File: moNJxZ3bYyDp2R9poFcAy6-970-80.jpg (76 KB, 638x493)

76 KB JPG

first

>>102374952
my american friend needs a $1000 pc for high fps gaming :)

>>102375061
buy used and turn off RT

any good pc cases with a vertical GPU mount?

>>102375124
where and which one?

File: 500dx.jpg (59 KB, 640x911)

59 KB JPG

got a Be Quiet 500DX case to fit my new 4090, seemed like good value
anybody have one?

File: 1597660711194.png (97 KB, 975x314)

97 KB PNG

OLEDBROS..

>>102373413
reposting to get some more suggestions.

>>102375345
If this is a gaming first machine, grab a 5700X3D off Aliexpress for $166:
https://www.aliexpress.us/item/3256806822307062.html
If it's apps/productivity focused a 5900X/5950X would be good.

File: Final ARL Leak.png (64 KB, 899x481)

64 KB PNG

https://videocardz.com/newz/intel-core-ultra-200k-final-specs-leak-core-ultra-9-285k-boasts-816-cores-5-7-ghz-boost-and-250w-max-power
Finalized Arrow Lake specs leaked ahead of the Oct. 24th launch. I'm a bit surprised by the clockspeeds I thought would be nerfed more if not due to architectural changes then shortcomings by TSMC. I also thought Intel be more gun-shy on power after the Degradation Crisis but nope Thermal Velocity Boost has actually been expanded to lower-tier models (previously only an i9 thing) and PL2 still at 250W

>>102375345
5700X3D
5800X3D not worth it unless it's cheap
Your monitor or headset could be a good upgrade

>>102375553
thinking of getting the 245k when it releases, I hope it's not shit

microled monitors by 2026 trust the plan

File: 1725670175117434.webm (688 KB, 1446x1080)

688 KB WEBM

literally what the fuck is this shit?
why do you faggots shill this chinkslop?

>>102375699
TWO MORE YEARS

>>102375756
>random /g/ seething and falseflagging
fucking lmao
embarassing

>>102375770
it's one of the GPU brands bots malfunctioning

What is the best Anti Aliasing mode for games

>>102375818
supersampling

>>102375756
It's literally just autistic pedants who have a single bug bear with windows that they're willing to make every other aspect of their life inconvenient out of spite

>>102375553
>PL2 still at 250W
Pretty insane. CPUs really shouldn't consume that much: 65W is enough unless we're talking a huge number of cores

>>102375818
DLAA or DLDSR + DLSS/DLAA. Can't tell if troll

>>102375818
Super-Sampling Anti-Aliasing

>>102375553
Ecores trash
lower clocks
250W boots (actually 400W)
just get to 18A already

>>102375850
>Can't tell if troll
It's not I know TAA is bad but I don't know what the others are really doing

>>102375756
>win 11 removed the ability to drag files to the taskbar to maximize another window where you would drop them
what the fuck steve jobs? this is why people dont even want to pirate that shit

>>102375850
>CPUs really shouldn't consume that much: 65W is enough unless we're talking a huge number of cores
24 is a huge amount of cores
65W is sensible if you have 4 cores

>>102375945
they should make strong single/dual core cpus that can compete with current gen

>>102375963
yeah sure
i would buy 2 cores at 7 ghz

>>102375978
Have you tried running 2 cores at 5.5-6ghz? It's a shit experience

>>102375898
linux sisters.....

>>102375699
>>102375758
>microled monitors by 2026
I've seen no evidence or recent news that MicroLED is anything but vaporware for the next 10 years outside of panel sizes comparable to the broadside of a semitruck costing as much as one to boot. I have more faith (and still not much) in Sharp commercializing QDEL at this point.
https://www.youtube.com/watch?v=eONWY3kbZc0
https://www.youtube.com/watch?v=p_L9JqOTqNw

>>102374876
What you cooling with anon

File: 4k native vs taa.webm (3.74 MB, 1920x2022)

3.74 MB WEBM

>>102375897
>I know TAA is bad
Temporal techniques are generally the best if implemented well. High resolution will really minimize the blur, which is why it's often worth using DLDSR to reconstruct to a high res if you have a 1080p or 1440p monitor (using DLAA, or DLSS if your graphics card isn't strong enough otherwise). Then it downsamples. The only other AA that's decent in my opinion is SMAA, but it's still pretty jaggy and sparkly in motion. Here's a comparison I did with BG3
https://imgsli.com/MTk3MTUz
You can use the drop down boxes to switch. Doesn't really paint the whole picture, as those are just stills, but you can see that 4K DLDSR + DLSS is clearer and more detailed than 1440p DLAA.

>>102375945
Even for eight P cores 65W is enough

>>102376128
modern slop looks like shit
old games that support msaa still looking good today apart from blurry textures

>>102376128
>Even for eight P cores 65W is enough
one P core can pull 50 W on its own if you allow it
you only get 8 W per core if you add an artificial power limit (which you're free to do, but i don't see the point unless you're having cooling issues)

anons , in terms of performance. Just how reliable are lesser known Cn brand gpus?

>>102376223
>msaa
Eh, I used MSAA in Nier Automata because there was no other option, and it wasn't that great. Also extremely heavy in terms of performance. I think you're remembering with rose colored glasses, especially considering how simple a lot of those old games are. The only relatively modern game I can think of with MSAA is RDR2, and it's horrible compared to TAA, both in image quality (whole screen shimmers) and in performance

File: 14900k power scaling rela(...).png (110 KB, 560x2170)

110 KB PNG

>>102376256
>one P core can pull 50 W on its own if you allow it
That's my point, it's too much

>>102376330
desktop PCs don't run on batteries
personally i don't like losing a whole generation of performance because i was scared of a few watts
up to you though, this is why we have BIOS options

>>102376330
S T O P
P O S T I N G
S H I T
C P U S
R E T A R D

Is waiting for the 9800X3D worth it?

>>102376566
You've waitedTM this long, but probably not

is there a significant difference between the 7735HS and 7840HS for performance / gaming?

>>102376566
No, I don't see anything AMD could reasonably do for it to escape the Zen 5% Death Sentence hanging over it's head. The only thing worth waiting on is Arrow Lake (and that's even iffy since mobos could be stupidly priced) or sales on the 7800X3D as we close in on the holidays.

>>102376724
>No, I don't see anything AMD could reasonably do for it to escape the Zen 5% Death Sentence hanging over it's head
zen 5 is severely memory bandwidth limited (the infinity fabric is limited to DDR4-era speeds)
more cache will give the cores the chance to do more useful work
don't expect the same uplift as 5800X3D to 7800X3D, but expect more than 5%

>>102376712
I can exploit both the same, bruh.
Everybody can, why do you buy these in 2024?

https://m.youtube.com/watch?v=MRNo-yYeDu0

>>102376712
yes. the former is one generation older (zen3) and the latter has faster cores and igpu

>>102375199
yeah had mine for almost 4 years now, its working great!

Got a 4070 super as a gift.
Whats a good cpu to upgrade from a 5600xt? I don't feel like dropping $400+ for a 7800x3d, but don't want to keep bottlenecking this thing on video games.

>>102376984
>bottlenecking this thing
You aren't in most modern games

>>102376984
cheap 5800x3d?

also don't worry about bottlenecking

>>102376984
The only time you would ever get GPU bottlenecks with those cards is playing CPU heavy games. Like City Skylines, or the airplane sim game.
Just ignore the rem bottleneck entirely. People don't know what it actually means and how it impacts their performance and they freak the fuck out for no reason.

>>102376915
is it worth paying about $100 usd for it

>>102377081
absolutely

>>102376984
Your CPU is fine dude, are you actually bottlenecked that hard? your choices are to spend $200 on a 5700x3d or $700 on a full 7800x3d setup and both sound awful to me. I would rather just get a higher res monitor and be GPU bottlenecked

Radeon sucks ass

>>102377135
what GPUs don't?

>>102377081
2024 is 1000 + 1024

live free or perish

my wallet votes

>>102377135
How do I stop my 4070 ti super from driver crashing so much?

>>102377191
kick the side of the mac

Anyone here have any experience building for local LLMs? Don't need anything too crazy and I'd like to focus exclusively on that for a crypto project I'm working on.

I already have a 850w power supply but need to replace everything else, budget would be around 1200$ but I'll wait for cyber Monday anyway. USA

did i say mac? i meant hack

Can you believe it? A crypto shitter in year 2024

>>102377230
i told you to buy btc in 2012 bro not my fault you didnt listen and resent me because you're still poor

>>102377248
Still not buying your shit coin

Are there AiO coolers that are worth it?

>>102377363
No because they malfunction quite often

Do AMD really think 9 month worth of r&d is enough to catch up to everyone else in AI?

>>102377374
no they dont.

i have had mine since 2017.

>>102377363
the Arctic Liquid Freezer III

File: 1710765788721087.jpg (38 KB, 1280x720)

38 KB JPG

>upgrade cpu
>the microstutter in games is gone

>arrow lake max power draw is still 250w
Man fuck it I’m just going to set the power limit to 190w like my 12700k and continue using my nhd15. I’m not getting an aio you niggers can’t make me

>>102376772
didnt know this guy, im watching part 1 right now
disgusting stuff, i guess all those 7gods like me that waited right after steam dropped support for it "upgraded" to 10 ended up being a good move

File: googles.jpg (32 KB, 420x299)

32 KB JPG

In the last thread I posted about checking for GPU bottleneck and an anon posted advice but the thread got archived.

>>102375141
>>102375520

Anons advice was helpful but not what I am looking for - I am looking for a synthetic benchmark that will identify where a bottleneck might be between my system and GPU, not for any specific game but to benchmark the components and report which (RAM, CPU, PCI bus) is getting saturated first.

Any advice is appreciated.

>>102377522
>I am looking for a synthetic benchmark that will identify where a bottleneck might be between my system and GPU, not for any specific game but to benchmark the components and report which (RAM, CPU, PCI bus) is getting saturated first
different workloads saturate different components
you don't understand what you're asking for

Can someone explain the Zen 5 disaster?

>>102377568
overall 5% more powerful
50% more expensive
2x core-to-core latency (performance regression in certain tasks)

>>102377568
AMD marketing made bullshit claims after engineering made a statement about how much they're expecting to see an uplift and then those uplifts didn't materialize because they're bottlenecked by the infinity fabric which they couldn't get to run at 2400mhz on everything for Zen5 so went with the safe 2000mhz of Zen4.

>>102377568
CPUs are advancing so fast that software (OS/Program schedulers) cant keep up. This has been going on since at least Ryzen 3000. You needed power plan and additional updates to make use of it.

>>102377568
the aliance between intelaviv and mcrosoft stopped being subtle some time ago
take a guess who loses in this situation

>>102377388
What are you using?

>>102377522
in a gpu theres only two types of bottlenecks. core bottleneck and memory bottleneck. you run into memory bottleneck at higher resolutions, you run into core bottleneck at lower resolutions. you can verify which one is bottlenecking in hwinfo64 by staring between memory controller usage and core usage.

>>102377602
Amdisaster

>>102377649
AIOs will fail or corrode over time anon, just get a good reliable noctua cooler

>>102377672
most aios are using anti corrode fluid. the failure point is likely losing water overtime but you can always refill it.

>>102377521
No wonder my 5600x and 5900x ran like shit feel stupid selling them now even if my 13600k is faster than a 5800x3d in w10 and 11

Will definitely be running w10 and Linux on my htpc build

>>102377692
Most AIOs are non-servicable. There are a few but they stopped doing it. The pump can also fail.
I still like using AIO since it's quieter than a tower cooler. But ever since I put my PC in another room sound is not an issue anymore. Once it fails I'm just getting the fattest CPU cooler ever.

File: noctuafags.jpg (67 KB, 800x633)

67 KB JPG

>>102377672
>noctua
kill yourself

>>102377383
No but they want everyone else to think that so they will buy their products

>>102377626
REtard

>>102377383
>>102377871
what are you talking about?

>>102377877
>>102377814
REtard

>>102377568
its the j__s

>>102377877
>>102377890
resident evil tard?

looking at B650E variants for a 7800X3D rtx4090 4k gaming build, budget itself isn't important, I don't wanna be kicking myself when I try and upgrade in 6yrs time that I got the wrong mobo. I was looking at the strix B650E-E, but it's either really hard to come by in Australia or its like $700-800AUD. if it's worth it I don't mind, I can get a B650E-F for $469.00, or NZXT N7 for like $450, but it looks worse performance in nearly every way

any help choosing between all the B650E variants would be helpful bros

>>102377814
If you can remove the water block then it's serviceable. Most are able to let you do that. The whole Arctic Liquid Cooler problem had you open the block to change a gasket and refill instead of sending it back to Arctic (but that's an option too)

>>102377649
its an arctic aio, not sure which one.

i still runs as quiet as when i installed it 7 years ago.

>>102377942
get an asrock b650e taichi lite from newegg instead

>>102377602
>>102377607
Wait wait wait…why would they release a part that’s only 5% faster than previous gen? Don’t they have tools to model performance? Was it that the Infinity Fabric tech was expected to develop apace but couldn’t realize that for some reason, so they just put out an under-performing part to at least get back some sales on the Zen 5 product cycle?

>>102377631
So, is Microsoft giving AMD the finger on thread scheduling again?

Also, can someone explain what’s going on with AMD’s high-end GPU business? It’s being dropped?

File: thestupid.jpg (94 KB, 600x682)

94 KB JPG

>>102377542
kek i think i do know what I am looking for after 30+ years in systems support

a synthetic benchmark to stress test everything, really saturate the CPU/GPU/RAM/storage *all at once*, and record the result in detail but it seems this is not available for free.

was going to have a look at performance monitoring while using something like Benchmate or OCCT in Windows but never mind because I think I found my issue - had to change QPI snoop mode to "Directory with OSB" and turn on PCI-E performance mode on my antique system, disk performance went up in Passmark and stutter disappeared in my CAD app when rendering large models (so, probably disk was throttling transfers to the GPU).

>ancient Z840 workstation
>forgot to backup BIOS settings before I started swapping cards around to accommodate my new 4070 Super
>card swap resulted in a no-post condition
>had to reset BIOS to troubleshoot
>old workstation has a lot of things to twiddle in BIOS
>i am most having the stupid today
>probably all that weed i been smoking

>>102378296
do you udnerstand how GPUs work? and it haven't changed since 1999. Arch scalability is just a question of wasting wafers.
I'd rather they had 7800XT every year for $500 than a 7900XT for $1200.

>We have accelerated our AI roadmap and are on a one-year cadence of new products.
>- AMD's CEO Lisa Su
Market going to be way more fun if they can pull it off. Fuck xx90 who cares, I am and many others not buying it ever.

>>102377653
bottleneck spot on but was disk drive bottlenecking the RAM >>102378388

in my system the PCI bus is only version 3 so there is a cap there but it didn't explain the stuttering I was seeing

Why does literally nobody care about AMD GPUs? They're not even hated, just nobody cares about them

>>102378036
>i still runs as quiet as when i installed it 7 years ago.
you sure it's running?

>>102378469
same reason women don't care about android.

>>102378296
It's pointless for them to compete in the halo product category because no one buys them anyway even if they're close, just look at the 7900XTX that got its priced slashed so much already.

>>102378477
I would argue if they put it in correct price bracket from the start it would fly off the shelves.

>>102378469
Brand.
Nvdia always won in that regard, whether their gpus are shit or not.

>>102378469
Radeon drivers have pissed a lot of people off

>>102378485
It's a massive 500mm wafer. There's a limit to how low they can price it before it becomes a waste of money to even make them.

Meant die size in >>102378528
I can't imagine the yields are that good to begin with.

>>102378528
And they priced it that low anyway. so what was the point to sele those 50k cards before cutting price on another 300k they made?
>>102378541
yields were fine, main chip is only 300mm2

>running a Samsung 840 evo as my boot and gaming drive for 8 years
>Finally upgrade to a pcie nvme because I need space
>No difference in boot or loading speeds
Hmmm...???

>>102378496
I had mroe problems wih nvidia than AMD personally, but I started juggling brands only after they adrenaline became a thing in 2016-ish. Before drivers were terrible for real, now it's just memes marketing uses well.

>>102378567
there is a difference, you just cannot perceive it, you probably shaved off a quarter second

>>102378567
loading is CPU dependent more than SSD speed.

>>102378469
AMD GPUs are extremely popular in DIY and consoles. But DIY and custom assembly is niche sector compared to OEM which is always Intel/Nvidia hence why they got sued for antitrust.

>>102378609
wait, office OEMs still buy dGPUs when superior APU exist?

>>102378609
>AMD GPUs are extremely popular in DIY
AMDelusion

>>102378631
offices are buying laptops
these laptops by default ship with intel+nvidia

>>102378632
you will find a amd gpu in a custom build before you find it in a oem pc.

What are the GOAT CPUs and GPUs from their time?

>>102378609
>AMD GPUs are extremely popular in DIY
occured to me, it was a clever move to wait for Sony to adapt FSR4 for their console to force dev implementation in every single PS5 game.

>>102378609
>>102378659
What if Microsoft put a 5090 in the next Xbox?

>>102378567
This is why I always, always recommend not getting the Samsung 990 or 980. Boot times and load times between a 990 and say, a Crucial P3, is very minimal. Only reason I would ever recommend the Samsung NVMe is if they write a lot and need the endurance, it's never about the speed.

>>102378649
Cpu: q6600, 2500k, 7500f(for some reason not available in NA)
GPU: gt 8800, gtx 970, gtx 1080ti,

>>102378671
chance of that happening is 50/50 they either do or not.

>>102378567
You fell for the sequential read speed meme
Shoulda got optane 905p

>>102378690
>no Athlon 64
>no R300
you know nothing anon.

>>102378727
Go bed grandpa

>>102378690
>gtx 970 (3.5GB)

>>102378671
Microsoft is never putting Nvidia in a console again. They hated their partnership with OG Xbox. That's why 360 is powered by Radeon and was many peoples first real experience with HD gaming with accurate HW T&L.

>>102378735
pretty sure Intel still pays AMD for x64 and Nvidia pays AMD for HBM, haha

>>102378767
nvidiots still think unified shader was nvidia invention because of x360, AMD put it in consoles first

>>102378767
>accurate HW T&L
It's schizo hour

>>102378789
>>102378794
What are things AMD/ATI did first before Intel/nvidia?

>>102378727
no AGP cards
no buying a Gigabyte GV-N68128DH GeForce 6800 and BIOS modding it to unlock the additional 4 pipelines

>>102378735
no

>>102378794
when people say consoles look better then PC I don't think theya e lying, they simply used Nvidia GPUs which cut rendering corners too much. Nvidia hardware I have no issue with, but their software layers do cut corners and make picture thinner. I owned equal number of Radeon and GeForce cards btw.

>>102378818
tessellation comes to mind

>>102378835
>when people say consoles look better then PC
They don't say this

>>102378846
you don't speak to many normies.

>>102378850
Normies don't have PCs retard

>>102378835
>>102378846
>>102378850
>Xbox 360 GPU is somewhere between an AMD X1950 and 2900 XT

Was it kino in 2006?

>>102378794
back in the day website had to include gameplay footage from both ati and nvidia to highlight the differences in antialiasing, AF, color depth, etc. radeon always looked better because it did everything accurately. geforce took driver level shortcuts to make weaker hardware better. its really no different from today how they're pushing dlss so they can sell you less vram capacity+bandwidth.

>>102378868
>back in the day website had to include gameplay footage from both ati and nvidia to highlight the differences in antialiasing, AF, color depth, etc

Post examples of games where they did this

>>102378727
How about the HD4870? I remember it being good back then and that was one of the apex periods of AMD market share.

>>102377885
Amd is switching to AI upscaling for fsr4. The only problem is AMD only have 9 months worth of r&d

>>102378874
https://www.anandtech.com/show/1546/8
https://www.anandtech.com/show/1546/9

>>102378927
Will Valve ever remaster HL2 + EP1 + EP2 or are they afraid of the backlash for no HL3?

>>102378470
yeah.

my pc goes up to 65C at full load.

>>102378874
nvidia shipped 3dmark2005 with their gpu and it couldnt even run the hw t&l shit
https://youtu.be/ewkmWoCo574?t=478

File: ss_62c484ade710adabd62d65(...).jpg (2.08 MB, 1920x1080)

2.08 MB JPG

>>102378938
it would be interesting to see a remaster, but the game still looks beautiful on modern hardware without changes
just run it at 300FPS 8X supersampling

File: Screenshot 2024-07-01 at (...).jpg (1.04 MB, 2560x1440)

1.04 MB JPG

>>102378927
>From the screenshots above it is tough to tell the difference between the two competitors.
>The two look quite similar in our opinion
Compare with today

>>102378854
You have to be 18 or older to post here

File: 1725324622998.webm (3.86 MB, 852x480)

3.86 MB WEBM

Halo 2 was too ambitious. It was like two generations too early for the tech it was trying to push. Even if they did manage to ship stencil shadows, the resolution/performance would have been too low. pic-related is only simulated/approximated btw, it's not even the real deal that could have been with pixel perfect shadows and dynamic lights.

Xbox had a high end PC-level GPU coupled with a low-tier Pentium III 733 CPU with a crippled cache. The system has Direct X 8.1, shader model 1.1, vertex shader 1.1 and two Vertex Shader units, while the Geforce 3's only had one. The Xbox GPU was quite powerful.

File: Starfield 1440p taa dlss fsr.webm (2.51 MB, 2048x1152)

2.51 MB WEBM

>>102378973

>>102377383
idiot, AMD is in AI as long as ROCM exists, nvidia simply has CUDA dominance

File: Half-Life-2-Lamaar-RTX-On(...).jpg (155 KB, 1456x819)

155 KB JPG

>>102378938
>>102378957
You guys gotta be kidding me
I'd rather see a HL1 remaster though
https://youtu.be/R0-F8sPprmA?si=SmGQME18hzwJt9t-

I've never updated a BIOS before, and this ASUS FAQ looks like it was written by a pajeet with bad punctuation and broken english.
What guide would you use for TUF Gaming B650 plus? Just got done building the pc earlier

https://www.asus.com/support/faq/1044348/

What are the worst high end Nvidia cards?

>>102379107
RTX OFF is clearly better lmao

>>102378986
they all suck in different, fundamentally awful ways. where's RAW?

>>102379229
you put the update file on a USB stick
you reboot into the BIOS
you select the file in the BIOS flashing menu

that's it

>>102379263
Same way as installing windows?

Anyone here with a crucial T705, what's your boot time?

>>102379266
yup pretty much
right click > format > select FAT32
put the update file on the newly formatted stick

>>102379279
Thank you

>>102379266
>>102379279
to clarify, you don't boot from the USB stick like you would with a windows installer
there's a separate menu in the BIOS for updating, it's called Q-flash on gigabyte, idk what asus calls it but it's the same thing

>>102379291
I didn't really look at all the BIOS menu, would that be under the Boot tab?

>>102379300
they show steps here
https://www.asus.com/support/faq/1012815
it's under tool, ez flash 3

my mobo acts like it died for like 2-3 minutes after updating, so don't worry if it takes a while to boot after the update

>>102379333
Thanks. I just restarted and checked my version, I have 2613 while looks like the new one is 3035. Guess we're giving it a go

>>102379235
3080 family
>10-12gb
>Inferno Samsung 8nm
>Overpriced
>Artificially software limited by Nvidia

File: 6530-09-14-2024.jpg (70 KB, 485x451)

70 KB JPG

Latest BIOS
DDR5 EXPO On
All fine voltage wise on the CPU right?
Nothing touched in the bios beside EXPO on

Why is the 3060 12GB and 4060 Ti 16GB considered bad cards?

>>102379395
The cores are too weak to utilize all the vram so they only exist to pander to AI retards.

>>102379387
why are you worrying?

>>102379387
use hwinfo
is that vSoC? looks a little high (I have never played with am5)

>>102379235
Just off the top of my head.
>GTX 480
1.7% pork shoulders.
>GTX 970 3.5GB
If you consider x70 high-end.
>RTX 2080Ti
"just buy it"
>RTX 3080Ti
Would have been $1000+ even without the crypto boom
>RTX 4070Ti
12GB for $900
>RTX 4080
Crypto priced in a bear market.

>>102379395
They're just not a good idea for a pure gaming rig.

>>102378080
I like the idea of that one, can't find in Australia unfortunately

>>102379383
3080 is the only good x80 card in the last 10 years bro

>>102379395
I don't know who said the 3060 12g is bad. At launch it was nothing special but not bad. Maybe now its bad because it's a gen old?
The 4060Ti is easy, its because it should cost $250 with that performance

>>102379444
> looks a little high (I have never played with am5)
It's really not on am5
You can silicon lottery undervolting to 1.25V or lower for PBO and Cinebench minmax but 1.3V is what all mobo makers and AMD defaulted for pretty much the whole last year after ASUS killed CPUs with 1.4V
SoC voltages too high have not been an issue for the entire of 2024

>>102379525
it's the maximum allowed and the board makers just push 1.3 to all cpus
it doesn't need to be that high but it won't cause issues

>>102379235
3090, retarded split vcore power design that will occasionally fry the chip, half the memory is uncooled, it's a scam plain and simple.

>>102379444
7nm IO die goes higher
>>102379535
probably does cause some instability issues compared to lower voltages
you need clean signals for RAM and PCIe
gigabyte pushes 1.2V on AM4 for no reason and all it does is lower power budget for the CPU cores

Are AMD GPUs really that bad if you're only gaming?

>>102379536
Fuck your retarded preferences

>>102379535
Exactly it's a loose setting for stability reason.
While you can go lower your silicon could be low roll and be unstable so they push 1.3v, which is the safest voltage before 1.35V being the absolute maximun before the CPU start fry itself.
If your system is unstable at 1.3V the CPU was faulty out of the box

>>102379557
Not really. Especially at the low-midrange where you can't reliably run RT. The problem is at the high end where RT becomes the only way to meaningfully increase graphical fidelity. Or, to put it another way, if you're dropping $900 on a video card you obviously want to graphicsmaxx and spending another $100 to be able to /do/ that shouldn't be a burden.

>>102379581
Coping nGolem.
3090s are industrial waste from the enterprise line rebranded as top end goysumer cards.

>>102379642
>>102379536
>>102379235
Post the downsides of Nvidia

Did I do wrong by going with ATX instead of mATX /g/
I made a series of poor decisions throughout the year, starting with my PC dying and doing a partial upgrade where I'm stuck with an ATX motherboard on AM4 still
Since I didn't see it fit to upgrade my CPU in the very same year I decided to go for a better GPU
Ended up buying a GPU that would perfectly fit on a small build, but then I got to look for a compact case and they were all trash garbage
In the end I bought a Lancool 216 which is massive on my desk because I refused to buy another AM4 motherboard in 2024 just so I could get something compact
yesterday I hit my new case with my elbow and thank fuck it's heavy because it's at the edge of the fucking desk

>>102379671
No, your desk is too small. 190x80cm is the bare minimum for a normal male imo.

>>102379671
What is the point of having your case on the desk in the first place? To better hear fan noise?

>>102379333
Thank fuck that it won't let you install the wrong stuff. I tried to install B650 plus instead of plus wifi.
Computer started again though so I guess that's good

File: 1700642973188.png (129 KB, 3186x430)

129 KB PNG

>>102379671
Even a cheap mATX case fits all GPU's all the way up to the 4090.

>>102379706
lol oh shit
at least it gives you the exact model in the BIOS, which is important because sometimes there's multiple revisions of the same board

>>102379546
It's not really for no reason.
The reason is:"At these setting and values we are guarantee 99% of the user machines will work and we won't kill our customer hardware because if we do we are going to become a Gamer Nexus headline and that WILL sucks".
Are the settings optmized? Not really
Are they safe though? No Gamer Nexus headline related to mobos for a while so yes
Can you actually optimize and go lower? Maybe, go and tweak on your own and see if your system crash or stutter because RAM or CPU are on shit side of the silicon.

>>102379693
Also this, if you can grab both ends of your desk, it's too small.

>>102379729
>At these setting and values we are guarantee 99% of the user machines will work
1.1V is more stable than 1.2V on the vast majority of 3000/5000 series CPUs
the SOC does not act like the CPU cores, more isn't always better

>>102379699
Simply put the case on the floor would get very dusty and either A. being really kickable or B. getting in the way, it's a lose lose situation
I actually have a corner desk, but the way I set it up, I really want the corner, not the sides

>>102379736
Yeah, 190x80cm just about fits 3 24" monitors. Never mind larger monitors.

>>102379557
amd is fine for gaming
nvidia shines when you do stuff that's not gaming

>>102379787
like gaming with raytracing or dlss

The problem with PS5 isn't the specs

The problem with PS5 is no games or exclusives and that Sony are porting everything to Windows DX12 lol

How long does the initial memory training take for a new AM5 Zen 4 build?

>>102379804
depends on your mobo and bios

>>102379804
depends on the BIOS version
for recently updated boards it should boot in like 10 seconds and then take a minute or two once you enable EXPO for the first time
AMD figured those issues out a while back, but obviously not all boards are updated out of the box

>>102379803
The problem with both PS5 and PC is that the games are boring

What do you use to benchmark after you finish a build? I'm looking at todo lists and people are saying that's something you do

>>102379811
>>102379812
7600X + Asus Prime B650M-A WiFi II-CSM
Hmmm...everything turns on but even after several minutes, no output on my monitor
Should I give it a few hard reboots?

>>102379821
Software that you actually intend to use are the only performance metrics that matter.

>>102379815
Nintendo was always king

>>102379830
No old games like Settlers 2 are engaging, but they don't make them like that anymore

>>102379826
Latest BIOS?
try to turn on memory context and dram power down if they are off

>>102379829
I just play wow now. But I mean I assume they do benchmark tests to make sure everything is installed correctly? What would you use to find that out?

>>102379826
you should make sure everything is seated correctly
>>102379839
he has no output dummy

>>102379844
cinebench, aida64 and prime95

>>102379849
that no output could be an insanely long memory training or yeah, something is not built right
>>102379826
Any leds lighted up on the mobo?

>>102379839
>>102379849
We did it boys
I swapped the display output from GPU to motherboard and then did a long press power off
Turned it back on and got display output

>>102379804
>>102379826
Walk away and give it like 10-15 minutes. Also it should power cycle itself a time or several during the process too.

>>102379803
The problem is that hardware costs have gone up but console people aren't as aware because they only buy one every few years and don't need to care about specs. Then they look at the price compared to previous consoles and think it's overpriced when in reality you can't expect things to be like before.

>>102379855
>plugged in GPU no output
>plugged in the iGPU yes output
....are you sure the GPU is actually
A) fully inserted in the PCIE slot
B) plugged in

>>102379855 posted right before I hit submit on >>102379859
Glad you're in business now.

>>102379855
oh yeah
that is an issue with AM5
stupid iGPU lmao

>>102379867
Could be the case that the BIOS is set to use iGPU by default and won't look at the dGPU even when plugged in for an output on first power-on.

>>102379871
How you disable the iGPU at bios level to make sure it doessn't schizo like that?

>>102379892
idk i'm still on AM4
i remember when ryzen came out and everyone celebrated that they didn't have to pay for an unnecessary iGPU

>>102379867
It's fully seated, fans spin, power cable fully clicked in, and the LEDs on the shroud turn on
Could still be GPU but I'm going to let it memory train after setting EXPO timings and boot up again before I mess with any other BIOS settings

>Buy RM850x shift considering better cable management
>Motherboard cables are a fucking mess anyways because you -have- to twist them to fit them and they're not long enough to tuck away but not short enough that there's no room for these awkward fucks just hanging around
Cable management, what an autistic thing

File: IMG20240525213226.jpg (1.29 MB, 2645x2365)

1.29 MB JPG

>>102379954
There are two types of people;
those who use their computer
and those who do cable management

>>102380011
KEK

>>102380011
Yeah but what if someone just sneaks into your room at night and opens the side panel and then sees all of that shit

>>102380065
worse what if a girl opens the side of your computer case and sees that

>>102380065
It's a sign for them to leave it alone, like how you don't go bothering the duded swaying violently in the corner on the subway staring out into space with blinking for 6 solid minutes.

>>102380087
Anon it's virtually impossible for that to happen

File: cable management.png (1.1 MB, 924x955)

1.1 MB PNG

>>102380011
This is how it looks with those corsair shift PSUs

File: file.png (67 KB, 545x548)

67 KB PNG

Is this fine? TGP being above 110% TDP with 400W concerns me.

>>102375553
Huh, they're ditching hyperthreading? I wonder if that's because it lets them optimize the scheduler further.

File: power-gaming-peak.png (70 KB, 500x1210)

70 KB PNG

>>102380188
> TGP being above 110% TDP with 400W concerns me.
Your HWMonitor says temps are still in the low 80s so the cooler is doing it's job with respect to the thermal power dissipation not causing issues for the wattage being expended, and moreover EVGA's flavor of the 3080 (prolly includes a slight overclock) ran ~400W at max load anyway which was normal for when it was released.
https://www.techpowerup.com/review/evga-geforce-rtx-3080-ftw3-ultra/29.html
^Actually the card even hits 460W under power viruses like Furmark.
Nothing to see here and worry over.

>>102380188
>FTW3 Ultra
That was their heavy OC model so I wouldn't be surprised to see power limit shenanigans.

>>102375199
I opted for the shadow base 800 to fit my 4080 super. Still waiting for it to arrive, but their cases look nice.

>>102380188
>>102380236
Huh, low 80s? Max temperature there was 47.4 (hotspot)

>>102380225
>Huh, they're ditching hyperthreading?
Yes. To compensate the loss of MT Intel is roiding up the E-Cores. Skymont is on part with Raptor Lake's Raptor Cove P-Cores as far as IPC gains expected.
>I wonder if that's because it lets them optimize the scheduler further.
Probably a little of that, and also being tired of the security risks that kept happening that caused them to need to release microcode patch mitigation that would nerf performance.

>>102380245
Whoops looked at the wrong row, that makes it even less of a concern.

File: DLSSbutBetterGhostingThan(...).webm (1.67 MB, 1272x1080)

1.67 MB WEBM

>>102378964
Compared with today it’s quite awful

File: Intergraph InterView 28hd96.jpg (40 KB, 640x480)

40 KB JPG

>>102376128
>High resolution will really minimize the blur
We could've just continued to develop CRT or Plasma, both got pixel overbleed with which render most AA redundant for anything 1080p and above.

>>102379754
You are not wrong but what mobo and AMD cares is maximized compatibility and no hardware damage.
SoC at 1.3V is not optimized but it's safe and guarantee to run regardless of silicon lottery on the cpu and whatever RAM modules are used on normie machines, thats all they care.
Most user would never notice issues that are not explicit full system crash

>>102380382
i don't know enough about AM5 to agree or disagree with you, you're probably right because old BIOSes were running like 1.4V by default
i'm just commenting on AM4, you should NOT run more than 1.1V on that platform
you lose stability and power budget for no reason, cranking voltage is not always the solution and mobo manufacturers need to learn that lesson before they do something really stupid again
https://github.com/integralfx/MemTestHelper/blob/oc-guide/DDR4%20OC%20Guide.md
>By contrast, memory instability can occur when SOC voltage is too high. This negative scaling typically occurs between 1.15 V and 1.25 V on most Ryzen CPUs
>All of the tested samples illustrated negative scaling (i.e. more errors or failures to train) when higher than 1.150V SoC was used. In all cases the maximum memory frequency was achieved at =< 1.100V SoC voltage.

File: hot_singles_in_your_area.jpg (381 KB, 1439x2144)

381 KB JPG

any reason to spend more on the NT-H2 or nah?
$AUD

>>102380493
spend $10 on PTM7950 sheet

Alright
2 frontal 160mm intakes
1 rear 140mm exhaust
2 intakes at the bottom (on top of the PSU shroud, there's just a bit of clearance between the PSU and the rear bottom fan)
Should I add anything on top? Maybe one extra exhaust fan considering there's too many intakes?

>>102380506
no, and remove ones at the bottom. one direction. do not create vortexes of warm air in the case.

>>102380501
I'm not quibbling over a dollar lol. is that one any better?

Just to be curious i tried to go and set SoC from 1.3V to like 1.25 but when in bios when going ovveride it stay on auto below and does not allow me to make any change in the values.
Ok?

>>102380493
MX6 is $16

>>102380550
think you forgot a screenshot
anyway don't bother if it's working for you, AMD has already determined 1.3V is safe

Best cheap pwm case fans? It's for an old PC that I have so I don't want to expend much

>>102380515
What do you mean vortexes? Bottom are intakes (pushing air up, not down). They're not taking air away from the GPU. They're still pushing air towards where the exhaust fans would be.

Anybody can explain to me why people rather buy AMD AM4 Socket Mainboards that support only DDR4 and CPU than Intel 1700 Socket Mainboards and CPU that support DDR5, apart from the price? I mean the Intel stuff costs more but it's more future proof.

>>102380633
People don't want to buy intelaviv

>>102380601
the Arctic P12s

>>102380633
>future proof
both are dead sockets

>>102380633
I'm not a fan of rust in my computer.

>>102380633
>>102380662
>>102380673
What made AM4 so kino?

>>102380639
I mean I don't like Intel too, just looking at vulnerabilities like Spectre but from a cost/benefits view, shouldn't intel be the better system if I want to upgrade?

>>102380633
>future proof
intel shill script never changes, huh? still same old song and dance routine?

>>102380690
Nearly a decade of support with some of the fastest cpus money could by for a given time and the 5800x3d being the GOAT.

>>102380709
>and the 5800x3d being the GOAT.

QRD on this?

t. Intel user

>>102380690
no arbitrary restrictions
you can use PBO and curve optimizer on any AM4 CPU
you can do memory overclocking on any AM4 CPU
you can use any AM4 CPU on any AM4 mobo

I bought a 7950X3D. What am I in for bros?

>>102380633
>future proof
Hind sight is 20/20. Only idiots believe that remorse change anything.

File: Screenshot 2024-07-26 003054.png (2.53 MB, 2562x1742)

2.53 MB PNG

>>102380730
>cheap enough
>wipes the floor with every other cpu for vidya
It existing cemented AM4 as an amazing platform and the reason why the 7800x3d came into existence. Pic is an extreme example but highlights it nicely that cache matters.

were any of the X3D Zen 4s OCable?
Is the 9800X3D allegedly being unlocked a major thing or a nothingburger?

>>102377942
For AM5 anything above the very cheapest boards is more than sufficient for a 7800x3d which sips power. You are buying based on desired features like number of PCIE, M.2 slots; SATA ports; USB/Ethernet ports; built in WIFI; etc. If you are keen on the ROG aesthetic the B650E-E looks good with 4 M.2s. Alternatively, the Asrock X670E Steel Legend is $379 AUD and has plenty of power stages, a 8 layer PCB, shit tons of USB and 4 m.2 slots.

Are sata ssds good enough for gaming?

>>102380923
for 99% of games yes but with Sony games (Ratchet etc) they prefer NVME SSDs and I imagine most games going forward from next year will want Gen 3 NVMe at a minimum

https://www.amazon.com/Alphacool-Eiswolf-AIO-Reference-Backplate/dp/B0CXM8D97F/

is this thing worth it or am i better off just keeping the card stock?

>>102380923
>>102380941
Are PS5s actually good hardware as consoles for devs?

File: apu2.png (218 KB, 2173x1150)

218 KB PNG

Am i retarded? What is this apu2 reading mean? I used 2 separate pcie instead of plugging the same cord, is that it? Is that bad?

>>102380969
Let's be honest, we only buy better hardware because developers are really fucking lazy and manufacturers need to drive hardware sales

>>102380978
Anon, stop taking drugs.
It's just iGPU in your 7600X3D.

>>102380969
As a base, it was really good at the time. 8C/16T Zen 2, Gen 4 SSD level performance and a 6700XT equivalent but it massively over-promised as per normal
>>102380978
that's the onboard graphics (APU) from your CPU

>>102380997
x3ds have igpus?

qrd on DDR5 memory?
is it bad to go over 2 sticks?

>>102380997
>>102381006

>>102381009
>x3ds have igpus?
I guess I didn't know that either.

File: AMD Ryzen™ 5 7600X3D Desk(...).png (309 KB, 1600x1018)

309 KB PNG

>>102381009
Yes, all desktop Zen4 and Zen5 come with iGPU.
It's very basic but it's there.

>>102381013
both Intel and AMD on CPU memory controllers fucking suck so if you want anything over 5600MHz (you do) you're best going for 2 sticks. AMD typically gets to 6000MHz CL30 but sometimes can push to 6400 CL32 depending on the controller and Intel can sometimes hit 7200 CL34 but the sweetspot seems to be 6400CL32
>>102381031
crazy how a quick google search would've helped

>>102381039
can it run anything that a 750ti wouldnt be able to?

>>102381044
>crazy how a quick google search would've helped
I trust you spergs more then google

>>102381049
The zen igpu basically exists to give an image out and render spreadsheets.

>>102381059
*Than
fick

>>102381049
https://www.techpowerup.com/review/amd-ryzen-7-7700x/22.html
it could probably play older games with FSR
>>102381059
love you too anon

>>102381049
750Ti is much faster than that iGPU.

>>102381071
It's all love baby, no troll. Thank you for helping me and others here

>>102381094
yeah but some games refuse to even start because muh low vram even though it would reach 20-ish fps for example forza horizon 5
I know doom eternal runs at low settings >720p and aside from cutscenes it's basically the classic console experience

>>102381138
>>102381138
I got the bundle. probably won't help much but seemed like a good deal.
https://www.microcenter.com/product/5006834/amd-ryzen-5-7600x3d,-asus-tuf-gaming-b650-plus-wifi,-gskill-flare-x5-series-32gb-ddr5-6000-kit,-computer-build-bundle

>>102380968
does anyone know?

>>102381167
>all that for 450 dollars
ameritards do not know how lucky they are to have microcentre

File: 1695405634103805.gif (37 KB, 69x60)

37 KB GIF

>>102381167
>asus
whatever you do make sure to flash the bios before you place the cpu in the socket

>>102381173
repaste with PTM7950 before considering watercooling, unless you already have a custom loop

>>102381186
I didn't do it before, but I did do it after. Didn't have any issues. Typing on it now, finished building like 5 hours ago.

>>102381193
i dont and thats an AIO the only reason i am even looking at it is cause i have no interest in doing a custom loop at all

>>102381202
After, I mean like, after I installed steam and bnet and everything else. I did bios last. Maybe not smart but I don't know any better

>>102381204
PTM7950, your cooler will be plenty good enough as long as you're not doing crazy overclocking (because why would you overclock the reference model lmao)

>>102381222
sighs and do i have to buy that from that faggot linus? i hate buying things from losers

File: slideshow.png (984 KB, 606x4285)

984 KB PNG

>>102381138
It's still awful experience, it's average FPS so it doesn't show all those stutters you will get.
If you care about having better iGPU there are 8x00G processors but those are slower in CPU tasks because of smaller L3 cache.
Buying used Rx6600 is still preferred.

>>102381229
no, but i'm not in the US to say where else to get it from. At least from him you know it's real

>>102381238
dammit you are right

File: 20240914_064210.jpg (1021 KB, 8160x3768)

1021 KB JPG

>>102381186
Wow stats are ok though, I think?

>>102381281
it's to avoid the doubled voltage limit from permanently fucking up your cpu in the long run

>>102381298
I read they fixed that awhile ago

>>102381306
yeah with a bios update

>>102381298
isnt that what intel is going through right now

>>102381315
Ya I did make sure to do that, good call. Better late than never.

>>102381330
>>102381330
>>102381330

>>102381317
intels are just drawing 300 watts by design
7800x3d draws about 100W but has a voltage tolerance of about 1.3V
the retards at asus fucked up big time and made the max voltage of the board 2+V and I think the bios settings displayed it incorrectly
then they refused to accept warranties because >customer induced damage lol xd

>>102381340
I'd be pissed. Only my 2nd time making a pc so I didn't know about updating beforehand. First one I made was long ago, Ryzen 5 1600 1060. Extremely happy so far.

>>102380968
does that say $340? wtf lmao

>>102381375
ya the main reason im asking here rather then plunging off of the cliff blind

>>102381528
why not get one thats 3x cheaper and has good reviews? there's suggestions in op

well not this op i guess but there's better suggestions in this thread
>>102381330

>>102379485
REtard

>>102380533
yea, it doesn't dry out for along while

>>102380624
it passes through GPU air, GPU air hits intake air from the front, they whirl together in place. ONE direction, make a tube out of your case.

>>102380830
thanks anon, looking at some X670E's now, hard to get B650E-E in Aus atm for some reason

>>102381710
I'm lost here, can you open mspaint real quick?
As I understand it the GPU needs fresh air from the bottom. It sucks up the air from the bottom front panel and exhausts it through the sides (and in this case, the GPU has a small hole through the backplate so it exhausts air there, it's even illustrated that way).
What I'm doing here is taking what would be a small hole at the bottom front and expanding it to the bottom as well. More air comes from there. The shroud cover at the bottom of the glass side panel is mesh, so it takes air from there without being too obstructed or rerouting it from the bottom front intake.
I just fail to see where this vortex would even be produced.

No one ever warned me when you go to apply new thermal paste to the cpu it sticks to the heatsink and rips off the board and fucks the pins.

Well I'll warn you all right here right now. If the cooler don't come off effortlessly release the cpu from the mobo and pry it off the heatsink.

Almost had a heart attack bending the pins back. I didn't even yank on it that hard.