do that?
ima send the files anyways
GPU errors don't cause it to skip the memory dump, they're pretty good at recovering.
The kinds of things that cause that are mainly CPU and RAM.
Start > type ' reliability' first option should be it
Then you click on the errors and examine the codes it gives
just spams of
windows shut down unexpectedly
The 4 dumps you sent all indicate the Nvidia driver as the culprit. Have you tried DDU?
?ddu
DDU (Display Driver Uninstaller) Guide by Pato
https://github.com/NeweggTechie/Graphics-cards/wiki/How-to-use-DDU
yes, i belive we had talked about this before with driver issues. i had thought it was driver issues, turns out the GPU was sagging so it wasn't makin proper connection, then fixed it with a support bracket.
so driver is out, pcie connection is out, im not sure where else to go from here
So if those errors have been solved, then we're back to "why is it instant crashing and skipping memory dump" which is usually a CPU or memory issue.
i've done a memory stress test before with the program that was recomended before and results came out perfect. so maybe the cpu is the fault or, i have to run the memory test again
Try a CMOS reset, that'll reset the settings just in case there's a corrupt or bad BIOS setting
?cmos
How to clear CMOS (works for AMD and Intel)
https://www.intel.com/content/www/us/en/support/articles/000025368/processors.html
Defines CMOS technology and how to clear CMOS using the jumper or battery methods.
I said set a lot in that sentence
just remove battery and put back in right?
How could bios mess everything up though? Im doin the reset now
If one little setting gets corrupted suddenly a 0 becomes a 1 and your entire boost algorithm is messed up
HOLY, maybe i did that... probably... i can't remember
Yes
I've got some old dumps from December and January that are no longer relevant
Mmm
So no details?
They were all either "page fault in unpaged area" or "video tdr error" and all of them root caused to the Nvidia driver. They were fixed with a sag bracket.
Uh, just removing and reseating in a short time is okay? Or did i have to wait? My XMP toggle is on
No it takes about 10-15 minutes to clear
SHOOT! okay, turnin off to do this
Pointing to yeah likely not the gpu tbh
More ram, vcore, imc
Or psu
Thank the beautiful overclock psa
We should really pin this somewhere
Page fault being common would point to ram imo
Same with sudden restarts (or psu)
Tech what's your cpu and psu
CPU is 3900x, psu is a corsair RM 850X
Ok so we can rule out the psu then for the most part unless you're running the HOF vbios
HOF?
If you don't know what that is then great that rules it out too lol
All i know on bios is XMP, some default fan control, and boot sequence
I don't think the hof bios works on FE models anyway
I'm suspicious of ram or maybe dead psu cables
Though, i did use amd ryzen... master...
Yeah its OC stuff...
It may not be, but I'm immediately pointing at it lmao
WELP, hopefully!
Once bios is done in 10 min, ima boot a intensive load game. Usually crashes after a few min or something.
Run it ryzen master disabled
We can try and skip a step here, save what you had, disable it and just run stock + xmp
We'll see if it still happens
It's just for testing purposes, we can give you a decent overclock run down in 10 minutes to get you up to something that properly works if this ends up being the cause
I mean there's that but also properly doing ram >>>> ryzen master any day
Oki, time to boot
Bios has been reset!
And ima check ryzen master anyways and default it or naw?
K, says its already reset. Then ima just uninstall that later...
Ye leave it default
Loadin an intensive game now
Just make sure xmp is on basically
Similar loads before crashin problem so far, now to wait and see if it'll poof to blue
20 min in and not dead yet... seems promising
Ima try the main problem game. Usually that insta crashes
Nope, crashed upon loading
Nope, screens flickered and just reboot
No click
describe the flicker
Loading screen came up, both monitors flashed black, loading screen was stuck frozen on main monitor, flash black on main monitor, then rebooted
Is there any OC on the gpu
I don't belive so, i reset the power option back to default before
Just double check it's defaulting
If it is then we should download OCCT and do a gpu 3d test or system power test
Gettin MSI after burner up
Okay, links on those?
OCCT website
Ocbase is the home of OCCT, the most popular all-in-one stability / stress testing / benchmarking / monitoring tool available for PC
Okay, ima do gpu first
Is the system power test also 1hr long by default?
Mhm
If it didn't crash during the gpu or power test then odds are it's probably not the gpu
3dmark stress test on a loop would be another good thing to try
54 min through the 3d standard test, no issues
So what the heck did i do???
Running power soon
Starting power
Go check reliability history to see what the error was that it reported
Wait nevermind, its just.
"Didn't properly shut down" again
Sorry
Windows logs
System
Then scroll until you see an error or more at around the time you shut down
Shut down unexpected
Shut down unexpected
TPM error
the "shut down unexpected" just repeat those words. doesn't go into detail, this is the only one with some info
tpm data
wai, i remember that code... TPM is just a storage issue right?
Trusted Platform Module
It's part of the motherboard I believe, with tpm 2.0 it can be on the cpu as well
Sorry I went AWOL a minute @frank dagger
Shut down unexpected would hint me that it's possible to be the IF
What xmp is the ram
XMP 1, can reach up to 3600 Mhz
Which motherboard are you running
Aorus B550M pro-P
Bios is up to date
Jesus
Based board
Hm
Can you try this for me
Go to bios
Set manual frequency with XMP on to 3200
infinity fabric to 1600
Then try the power test again
FCLK frequency?
Infinity fabric is FCLK yes sorry
So where do i input the 3200 onto?
In bios
You go to advanced section
F2 or F7
One of those two buttons
Second menu
Right next to XMP there will be Manual frequency and FCLK
Theres:
XMP frequency support
System memory multiplier
UCLK DIV1 MODE
Advanced memory
Show pic
system memory multiplier 32
FCLK 1600
F10, save and restart
UCLK div1 mode you could also change to 1 to 1
UCLK=MCLK
Doesn't really matter tho
Crashed: Uncorectable_Error
Nope...
"UNCORECTABLE_ERROR"
Blue screen??
Yeah, with that error name
I'm not aware of that error specifically
But it never showed the "WHEA" part, i find that wierd too
Hmm
Can you go to bios
Now
I don't remember where this is
Look for C states
Disable c states
Then we'll try power test again
If I could remember where I'd tell you lol
It's in one of these menus
I think 2nd or 3rd iirc
Global C-state control
Aaaandd disabled
Wish LA has PC technicians, got nothing close by
LA as in Los Angeles? Because you could try microcenter if you do need to bring it somewhere
Fal
Or Louisiana
Not that local lol
4 hours via public transit
Wait I'm dumb I'm thinking of Cali
The entire LA area, including the 'burbs? Hell no, longer than 30 minutes even before you take trafiic into account
Wha... It takes 4 hours to get from the coast to the next state East, what are you really thinking of lol
Hm a lot bigger than I thought
30 minutes by F15

America has such weird like, aspect ratios
We need computer technicians in cali. Barely any places
It fits in australia yet somehow you guys take super long to go places
Cali is the third largest state, larger than most European countries
Usually 15 min into the test it fails, so I'll see again
wtf it's so big
Guys you need to draw USA better wtf is this
I always thought it was like a third the size of texas not the same size
Look for one with Alaska overlaid for comparison
YALL WHACK
One state is 2/3 the landmass of the rest of the country lol
Jesus christ
Yet it's one of the least populous states
Oh yeah I can see it now wtf
I thought the state was 1/3 the size
Like honestly
Cali big af
Like imagine florida size
Last I checked we were also the eighth largest economy in the world, producing about 1/8 of the country's food
Florida should have stayed swamp land where no one lived
You say that but there's a whole country over here staring at your statement
🇦🇺 👀
No but really all the former swamp land has foundation issues, and that's before the annual 5 hurricanes that pass through
A fair bit up north is like that here
You go fishing in marsh and swamp
Amongst the crocs

With our less regular cyclones
Wth



Crashed: this code
That's generally infinity fabric innit

Ima continue this later. I've been tryin to fix this for half the day. Its terrible
Related to ye
All good mate
I'm pretty sure though it's the RAM/IMC now
WHEA errors are one or the other
Solution: MOAR VOLTS
Consistent WHEA is IMC iirc
I wanted to binge a game this weekend but this happened. Its a long 24 hours out of my weekend...
Unplanned maintenance is never fun
If you had a few other errors like PAGE_FAULT that would be more ram related
So yeah
Most sus of the IMC/IF on the cpu rn
What is IMC?
So ram or motherboard?
CPU
I'm not kidding about this btw, VDDP, VDDG, SoC voltage up often helps
We may be able to fix it in the next sitting
But it'll take time
If not, 5700X3D kinda slaps tbh
💀
I suspect the auto oc of ryzen master may have had a hand in it but I could be wrong
Auto OC may have messed with a voltage and degraded something yeah
Now this is a CHONK Boi card @odd siren
But damn i didn't even bother to check the size specs Holy 💀💀
It might not fit in my d41 case crap
Sadly it didn't clear
It was hitting my psu box
Taking out psu box to test the 7900xt
Before going full new pc case install
Fans spin up to 100% on 7900xt GPU
Is that normal?
29 c idle
Yes it is
All case fans set to max.
max.
Okay it's fine
Now testing games..
What profile do I use cause my game is acting weird
Like
It sharpens constantly
test mem prorgram, extreme profile right?
Never mind fixed it
If its the CPU i have 2 atm to swap in. A 3600 and 5600
So i may try if thats the problem
1 hour in and nothing bad for RAM
I stopped the RAM test. If there was a problem it would've came in that hour
Ima do the OCCT CPU test then
Not cpu, cpu + ram starting
no error, so it's either motherboard or PSU problem now theb?
Had an epiphany, what if the power strip plug socket was a problem?
I'm usin some industry office power strip, maybe it was done for its time???
I'm plugging directly to the wall socket
NOPE, still crashed
Got a question: so i got offered an rtx 4070 ti super (msi ventus 3x) for 950 cad and an asus pro art rtx 4080 super for 1200. Which one would be better? Kinda stuck because theyre both really good deals imo.
Id brefer nvidia
what do you do?
if you're solely gaming, nvidia is bleh usually
in terms of perf/$ ratio
Do this
https://www.overclock.net/threads/memory-testing-with-testmem5-tm5-with-custom-configs.1751608/
Anta777 config
So far so good?
Both awful
Like
I've seen 4080 supers for 950
Oh wait cad
Hm
I did, no issues
Do this
Run the vt3 test
Yes the website has no certificate, it's the official site
Vt3 via component stress tester
The 4080 still better cos it's not ventus
But both kinda meh for used price
Yep I see no hard crashes
But I need to optimize my 7900xt
Optimise wym
I get little fps dips in forrnite when I enter an building
Make a fan curve that's all
Oh ok
Set an fps cap in fortnite and use dx12 pro
Dx12 pro?
Don't EVER use performance mode
On fortnite or amd software
Specific graphic setting in game
Fortnite
@sharp trellis
https://youtu.be/yP3dQ-fF-9U?si=COvkPJOIZB81I5H1
► (Screwdriver) Get 37% OFF on Fanttik E1 NEX Workstation:
► https://www.amazon.com/dp/B0CSJZFNY9?maas=maas_adg_5A707068EC1A3DB6DC85E3031D7D2B56_afap_abs&ref_=aa_maas&tag=maas
► (Apply the Coupon + Use code WORKSTATION)
► Buy my PC builds: https://zttbuilds.com
► Let me help you build a PC: https://zachstechturf.com/consulting
Today, we settl...
Oh thanks
Yes directx 12
There's a "pro" option
Directx11 also should work really well
Just don't touch perf mode, it woefully underutilises radeon
what option do i select?
nevermind, i didn't read. started benchmark
i have 32 gb but 22 was the closet, multi threaded
Interesting
Gotcha thanks
Theyre both brand new
Also, i wanted to ask what was bad about ventus?
Both sealed
Still would pick the proart
Ventus is like, ground floor at nvidia
Not even ground
The ground is where humans live
Ventus is the basement
It's got barely acceptable build quality
Yikws
They get better with that each gen but they still surface with issues
Ok. Yeah i just wanted some opinions on the deal.
Yeah ive been discouraged by friends though they never say why
O lmao
Actually i was wondering abiut pro art. I saw online theres a bunch of issues with how quiet it is
?
Or rather loud it is
Ventus is also loud
Welp no ventus then 😂😂
Lol
Pro art is good though right? Ive never had an asus gpu before so idk
Like they're both acceptable cards but in a like for like situation I know which one I take the chance with each time
I would give it a bit more credit tbh
I see. Alr. Thx for the second opinion. Gonna get the proart
What food u getting?
Last of the stress test is goin, everything passed so far. VT3 is left
It all passed, its running the tests again
So its not the ram if we've tested it 3 times. Probably the PSU or power delivery options if its the case
Hmmm
Left me extremely confused though as whea errors, and especially repeatable whea errors have everything to do with imc/ram just about
Maybe we need to do cycle 18 cpu extreme on occt
It's really unlikely to be vcore tho with c states disabled
Like, 99% unlikely
Lmao
Something is up, and it has everything to do with this but we haven't been able to replicate it
Just only the power OCCT is what failed. I did cpu on normal, I'll do extreme then
Maybe the windows install is buggered
A bios update or windows wipe might be the play here
The cpu isn't overheating is it?
No,no,no.
I can't afford to do that again.
I've done it so many times with y'all. I've even re-installed this year! Windows can't be the problem
No, again, in this small case its fine temps. The aio keeps it at low 70 for gpu and cpu during games
That's what my footnotes are saying here for whea
Even so, those problems didn't go away before. It has to be some thing else
Cpu extreme test is starting
I've got here vcore
Which we did by disabling c states
So rules it out
Raise imc voltage (vsoc/vddg)
Test agesas
Cpu overheat
And windows corruption
Ever do ddu?
Yes
I said yes
Wheas are unrelated to gpu entirely
If occt cycle 18 doesn't work then it's either raising imc or testing agesas
If either of those don't work it has to be windows
Otherwise there's some serious hardware fault there we are unable to diagnose
Which makes no sense as these stress tests are harder than games
Try this for imc next, leave ram at 3200/1600
Then bios revisions
If the power test is the only thing that's failing, maybe try to see if underclocking GPU will get you to a point where you're stable
Might be some weird thing with pulling power making 12v rail droop to a point where it makes the CPU unstable
1 hr and 40 minutes in fortnite gaming at max settings
I'm suspecting the cpu cable yes
Maybe changing vsoc loadline calibration from default to normal in bios
I really feel like
After all that


