>UPGRADE & BUILD ADVICE
Post build list or current specs including MONITOR: https://pcpartpicker.com/
Provide specific use cases
State BUDGET and COUNTRY or you will NOT be helped

>CASE
mATX: AP201, Lian Li A3, O11 Air Mini, XT M3, CH260
ATX: XT PRO (ULTRA), AIR 903 Base/MAX, Lancool 207, Flux Pro, Y40, Meshify 3, 4000D FRAME
Dual Chamber: Y60/70, Vision (Compact), Antec C8

>CPU
Budget (inc. gaming): 7500F, 7600/X, 9600/X
Gaming: 9800X3D, 9700X
Workstation: 9950X, 9900X
Premium: 9950X3D
Intel: If deeply discounted (265K for value & reliability)

>COOLER
AIO: Thermalright Frozen Edge, Cooler Master Atmos, Liquid Freezer 3/Pro (unhandy mounting solution)
Dual tower: A620 PRO SE, Peerless Assassin 120 SE V3, NH-D15 G2
ITX/>42mm RAM: Mugen 6 Dual Fan, AXP120-X67

TIM: MX-6, Duronaut, PTM7950 (phase change)

>MOTHERBOARD
ASRock AM5 board issues still reported; update BIOS before CPU installation

>RAM
DDR5: 2x16GB or 2x24GB, 6000CL30 (AMD), 6400CL32 (Intel)
Workstation: 2x 32GB (budget), 2x48GB or 2x64GB (high-end)

>SSD (OS drive)
Budget: SN5000, SN7100
High-end: SN850X, 990 Pro (Windows)
Premium: SN8100
https://borecraft.com/

>GPU
Budget: Arc B580 (for new builds), used card
Midrange: RX 9060 XT 16GB, 5060Ti 16GB
High-end 1440p & Midrange 4K: RX 9070 XT
Workstation: RTX 5090, used 4090 or 3090 Ti

>PSU
Budget: Corsair CX650
PSU buying guide:
https://hwbusters.com/best_picks/best-atxv3-pcie5-ready-psus-picks-hardware-busters/

>MONITOR
1080p: KTC H24T09P
1440p: KTC H27T22, Gigabyte M27Q, ASUS XG27ACS, KTC M27T6 (miniLED)
4K: Gigabyte M27UA/M27UP, KTC M27P6 (miniLED), ASUS PG32UCDM (QD-OLED), PG32UCDP (WOLED)

>OS
Activate Windows @ >>106123743

>CASE FANS
Meta: Case with good stock PWM fans
Midrange: Arctic P12 Pro (loud @ higher RPM), P14/Max (5-pack)
High-end: Fractal Momentum, Noctua G2 (140mm or 120mm)

Prev: >>106173428

>News & Rumors
RTX 50 Super: might launch in Q4 '25, 24GB VRAM for 5070 Ti & 5080 Super, 18GB for 5070 Super.
ASUS WOLED with TrueBlack Glossy film launched (XG32UCWMG, XG32UCWG). Limited availability.

>Omissions
RTX 5070 Ti: buy if you need CUDA, if you value DLSS/Reflex that highly, and if the 5060 Ti 16GB or 5070 isn't performant enough.
RTX 5070: buy if it's reasonably priced & if you plan to work around potential VRAM bottlenecks.
RX 9070 non-XT: worth considering if it's significantly cheaper than the 9070 XT.

Intel 13 & 14th gen i7/9 -K CPUs: BIOS update required due to reliability issues. Recommended to set a CPU VR voltage limit (1.4 or 1.45V).

>Product Notice
Gigabyte RTX 50 & RX 9000 video cards: thermal gel has QC issues, Gigabyte claims only early production batches were affected.
https://www.igorslab.de/en/update-on-thermal-putty-on-gigabyte-graphics-cards-after-further-storage-oil-separation-to-an-alarming-extent-and-silence-as-a-strategy/

Avoid updating Gigabyte motherboard's RGB controller firmware, as it can disable board features such as Q-Flash Plus.

>PSU Reviews
https://www.tomshardware.com/pc-components/power-supplies
https://www.lttlabs.com/
https://www.kitguru.net/reviews/?category_name=power-supplies
https://www.techpowerup.com/review/?category=Power+Supplies

For tier-list enjoyers:
https://docs.google.com/spreadsheets/d/1akCHL7Vhzk_EhrpIGkz8zTEvYfLDcaSpZRB6Xt6JWkc/

>Storage Reviews
https://www.phoronix.com/reviews/Storage (Linux performance)
https://www.tweaktown.com/cat/storage/index.html
https://www.techpowerup.com/review/?category=SSD
https://www.tomshardware.com/pc-components/storage/ssds/reviews
https://www.servethehome.com/category/storage/

Still no ray reconstruction alternative from AMD

>>106185846
don't worry they'll drop some shitty half baked terrible competitor in 3 years that wont be an actual viable alternative for another 4 years

>>106185846
still no black screen alternative from AMD

>>106185846
AMD Radeon's fully on-board with the AI & RT stuff, their executives are very excited to shill FSR4 and their frame generation in the future.
https://www.pcworld.com/article/2569453/qa-amd-execs-explain-ces-gpu-snub-future-strategy-and-more.html
https://www.tomsguide.com/gaming/playstation/sonys-mark-cerny-explains-why-your-ps5-pro-is-getting-an-upgrade-in-2026

>>106185846
they had that (shit) demo for path tracing at 9070xt launch
now its truly terrible image quality but it shows the trajectory where they likely get it done by the time next gen consoles launch

nicotroon ready to immediately drop some hot fresh cope

Objective and vehicle way points have a weird mirror image-ghost thing going on. Don't really notice any other issues with x4 framegen in this game though
This is 540p internal
>>106185874
>by the time next gen consoles launch
Yeah but considering amd's record that means it'll be an rdna5 exclusive
>>106185863
real
>>106185867
meme

>>106185873
We're in the second half, where's Redditstone? Do they mean "latter half of December 2025"?

>>106186063
stop noticing things goy

Fractal cases are unironically the best shit ever, change my mind.

>>106186170
I like how they combine good looks with good performance and ease of building in and good cable management with decent support and ability to buy replacement parts

>>106186063
and that would still be alright since 31 December is still part of H2

>>106185883
All these posts your keep schizoing out about aren't even me silly boy, the 5070 ti driver issues spam isn't me either btw. Still living rent free inside your head regardless it looks like doe.

>>106185979
>4 times the fake frames
>540p resolution
Nvidiots really playing their games like this lmao. Can't see any of the artifacts cuz they are blind I guess.

>>106186304
post more nico

How much vram does my gpu need for any game at 1440p max settings? Is a 5070 ti enough?

>>106186334
5070ti should be good for 144hz if you use dlss quality and turn pathtracing off

>>106186334
16gb

>>106186313
Of course.
Smart Access Memory>>>>Novideo rebar

>>106186334
the 24gb versions launch in few months

>>106186170
I love my North, would love to build a Terra for something some day. Maybe a media server.

Which of these back panels is better?

>>106186510
I need my speaker port. So the right

>>106186170
I wish they had a jade green matx case

Why do these cables have two sets of heads if you aren't supposed to power a GPU with a single cable?

>>106186510
>Dp
>More usb A
Left is the better back panel but it lacks clear CMOS, 5G networking and spdif

Couple of questions for AM5/DDR5, I've manage to tighten up my RAM timings a bit after flashing to the latest BIOS (previous one would crash almost every time I made any changes), what do you guys recommend for benchmarking CPU & RAM OC and testing stability? Preferably free shit, AIDA64 is trial only and kinda sucks, but at least I can tell my RAM latency has dropped from 72 ns to 63 ns, which is a nice improvement. Also is it worth trying to increase the FSB clock? CPU is a 7900X fwiw.

Lastly, is it worth trying to squeeze more out of my RX 6800? I'm running it with a 950 ma undervolt, 2333 MHz boost clock, fast RAM timings, +10% power, seems stable enough.

Odds that and makes a 5090 competitor under $1500msrp?

>>106186585
don't try to fuck with clocks unless you have a board with a clock gen

>>106186602
0%

>>106186602
They still don't have a 4090 competitor so good luck with that

>>106186602
All AMD needs to do is double the vram on their current cards and leave everything else unchanged, and there would be zero reason to buy a nvidia card for the next half decade.

But they refuse.

>>106186620
Yes extra vram for running the massive 4k market segment that they can't compete in

>>106186441
I already have one I got at msrp. I'll probably hold until it doesn't work anymore for what I want

holy shit I just saw this. I knew it was this bad, but I didn't know it was THIS bad

>5080
>9800x3d
with MSI X870 GAMING PLUS WIFI ATX AM5 board. Motherboard thoughts? I just like the dragon slapped on it. But if theres a board thats better suited for the combination, then I'll just toss it. I'll just paint a damn dragon on it.

Vram doesn't matter much if you're still slower. That being said give us more damn vram

>>106186675
What is this measuring? Good boy points? I swear you all cherry pick random benchmark data from one game and preach it as gospel without even showing what you're measuring

>>106186620
7900 XTX has more VRAM than the 9070 XT, why don't you buy that?

>>106186710
I just roll my eyes whenever that site is posted
you don't need to skew the numbers by including dlss to make amd look bad

>bought intel and 3dfx
>bought intel and nvidia
simple as

>>106185846
>fake resolution
>fake frames
>fake rays
I LOVE NVIDIA

>>106186558
>clear CMOS
Just plug your front reset switch in the clear CMOS header?

>>106186510
I would buy the one with clear CMOS
>>106186789
Yeah but the other is much more convenient.

What happens if I piss on the CPU

>>106186646
My guy, nvidia has gone three generations without a meaningful vram upgrade. There's zero reason why their budget cards are still on a mere 8GB, outside of the fact that they know you'll buy it anyways.

The 9060 is 8 fucking GB. If it started at 16GB and the 90XX's top end went up to 48GB, there would actually be reasons to buy team red at every single price point that aren't "I'm poor".

>>106186585
A day of y-cruncher and a day of prime95 should make sure that it is stable.
Then run a GPU gaming test for an hour or two to make sure heat isn't making it unstable.
>Also is it worth trying to increase the FSB clock?
Don't fuck with that unless it is de-linked external clock gen.
>Lastly, is it worth trying to squeeze more out of my RX 6800? I'm running it with a 950 ma undervolt, 2333 MHz boost clock, fast RAM timings, +10% power, seems stable enough.
This is probably more or less as good as it gets realistically.

>>106186853
It would get wet I think

>>106186853
dad gets very upset

>>106186868
because they can get away with it by being so far ahead
geebees dont matter if you don't have the compute to back it up

>>106186689
Dunno about that or X870s. My B650 MSI Gaming Plus Wifi has been working fine for me last 18 months though.

>>106186689
I have the 870 tomahawk and it's great. I'm sure that one is as well

>>106186853
Might cool better than mayo, give it a shot and report back.

>>106186878
>because they can get away with it by being so far ahead
Which is exactly why focusing on numbers would be a winning strategy.
If you want more than 16GB on a green card, you either have to pay close to $2k or get a card that's old enough to fuck.
Releasing a 32GB+ card that competes with nvidia's budget lineup just dominates the market. Why wouldn't you buy that over a 5060, or even a 5070?

>>106186975
So the entire strategy is just to have a higher number? Why not just advertise some arbitrary thing like number of capacitors, see how well that works out when the performance is still shit

>>106187010
I mean, they already do exactly that.

>>106186868
VRAM is like regular RAM - you either have enough of it, or you don't. 16 GB is fine, 24 GB is plenty enough, anything more is a waste unless you have the raw performance to back it up. AMD needs to put a fucking herculean effort into their next gen GPU core if they want to throw punches at nVidia and claw back their market segment, and throwing more VRAM and the problem simply isn't the answer.

Is an 850 mobo a worthwhile upgrade from a 650?Is there any benefit outside of the 850 having pcie 5.0 which is just a meme anyway?

>>106187327
no

>>106187267
Just one example, the 3090 is still the go-to budget card for AI exclusively because it has 24gb of vram, even though there's multiple cheaper 16gb cards that are more powerful all around.
And we're long past the point where AMD sucks for consumer AI, you can even run it comparably on Windows nowadays.

>>106187342
>AI
completely different use-case from gaming, they should be selling dedicated consumer AI cards at this point. The majority of people who game don't care about AI slop.

>>106187365
they already do that but part of why cuda is everywhere is because the software scales from weak ass gt730s all the way to multiple gb300 dgx systems

>>106186934
piss is warm doe

>>106186604
>>106186870
>no touchy FSB
What gives easy/linear performance gains, then?
And what benchmarks should I be using to verify my shit is actually running faster?

so done with these fucking driver issues on my 5070 ti. i was in the middle of typing up a shitpost dunking on the AMDrones ITT when my screen went black and I had to turn my whole pc off. im rmaing and switching to 9070 xt fuck it idc anymore.

>>106187414
Most of the modern hardware comes with 90% possible performance out of the box (except in heavily battery or thermal limited conditions)
The answer is that there is not that much to do in terms of performance gain anymore, excluding inefficient meme OCs.
Just undervolt the CPU as much as you can while remaining stable and let its algorithm decide how to push its clocks.
>And what benchmarks should I be using to verify my shit is actually running faster?
On Linux there is phoronix test suit but on Windows you can do a combination of geekbench, ffmpeg encode job, 7zip file compression job, cinebench etc.

>>106185756 (OP)
You won't use your overpriced hardware. You spend money on it because you are a no-lifer.

>>106186556
this cable is for cpu not gpu

>>106186510
The one that doesn't have the weird wifi antenna connectors

>>106186556
Realistically you could run an 8-pin + 6-pin GPU off a single cable and be safe, or even the full 8+8 if your PSU can handle it on that single line, but it's recommended for load balancing reasons to use separate power cables if available, particularly if you're running a fairly power intensive GPU.

>>106187505
Same exact shit just happened to me on my 5070, these drivers are truly heinous to the point I can't even meme about amjeets anymore

>>106187505
just happened on my 5090
i'm swapping to a 9070xt first thing tomorrow

>>106186675
Anon, why do you do this on daily basis 24/7/365?
It looks pretty weird and suspicious as well.

>>106186689
>buying a $1500 16gb card that will be replaced by a 24gb version in 5 months

>>106186620
The 9080xt is going to have 32gb. But it doesn't matter because stable diffusion sucks on amd and barely works

>>106186646
I've been playing in 4k since the 760gtx 2gb. I will never understand this meme

>>106186510
Personally I would chose one without wifi and with more full size pcie slots. This is becoming increasingly difficult.

>>106185979
There will be no RDNA5 it's called UDNA now also zen6 is coming with dual x3d cpu as well

>>106187505
>Im a gay faggot who sucks huge dicks and has to shill in every PCB
Have you thought about just ending it?

>>106187561
Anon, why do you do this on daily basis 24/7/365?
It looks pretty weird and suspicious as well.
>>106187505
>>106187537
>>106187543

Suggestions on an mATX mobo for a 9800?
Do they even still make mATX?

>>106186585
>what do you guys recommend for benchmarking CPU & RAM OC and testing stability?
for RAM stability:
https://github.com/CoolCmd/TestMem5 - use ABSOLUT or anta777 extreme profile, 1usmus is decent for quick runs while changing settings since it has a much shorter test cycle
https://www.ocbase.com - one hour free
https://www.karhusoftware.com/ramtest - paid
https://www.mersenne.org/download - run large FFTs for 30-60 mins
https://www.numberworld.org/y-cruncher - run VT3 for 30-60 mins
>AIDA64 is trial only and kinda sucks
AIDA64 SHA3 is a good stress test
>CPU is a 7900X fwiw
post a zentimings screenshot instead
those timings aren't looking good but i'm just gonna pretend AIDA is bugged
>>106187508
>Most of the modern hardware comes with 90% possible performance out of the box
lmao, no
but 20% of the effort does get you 80% of the results

>>106187670
>do they still make the most common (1st? 2nd?) form factor board?
Try googling AM5 motherboard sheet.

>>106187721
Most common? Almost everything I see online, whether it's on store pages or reviews, is full size ATX. If that's the case, do you have any recommendations on one?

>>106187736
There are tons of mATX cases out there because there are lots of people who don't need more than a few NVME slots and don't care to pay for more.

Google the sheet, link is blocked here. Pick a B850M that fits your needs. They're all the bloody same outside of buying from the lesser known brands. Technically speaking the few Chinese brands will work fine but the manual and maybe BIOS will be in Chinese.

Oh right, Asrock has problems with killing AMD CPUs. Might want to avoid them for AM5.

>>106187736
Full list:
https://pcpartpicker.com/products/motherboard/#f=7&s=41&sort=price&page=1

Filters for 9800X3D (it's a gaming rig so no X chipsets and no ASrock due to X3D issues):
https://pcpartpicker.com/products/motherboard/#f=7&sort=price&m=8,18,27&c=160,169

Price cap of $200 (again, gaming rig) leaves you with 23 boards:
https://pcpartpicker.com/products/motherboard/#f=7&sort=price&m=8,18,27&c=160,169&X=0,20000
Easily enough to sift through but you could probably get it under 10 with some basic filters. PCIe/M.2 slots, LAN, WiFi or not, etc.

Why don't we have a water cooling thread?

>>106187915
You're welcome to post one or talk about it here but not many people run custom loops these days. Hell, it wasn't that popular even back when it was noticeably better than air/AIO cooling.

>>106185846
What the fuck is up with that image? Dark, blurry mess, where the only sharp thing is the HUD. Muddy textures. Reflections and lighting that were achieved using raster techniques a decade ago.
>1920x1080
To achieve mediocre picture quality like that. An absolute joke.

>>106187658
Saar, 0 rupees were deposited on your account.

>>106187959
In the age of 500+ W gpus water cooling is a must imo.

>>106187977
I respect your opinion. Is that your rig in >>106187915? It's nice.

>>106187915
Anon, g/incels are dirty poor pieces of human trash mostly.

>>106187999
You mean kids?

AMD Radeon AI PRO R9700, also known as 9070xt with 32gb VRAM is supposed to be sold for 1200$.
Thoughts?

>>106186675
>be me
>dont use site for several months
>check back in out of curiosity
>its all the exact same posters posting the same stupid shit
Neither reddit nor twitter are as bad as this shithole, as they have at least baseline moderation possible.

>>106185756 (OP)
Why do nearly all cases have glass side panels now?

>>106188157
Too expensive. Shall use my 3070ti till it dies and then I shall use working 2080s.

9070XT at 720€
5070Ti at 890€
Worth the extra 170€?

>>106188325
sure

>>106188325
please consult the chart

>>106188325
Yes, coming from a 9070XT owner that is starting to regret my card because of all the heat problems

>>106188339
Do you have any proofes, saar?

>>106186525
L-Out is your speaker port dipshit they both have it wtf are you talking about

>>106187697
>lmao, no
That's just retarded to say. Show me how your GPU OC profile turns 100 FPS into 130 otherwise.
Even memory profiles, where most headroom is left, will result in single digit uplift in most games with respect to just enabling EXPO. (Though can have more impact on 1% lows)
In most extreme cases, you might be able to squeeze up to 30% improvements if you can get your memory to run at high speeds for productivity tasks that scale with bandwidth. That's about it, and not relevant to most people.
>inb4 meme synthetics

>>106187697
>post a zentimings screenshot instead
Here. I was following a few different guides earlier but my BIOS kept crashing on save, did an update to the latest beta version and it seems a lot more stable but I've only done some conservative improvements so far. AIDA64 says it's dropped latency from 73 to 64.5 ns so I'm not sure there's much more to squeeze out of it compared to the stock EXPO settings.

how trusworthy are the rumours about the 5000 series super refresh thingy?

>>106188600
extremely
dont buy a gpu before christmas

>>106188606
unirnically?
fuck me

>>106188456
s/pdif

>>106188606
Watch the 5080S have just 18GB VRAM for 1899 MSRP aka scalping price 2300 lmao

>>106188611
kopite7kimi has never been wrong about NVIDIA GPU specs, he even got the 5090 2 slot cooler right when everyone thought it was totally bogus

>>106188621
the super swaps 2gb gddr7 to 3gb modules
8 modules of 3gb = 24gb

>>106188632
well damn
>>106188635
any chance this is gonna slow down the card somehow?

>>106188525
first impression is that tRAS+tRC is loose, tRRDL is kinda loose, tWTRL is loose, tRTP is loose, tWRWRSCL is ridiculously loose
tRRDS+tFAW is good, tWR is good, tRDRDSCL is good, tRFC+tREFI is good (no point in pushing these to the limit, you're only setting yourself up for random idle crashes)

power down mode should be disabled
you can push higher voltage on hynix for lower primaries (VDD)
you can also try to push to 6200-6400 MHz with your 1.3V SOC, should give you a nanosecond or two of latency savings since it overclocks the memory controller

>AIDA64 says it's dropped latency from 73 to 64.5 ns so I'm not sure there's much more to squeeze out of it compared to the stock EXPO settings
you should drop down to 60-62 ns just by disabling power down mode

these three videos have good AM5 info:
https://www.youtube.com/watch?v=dlYxmRcdLVw
https://www.youtube.com/watch?v=BS_NeTwjOvY
https://www.youtube.com/watch?v=Xcn_nvWGj7U

Will there be a 5060 Super with more ram then 5060ti?

>>106188657
wouldnt make sense in the stack where a more powerful 5070 super exists with 18gb

>>106188638
There's a risk of the 3gb modules having looser timings and losing performance but the supers will be factory OC'd and pull ahead of the non supers at stock.

Buy a 5080 now, the Supers will most definitely run hotter

>>106187977
not really since you can just underpower to 400W and get literally the exact same performance at severely reduced temperatures

>>106188681
but 24 gb of vram to render my furry porn

>want to undervolt my 5080
>completely forgotten how to to it
is there a undervolt for retards video out in the wild?

>>106188689
download msi afterburner
either adjust the voltage from the curve editor or just limit the watts from the power limit

>>106188709
Powerlimiting is NOT undervolting....

>>106188716
That's why I said ///OR//// while listing two different things, you illiterate retard.

>>106187697
1usmus also has this sheet that gives a rough idea of what's erroring. Though I'm not sure if it's relevant for DDR5.
https://docs.google.com/spreadsheets/d/1MTKvkEI5mXchE_XFaiG9AAS89sOANAy7YY69bWLkdQ8/edit?gid=221662466#gid=221662466

>>106188157
I'm not sure who is going to purchase it.

>>106188735
these have never been accurate for me
in general i would be cautious of anything you find on overclocking forums, there's too many schizos on those who find patterns in data that aren't real

>>106188642
Thanks anon, I'll have another crack at adjusting stuff tomorrow. Running y-cruncher for an hour at the moment to make sure it's stable before going any further.

>>106188689
how much did you pay for your 5080?

>>106188681
Saar, I do not need a new gpu.

>>106188799
>Running y-cruncher for an hour at the moment to make sure it's stable before going any further
that's a sensible approach
honestly the only thing that REALLY sticks out is the tWRWRSCL
i assume that's on auto?

everything else is fine, way better than EXPO
no classic beginner mistakes like tRRDS>8, tFAW>32, tRDRDSC/tWRWRSC>1
those are the most important to get right

good luck anon

>>106188841
>i assume that's on auto?
It would be, yeah. I haven't touched the sub timings, just the tRAS, tRFCs and tREFI for now. Everything else is EXPO defaults.

>>106188960
just copy what buildzoid does in his easy timings with the exception of tRAS and tRFC/tREFI
leave that as is

the games that will scale the most with RAM OC are simulation games, RTS and MMOs
also large open world games like KCD2, MH wilds, modded skyrim, etc
and terribly coded unity indie games lmao

you can use these to see how much your score is affected by tightened timings:
factorio benchmark - https://pastebin.com/1BKuUtBS
FFXIV dawntrail benchmark - https://na.finalfantasyxiv.com/benchmark
MH wilds benchmark (use low graphics settings) - https://store.steampowered.com/app/2246340/Monster_Hunter_Wilds

for synthetics geekbench will scale with bandwidth and something like pyprime 2B will scale with latency
i wouldn't care about getting low scores in these though

>>106188816
$1649

Is it smart to buy a new PC now?

With all the tariffs going on nowadays.

Is new tech close to being released? Is Windows 11 unoptimized invasive garbage (wait for Win12)?

>>106185756 (OP)
I dont get it
So i could just buy random ddr5 ram and set it to 6000mhz cl30??

>>106189044
You don't need fresh hardware with CIA's back-doors.

>>106188725
He asked about undervolting, learn to read

>>106188816
i got it for 1.2k
>>106188709
i meant as in how to edit the curve
i completely forgot how to do this shit

>>106188835
SAAR YOU WILL BUY

>>106189070
It doesn't matter if you undervolt or power limit, power limiting will automatically limit the voltage, dumbass.
>>106189076
You find out what your max clockspeed is that your gpu is running at, you consult the graph, you adjust the node at your max clockspeed to your desired voltage as shown horizontally, then you drag and select all the nodes to the right of the max volt node, drag them down, then go back to the msi main window and apply settings. Of course it's easier to just power limit.

Instead of discussing the actual topic, you inbred mongoloids are constantly baiting and starting flame wars. And these nigger faggot jannies allow it. You should get gunned down like dogs and burned alive, retarded troons. May your deaths be slow and painful.

>>106189095
>power limiting will automatically limit the voltage
You are so fucking stupid jfc

>>106189161
>use voltage to limit power
vs
>use power to limit voltage
there is no difference, dumb fuck and the fact that you couldn't even procure an argument to counter this simple fact already says enough about you.

>>106189049
No you can't

>>106189177
>there is no difference,
Power limiting and undervolting are two different things. Undervolting can reduce power consumption so the CPU can run faster. For many games that are GPU limited, a faster CPU is not necessary. In this situation, limiting the CPU power and forcing the CPU to throttle might not cause any significant decrease in performance.

>>106189240
except power limiting also reduces the voltage, it's literally the inverse of undervolting and yields the exact same results

>>106189272
No it does not

>>106189322
Yes it does, retard, you can literally check yourself using hwinfo

>>106189334
Here, even a retard like you should be able to understand this

>>106189355
Literally second line in your image says that power limiting will automatically lower the voltage and/or clock speed, exactly what I said, you absolute buffoon.

>>106189272
>except power limiting also reduces the voltage
NTA, but this isn't necessarily true.
Voltage will be reduced only if your power limit is set too strictly, and you will lose performance compared to an OC + UV.
Increasing the GPU core offset slider + setting a power limit would achieve similar results to an OC + UV.
Setting a lower voltage limit would probably make the card run more power efficiently.
Test it yourself with a single benchmark.

>>106188689
Get the beta version of MSI Afterburner to unlock memory OCing past 2000MHz for RTX 50 cards.
https://forums.guru3d.com/threads/msi-afterburner-4-6-6-beta-6-with-unofficial-rdna4-support.456792/

If you want a quick and easy guide without fiddling with the V/F curve, set the GPU "Core (MHz) Curve" offset slider to anywhere between +200~250MHz, you can go higher with RTX 50 but this is a good starting pt.
Then set a power limit to 75~85% (270~306W from stock 360W).

If you want to use the curve editor, refer to the guide below.
But don't follow their frequency offsets or voltage limit, they are using a 3070 as a reference.
RTX 50 should handle the 200~250MHz GPU core OC I mentioned earlier, and their performance/efficiency voltage sweet spot is between 925~975mV.
https://github.com/LunarPSD/NvidiaOverclocking/blob/main/Nvidia%20Overclocking.md
https://www.youtube.com/watch?v=LPpW9yXHvOU

Don't follow most TechTubers' undervolt guides; they are usually downclocking the V/F points to the left of their voltage limits.

https://nvidia.custhelp.com/app/answers/detail/a_id/5676/

After 11 years of Game Ready Driver support for Maxwell-series GPUs, and 9 years of support for Pascal and Volta-series GPUs, these GPUs will only receive driver updates for critical security items beginning October 2025 and continuing through October 2028. Game Ready Driver updates, including performance enhancements, new features, and bug fixes, will be available for systems utilizing Turing, Ampere, Ada, and Blackwell-series GPUs.

PASCALFAGS STATUS?

Will the 5080 super have more juice? The performance is kinda ass too, not just the 16gb of vram.

>>106189394
>you will lose performance compared to an OC + UV.
No, you won't, I literally just tested this
power limiting to 400W yields 1.040-.050V, 2780-2900MHz, 70C temperature - 48.5s/it
undervolting to 1.040V yields - 400W, 2900MHz, 73-75C temperature (due to always running at max clock) - 48.2s/it
You aren't losing shit, the 0.00001% difference is absolutely negligible in every scenario unless you're some autistic number-chasing tinkertranny, it is completely unnecessary to ever touch the voltage graph when you can just adjust the slider to 90% and get the exact same results with no real difference in performance

>>106189423
>Will the 5080 super have more juice
5080 Super is using the same, fully-enabled GB203 chip.
It will have higher factory base & boost clocks, and a higher maximum power limit.
Comparing manually OC'd 5080 & Super, the Super might be slightly faster (around 3~5%) because of the higher power limit.

>>106189431
>undervolting to 1.040V
Is this an NVIDIA card? because that's nothing.
A stock 5090 FE draws 1.02V on average during gaming loads.
Set a proper voltage limit, like 975/950/925mV.
https://www.techpowerup.com/review/nvidia-geforce-rtx-5090-founders-edition/44.html

V/F analysis of stock 5090 FE in pic.

>>106189497
my 5090 draws 1.08v at the full 575W, which is completely unnecessary, but it doesn't matter if I drop the voltage or the watts, the performance is nearly identical, except undervolting doesn't limit the clock speed so you actually get higher temperatures

>>106189519
What matters is the GPU "Core (MHZ)" offset slider, whichever way you decide to reduce the power consumption of the card.
If you don't raise that offset, you're leaving performance on the table either way.
That offset helps you retain the stock performance you paid for, even with lower power consumption.

>>106189544
the 100MHz drop won't affect performance in any meaningful or noticeable way, though, maybe if you're running a GTX760 but then you should be thinking about buying a better card instead of playing with clocks and voltages

>>106189591
>100MHz drop
Except we're talking about RTX 50 GPUs that have at least 200~300MHz overclocking headroom across the stack.
It's not uncommon to see people reporting stable 5080 OCs running at 3.2GHz.

I really need to upgrade my 9 year old machine but I don't want to spend so much fucking money...
2000€...Is that how much you have to pay nowdays for a medicore pc?

https://de.pcpartpicker.com/list/hFJYGJ

anyway so save money her without losing quality?
big psu because I'm gonna upgrade to a 5080s when that comes.
(which will be another 1400€...)

>>106189646
you are massively overspending on everything that isn't your GPU, you fucking retard

>>106189613
And it's completely unnecessary and won't offer any real tangible benefits when your card is already fast enough to do everything in minutes or seconds, this isn't 2014 anymore, overclocking is a dead scene unless you're just a number-chaser

>>106189646
>mediocre
>top of the line gaming cpu
>paired with a 5060ti
>1000w
>128GB

>>106189646
this has to be a troll

>>106189666
>completely unnecessary
Over 10% performance boost from OC puts the 5080 much closer to a stock 4090's performance.

>you're just a number-chaser
People buy these kinds of video cards because they're chasing higher numbers (fps & lows) @ higher visual settings.

>>106189657
where can I save money?
>>106189673
how is a 5060ti not medicore?
>>106189691
unfortunately not

>>106189703
>110 vs 100 fps
literally noone cares, it's like running your kettle at 400W vs 500W, will your water boil 10 seconds faster? yes. Will you ever notice it? no.

>>106189751
>literally noone cares
Yet people ask how to OC + UV their cards in these threads.
You just have to move three sliders in Afterburner (GPU core offset, memory offset, & power limit) to get that 8~10% extra performance.

>>106189743
you don't need a 9800X3D
you can get a cheaper mobo
cheaper and less RAM
far cheaper SSD
cheaper PSU

anyone knows if there are benchmarks for video rendering, like davinci resolve, etc?

>inb4 just buy nvidia
no

people thought this was hitstop but this was actually hardcore nvidia driver stutter

>>106189876
PugetSystems
https://www.pugetsystems.com/labs/articles/amd-radeon-rx-9070-xt-content-creation-review/

>thought I'd get cheaper used parts and reuse some of mine
>can game on RT Ultra whole month earlier!
>Gigabyte B660 mobo cannot hold RAM in A1 B1 for shit
I mean it is fair that it even warns on boot to put sticks into A2 B2, but who knew it meant nothing fucking works otherwise?

I am now forced either to buy 2x16GB DDR4 in fucking 2025, grovel with 2x8GB DDR4 in 2025, or just give up, wait for paycheck and buy DDR5 mobo with new sticks.

>>106189795
Maybe in benchmarking software, sure, but the extra 5fps or 5 seconds you gain anywhere else that you actually use your gpu for will not be noticeable at all.

>>106189890
if anything overclocking causes more problems like >>106189883

>>106189886
what the fuck are you even saying

>>106189890
Very strange how no one talks like this when the conversation is about OCing Radeon card.
>>106181797
Is someone bitter that others are also having some fun?

>the extra 5fps or 5 seconds you gain anywhere else that you actually use your gpu for will not be noticeable at all.
Wrong.

>>106189925
That A1 B1 RAM slots no worky on used mobo that I bought.

>>106189962
no it sounds like you're trying to mix dimms
and what does that have to do with rt
try running in gear 2

>>106189955
>Wrong.
LOL
literally turn off your fps counter and you will never be able to tell if your game is running at 65 or 70 fps

>>106189993
Imagine being this bitter about OCing NVIDIA cards, having to resort to posting contextless clips of his favorite Chinkshit slop.

>>106190014
>H-HOLY SÖY, DID YOU SEE THAT, I JUST WENT FROM 210FPS TO 240FPS HOLY SHIT THIS IS AWESOME
>*uses a 60Hz monitor*

>>106190034
Bitter loser writing up fiction as a coping mechanism.

>>106190050
What fucking fiction, you retard, nobody besides hardcore enthusiasts are running 200+Hz monitors, idiot, you're literally shitting up your GPU and wasting power rendering those extra frames that you don't need instead of frame-limiting to your monitor's frequency, so what is the point of your shitty overclock again????

ordered nvidia gpu only because it can do AI gooner things better
I don't even care about ray rracing

>>106190084
240hz can be found for under $300 these days

>>106189991
>no it sounds like you're trying to mix dimms
Perhaps I wasn't clear enough.

On a Gigabyte B660 motherboard, 2*8GB DDR4 sticks from the same kit:
>A1 B1
Do not work.
>A2 B2
Work perfectly fine at their max profile. I have two kits so I tested both: they both work in A2 B2 and both don't work in A1 B1.

On my older ASUS Z270 motherboard, 4*8GB DDR4 sticks from 2 different kits work perfectly fine, picrelated. Cursory internet search suggests bent pinks on motherboard which I guess is possible but I examined it during purchase, the pins were fine.

>>106190101
Maybe some 1080p chinkshit, sure, but it's a complete waste of money if all you're going to use it for is to play minecraft and cs on your blurry-ass screen

>>106190084
5080 owners are very likely to have at least a 165~240Hz monitor.

>shitting up your GPU
It's running within the vBIOS & NVAPI's limitations, NVIDIA App's Auto OC doesn't break warranty either.

>wasting power
5080 is one of the most power-efficient cards.

>>106190084
>those extra frames that you don't need
What is system latency?
Just stick to your shitposts. You're hopeless.

>>106189991
>try running in gear 2
Also this is what I tried:
>2133mhz (JDEC) 15-15-15-36 tCMD 2, DRAM V 1.25, VCCSA 1.25, VDDQ 1.25, MEM FAST BOOT DISABLED, GEAR 2

>>106190112
Why are you trying to set two memory modules in the suboptimal A1/B1 config to begin with?

asrock or asus mobo?

>>106190433
ASRock, if you trust their claims that the new AM5 BIOS has solved the issue.
But I wouldn't fuck around and find out.
https://old.reddit.com/r/ASRock/

>>106190400
Because I am trying to run 4*8GB DDR4 sticks. I was not aware this is such a problem in 2025. My 2017 motherboard runs them just fine.

>>106190456
>6 hours ago
yikes... I might skip that one

>>106189886
>>106190491
gigabyte is ass for DDR4 on LGA1700
sorry you had to find out this way

Do I put ram in a1 and b1 or a2 and b2?

how is a Ryzen 7 5700X3D looking like for gaming nowadays?
I'm looking through a used PC with that and a 4060 Ti for USD 1,300.00 which is pretty decent price for my country (Huezil).

>>106190433
asus if want the ""gamer"" look
asrock if you want a ticking time bomb

>>106185756 (OP)
https://pcpartpicker.com/list/n7pdQd
>Use cases
Gaming, mostly indie/AA games, web browsing, multiple videos.
>Budget
$1300 and Perú.

Any way to save money on this build? I really want to keep the CPU and the mobo since i want this build to last ma at least 5-8 years.

>>106188337
Important to note that neither card can max games even today so you're going to need the better AI shit.

>>106190639
it's still good for gaemz

>>106190675
you should get a 2tb nvme
5060 at a minimum
KTC M27T6 would have gooder contrast and hdr but worse pixel response

>>106190730
Thank you, I will look into it.

>>106190675
no need for that giant cpu cooler for a 9600x unless your room is a literal sauna during summer

>>106190639
5700X3D performs about as well as a Ryzen 7600X on average, and better in CPU-demanding games.

>>106190675
The video card isn't strong enough for that monitor, if we're talking AA games running on UE5.
>Any way to save money on this build?
You can drop down to a 7500F or 7600/X if you don't need the better AVX-512 performance on Ryzen 9000 CPUs. AVX-512 improves PS3 emulation performance.
Drop down to an mATX motherboard if you need to cut costs.

Is 970 Sneedvo a value option for you? SN580, NV1, MP44 (all DRAMless) are budget alternatives.

>>106190675
https://pcpartpicker.com/list/YGhsNz
a minimum GPU is the 9060XT 16gb if you want it to last 5 years, stupid mountain ape

>>106190728
>>106190779
thank you anons

Threadly recitation of the Radeonman's Creed #3
>This is my Radeon. There are many like it, but this one is mine.
>My Radeon is my best friend. It is my life. I must master it as I must master my life.
>My Radeon, without me, is useless. Without my Radeon, I am useless. I must OC my Radeon high. I must score higher than the nvidiot who is trying to outscore me. I must burn his connector before he burns me. I will...
>My Radeon and I know that what counts in this war is not the benchmarks we run, the noise of our fans, nor the smoke we make. We know that it is the score that counts. We will score...
>My Radeon is human, even as I, because it is my life. Thus, I will learn it as a brother. I will learn its weaknesses, its strength, its parts, its accessories, its voltages and its clocks. I will keep my Radeon clean and ready, even as I am clean and ready. We will become part of each other. We will...
>Before Lisa, I swear this creed. My Radeon and I are the defenders of AMD. We are the masters of our enemy. We are the saviors of my life.
>So be it, until victory is AMD's and there is no enemy, but peace!

so everyone just gave up right? nvidia driver issues still not fixed on windows or linux.

>>106190822
my 5090 works flawlessly

>>106190834
no it doesn't.
https://nvidia.custhelp.com/app/answers/detail/a_id/5157/~/after-updating-to-nvidia-game-ready-driver-461.09-or-newer%2C-some-desktop-apps

>>106190617
A2/B2 is practically the standard recommendation.

>>106190834

>>106190845
>>106190855
coping troon trying to tell me to not believe my lying eyes lmfao

>>106190845
I'm sure NVIDIA drivers are the reason why Radeon 9060 & 9070 XTs are experiencing issues related to Microsoft's Windows MPO.
Get help.

>>106190845
is this AMDesperation?

>>106190878
still not fixed as of 2025

>>106190870
>>106189883

>>106188325
Wait until Christmas.

>>106190864
you mean the same eyes that thought 1707x960 is better than native?

>>106190883
you are a desperate coping lunatic troon

>>106190902
show me a single driver from nvidia with patch notes that say the problem is fixed.

you can't.

checkmate autistic faggot

worth waiting for the 9700x3d for when i want to finally upgrade to am5?

>>106190902
Nvidia is literally broken on Linux so by extension its broken on Windows. It's not even up for debate. It's how GIT works which Microsoft also owns.

>>106190911
insane frothing at the mouth coping troon with RDNA1 buyers remorse but too poor to upgrade

>>106190925
I won you lost because you can't show burden of proof

>>106190924
Microsoft owns github, not the concept of GIT, retarded tranny

>>106190890
Contextless webm. Why are you so obsessed with that game?
It's a poorly coded mobile port riddled with issues.

why is the 9800x3d so slow wtf? AMD sisters were we lied to again? This is bulldozer all over again

>>106190938
>>106190939
phone&desktop samefag
>inb4 inspect element

>>106190938
and if Microsoft owns github, you see how they deploy their code? It doesn't work. So neither does DirectX or your driver.

>>106190949
Show what it's fucking running

>>106190911
>>106190935
MPO isn't something NVIDIA or AMD can fix.
That is why that NVIDIA help page is providing Windows registry fixes until Microsoft resolves the issue.
Those are the same registry files used by Radeon card owners to deal with MPO issues.
You're too dumb & mentally ill to understand the scope of the MPO issue.

>>106190949
>>106190965

>>106190962
ridiculous seething insane coping lunatic tranny lmfao

>>106190939
>>106190974
>>106190925
>>106190834
nvidia still doesnt work as of todays date

>>106190939
The game literally works on AMD and it doesn't work on Nvidia.
>>106190855
>>106190855
>>106190855
Official game response

>>106190998
Is this true? I was thinking of getting a 5060ti 16gb but now I'm leaning towards 9060xt 16gb

>>106190983
Because Battlefield 6 is one of a handful of games that is coded to make the most out of the new Intel hybrid core CPUs.
https://community.intel.com/t5/Blogs/Tech-Innovation/Client/Intel-and-EA-Announce-PC-Optimization-Partnership-for/post/1706776

Expect free Battlefield 6 to be bundled with Arrow Lake or its refresh SKUs later this year.

>>106191018
That's the guy behind the biggest mods in PC gaming and he says Nvidia is broken. Who the fuck do you believe? rando retarded no names like >>106190974 or actual PC devs?>>106190855

>>106191018
No not really it only exists on reddit
>>106191029
Why do only unoptimized games run better on x3d? Seems like Intel gods who don't play slop win bigly

>>106190781
>https://pcpartpicker.com/list/YGhsNz
Thanks a lot!

>>106190779
Oh wow, those budget alternatives are solid, thanks!

>>106191018
Nvidia doesn't fucking work at all go for the AMD card. Buy based on todays reality

>amd retards are flopping like dying fish to meme driver issues
AMD had a chance to win me over with the R9 series, but failed miserably

>>106191048
>>106191018
Yeah its like still buying Intel in 2025. only retarded people and "tricksters" recommend it. you buy AMD no matter what.

>>106191019
Desperate coper linking stale & irrelevant article published in late April '25.
Clair Obscur has had over 300 bug fixes since then.
Not to mention the latest NVIDIA's driver is now at version 580.
https://www.techpowerup.com/review/clair-obscur-expedition-33-performance-benchmark/8.html

opinion on prime B850M-K for 9600x?
It has exactly as much as I need and nothing more

>>106191066
Then it's a good board for you.

>>106191057
TruthNUKE

>>106190949
Was this the source the commentators on jewfus' vid were using? were they right?
how come the jew man did not get these results?
14900k=9800x3d=9950x3d
https://www.youtube.com/watch?v=Wyubzqb-VuY

>5600MT/S
lol

what site is this?

>>106191029
Jewfus did not get these fps'
what went wrong

>>106191084
>>5600MT/S
>lol
I actually like this because I personally don't believe in XMP/EXPO

>>106191043
>Oh wow, those budget alternatives are solid, thanks!
They are DRAMless NVMe drives with a lower write endurance.
970 Evo has a DRAM cache.
You decide whether the savings are worth it.

https://phisonblog.com/dram-or-not-the-difference-between-dram-and-dram-less-ssds-and-why-it-matters/

>>106191084
Because that guy probably runs his 14900K with e-cores disabled.

>>106191052
Vega was good if you didn't have a blower card and overclocked at least.
>>106191018
Anon AMD shills always lie. Buying a radeon card will always lead to regret, they will abandon that 9060xt 2 years from now and it'll be worse off than a 5050.
>>106191084
I don't know who this is but he has a very Jewish phenotype and is 100% a liar and grifter. Game scales really well with core count, even the estrogen ones. An overclocked i5 is beating the 9950x3d at stock (no idea how well it would do overclocked)

>>106191084
nvm found it
https://www.pcgameshardware.de/Battlefield-6-Spiel-75270/Specials/Open-Beta-Release-Gameplay-Live-Benchmarks-Test-1479164/2/

>>106191120
that does not explain why the 9800x3d has half the min fps compared to 9950x3d though
jewfus got 170-180 fps lows in the testing
what went wrong

>>106191103
5600 is still running faster than defaults

>lg makes a 540hz 1440p monitor
>dp 1.4
jfc are the connectors made out of gold
why are they doing this

can I connect usb c 3.0 header to to usb c 3.1 on a front panel? are there physical differences?

>>106191141
>Vega was good if you didn't have a blower card and overclocked at least.
Wasn't Vega

>>106191143
Maybe it's offloading less latency-sensitive workload to the CCD effectively (including background tasks) while the CCX exclusively handles the latency-critical loads.

>>106191005
>it doesn't work on Nvidia.
Works fine, RT enabled as well. I don't resort to posting stale, contextless crap like you do.

>>106190949
>>106191084
>>106191143
BF6 gets HUGE speed bump on AMD 9800X3D by turning core tuning to level 2 instead of legacy in the bios. most reviewers leave it on legacy because it produces better aida64 results but its not the same for gaming.

>>106191229
Why are you googling random facebook posts when the devs literally the official game staff says it >>106190855
>>106190855
>>106190855
>>106190855
doesn't work on NVIDIA

>>106191057
Okay I'm buying 9060XT+5700X3D. Any other tips?

>>106191167
nevermind, i'm stupid
it's all 3.2, it's just different speeds
well, they didn't make it confusing at all

>>106191241
>when the devs
That's a boilerplate template for the Customer Service AI Chatbot to paste.

>>106191191
look at these cpus here, do you think this looks correct?
9900K=9800x3d is hilarious

>>106191231
I have no fucking clue what that is since I am just an avid shit poster and don't own these platforms but that sounds hilarious
are these defaults or not?
this is horrendous

>>106191143
>5600 is still running faster than defaults
It's maybe one speed bin faster depending on exact board and CPU
JEDEC spec DIMMs goes to 5600 even if it's not the default on overclocked kits, OC kitss for some reason always default to 4800 even if there is zero hard to adding 5200/5600 to the SPD

>>106191255
AMD recommends Level 2 Core Tuning for all gaming themselves.

>>106191267
pcgameshardware.de confirmed for retards

>>106191253
Holy fuck you're stupid. The staff of the game is more trustworthy than your NO NAME UNKNOWN STUPID ASS pulling random Facebook comments.

>>106191255
They're all 16-threaded CPUs, and BF6 is optimized to utilize as many available threads.
It can properly utilize Intel e-cores.
Not every game needs all that extra cache.
I'd be interested to see how well this game scales with memory OC.

>>106191279
Most reviewers operate on old fucking standards.

https://www.reddit.com/r/Amd/comments/1h8siwi/asus_intros_core_tuning_config_for_gaming_feature/

>>106191280
>staff of the game
Customer service agents are usually outsourced to India or SEA.
They don't have the latest information about the state of the game or the stability of the latest GPU drivers.

AMD sisters it's officially over, we're slower in the game of the century. I literally cannot stop shaking and crying, i'm doing it so hard ive puked on myself like 4 times in the last hour oh god nonononono

>>106191287
ok so you think a 9900k should have the same minimums and average fps as a 9800x3d in 2025
very cool
it's the whole games scale with cores meme again

>>106191293
techtubers confirmed for retards + not playing games
it's over

>>106185756 (OP)
Can we get more ssd /pcbg/ recommendations? Or western digital is the only best choice here?

>>106191319
They could be prioritizing Intel CPU optimizations at this stage of the beta, and the thread scheduling carried over to 16-core Ryzen CPUs.

>>106191319
>ok so you think a 9900k should have the same minimums and average fps as a 9800x3d in 2025
NTA but this could just be the harsh reality that CPU matters less than people have convinced themselves.

>>106191353
Western Digital is the best in all segments. Check the SSD review sources linked just below the OP.
Their performance doesn't degrade in Linux, which can happen on Samsung & Crucial SSDs.
Hynix had firmware-related issues that was recently addressed, but their options aren't better than WD or the 990 Pro (for Windows)
People shouldn't sleep on the SN5000 & SN7100 if they're planning a more value-oriented build.

>>106191319
Is this a 4k test? Why is the fps so incredibly low across the board?

>>106190949
corelets btfo

>>106190949
>7950x3d -> 9950x3d is more than double
hmm

>>106190554
>>106189886
>>106190491
So I did some more tests.
A1 and B1 work separately, when single stick of RAM put into either of them.
But they do not work in dual channel, when both are occupied. Wild.

I'm going to take cooler and CPU off to see if I fucked any pins up or smeared thermal paste over it while installing.

But it seems like Gigabyte is really weak for DDR4 on LGA1700.

>>106191681
don't bother. a1 and a2, b1 and b2 shares the same pins

>>106190949
12c/24t 10800x3d can't come soon enough

Stupid question.
Can Q-Flash be used to reset BIOS using the same version as the BIOS that is already installed or will it check that it's the same version and leave it as is, not reverting whatever I changed to fuck it up?
If removing the battery doesn't work, that is. I'm asking cause it already didn't work before and I used qflash to update, now it's on the last version and I'm worried that I won't have this to fallback on as a last resort.

>>106190949
wtf why is r7 9700x got BTFO by some gen 10 and i5-14600 shit?

>>106191781
you should be able to reflash the same version

>>106191800
Read the chain ESL
>>106191231

>>106191781
your cmos reset isn't working and you're trying to just reset it by reinstalling the bios? you're either fucking up or shits fucked

>>106191231
that's an Asus-only feature
I don't see it on my asrock board on bios 3.15

I have 1080ti + amd rx590. I may play around with muh fake frames (smells like a forced meme). but then I thought maybe can I use my rx590 to AI upscale to say 1440p? I use arch btw.

>>106191812
It's called different names on different mobos. MSI is "latency killer" for example.

>>106191812
Retard.

>>106191840
good to know but i went through every single cpu related submenu and couldnt find it. i only saw the ol' gaming mode setting that disables a ccx and smt

So will GPUs actually improve next gen or will it be a second "smallest leap in 15 years" event?

Realistically, what games can I not play on Linux or with wine? If it's just cod, fortnite, and apex then I'm installing linux

>>106191808
yeah, cmos reset didn't work so I had to update the bios
>>106191803
thanks

>>106191862
Pretty much anything that doesn't have anticheat. Games that have problems on Windows itself due to being too old might also be a bother. (Though in my experience Proton runs all my decades old boomer games without much issue)

That being said, you're better off asking on the friendly Linux thread, not here.

>>106191856
Probably better than this gens uplift but if you're waiting for a massive leap like we used to get it's probably literally never coming.

>>106191856
moving to tsmc n3+ will give a decent efficiency and performance improvement over n4/n5+

>>106191883
Will do

>>106191805
With that turned on how much fps can I get here?

>>106191921
pretty sure he's talking out his ass since that setting hasnt been benched for this game

>>106191936
That fps is way too low for r7-9700x isn't it? I thought it was a pretty good chip?

why gigabyte puts main nvme below gpu slot?
nasty

>>106191356
dude a 9800x3d is not as slow as a 9900k lmao

>>106191363
harsh reminder that testers publicly publish results that are completely unhinged without 0 thought in their mind

>>106191417
ask the tards who benchmarked it

>>106191720
Do you think a 9900k is the same as a 9800x3d?
ask yourself this one question
Do you have critical thinking skills?

>>106191604
is this bait or general retardation at this point?! I CanT TELL HELP

>>106192090
maybe they used intel's performance library that defaults to x87 instructions for unsupported cpus.
whatever the reason is, having more than 16 threads of the same cpu architecture is able to bruteforce performance looking at the 9950x3d and 9800x3d. the only unexplained result is the 7950x3d scoring that low.

>>106192112
>9800x3d having half the fps of a 9950x3d (the same cpu) is fine
>9800x3d having the same fps as a 9900k is fine
>nothing out of the ordinary
>this is fine

>>106192090
>harsh reminder that testers publicly publish results that are completely unhinged
The only people who are unhinged are those who believed that benchmarks at 1080p with a 4090 apply to every system configuration, that you should unironically expect a 30% boost with an X3D CPU despite your meddling mid-range GPU.

GPU is often a bigger bottleneck factor than CPU, always has been.

>>106192112
>nobody else except the germans got these unhinged results
what a mess

>>106192140
>already forgot about the intel math libraries
you must be new

>>106192140
>9950x3d (the same cpu)
Literally 2x moar corez you simpleton.

>>106192155
Who else has benchmarked cpu performance in the game other than Hub (userbenchmark tier)?

Hello bros,
My server seems to arbitrarily shut down and I can't pin the root cause. Netdata shows that the CPU was mostly idel 3-5% usage, memory usage ~50% not increasing, no I/O pressures or such and then suddenly poof the graphs are gone, my system is offline and then it restarts to do the same thing all over again.
This has been happening on and off for some time now. I checked the PSU is also not overloaded, it's 550W (440W at 80%) capable and my components barely add up to 300W max rated.
I'm running Proxmox on a i5 10400 (65W) with 2 RAM sticks, 16+32, 3 enterprise HDDs and a NVIDIA 4060 TI capped at 125W.
Help is greatly appreciated :)
Xposted from /hsg/

>>106192353
>>106192353
>>106192353

someone make a proper thread instead of nicotroon shill one

>>106192465
https://desuarchive.org/g/search/text/nicotroon/
>34 results
parasocial relationships are not healthy and you should stop

>>106192168
referring to where intel nerfed performance on amd in some instance? I forget
how is this relevant

>>106192175
you think adding 8 more cores in another ccd scales in games? you think it adds 2x the fps?
do you?
do you think a 9800x3d performs like a 9900k?
do you?

>>106192196
The jew man https://www.youtube.com/watch?v=Wyubzqb-VuY
If there is a person to hate on amd this would be the first man

would u buy a used 3070 for 184 USD(in my shithole not US) in the year of our dear lord 2025 anons? It is the best performance per dollar GPU atm but im not sure how it will hold up in a couple of years.

>>106192923
If you're willing to work around the VRAM limitations (lower settings, etc), sure.

>>106192528
>you think adding 8 more cores in another ccd scales in games? you think it adds 2x the fps?
yes
>do you think a 9800x3d performs like a 9900k?
yes