It's officially over at my local stores and my regions amazon
B580 going for between 290 and 340 now

>>103608106
GPU's are almost always in low supply in modern times so prices go up quickly if something is selling well. Christmas time is the worst time to buy GPU's. Usually around May-June the prices trend to their lowest points.

https://wccftech.com/amd-radeon-rx-9070-xt-the-top-rdna-4-gpu-radeon-9000-new-branding/

No fucking way...

>>103608288
>Source: my ass

>>103608074
looks like an obese hound lying on its side

File: 1721086590398745.jpg (108 KB, 904x1200)

108 KB JPG

>>103608400
Finished on the couch

File: Screenshot 2024-02-18 at (...).png (615 KB, 478x615)

615 KB PNG

I sold online some old hardware in a good shape by good price to a buyer.
And now that buyer sells my hardware by 3x price. I fucking hate people.

>>103608288
its obvious. the 9800x3d + 9800xt meme build will live

>>103608469
you mean 9070XT

>>103608288
Bigger number = more better
It's over for nvidia

>>103608453
Ye i usually undercut everyone to make a quick sale and it ends up going to some chink or repair shop

>>103608288
you confuse your customers by rebranding when you shit out a bad product so you can capture the thousands of sales from people who dont know what they're buying
>see: core ultra 200 series, rtx2000

>>103608415
cool case, I have a Fractal Meshify 2 waiting for my next build, but having second thoughts.
New Be Quiet 501 is also interesting.

>>103608509
>>103608574
5070 consoomers are going to go wild when they see 9070. Red wave.

File: 1733710032242554.jpg (379 KB, 960x949)

379 KB JPG

What monitor are you guys using?
Been on the lookout and the only contenders so far are
AOC Q27G3XMN
Xiaomi G Pro 27i

https://www.youtube.com/watch?v=ZiFclVrlmro

Not really sure what to pick or whether to waitTM for MiniLED to get better.

Ok anons with how power efficient and having good its thermals are. What are the chances itll get a single fan model? Maybe the B570 will be an sff?

>>103608574
Everyone who able to build a computer by themselves is aware about generations, models and naming changes. The rest buy prebuilds or listen to the shop consultants, it confuses no one.

File: ff70a8e5-f0f2-577e-880c-a(...).jpg (64 KB, 800x600)

64 KB JPG

This is what happens if you go full reta- brightness with your OLED.

>>103608706
>Not really sure what to pick or whether to waitTM for MiniLED to get better.
miniLED monitors don't completely solve the IPS blooming or the VA black smearing.
There are like 0 miniLED monitors that handle local dimming + VRR gracefully, plus inconsistent visual quality between SDR & HDR modes.
Mainstream OLED monitors are putting price pressure on miniLEDs too, so manufacturers have less incentive to improve the tech.

I don't mind paying $600-800 for a good miniLED monitor, but it has to be near-flawless to justify that price.

>>103608949
>oled burn in has been sol-

>>103608949
if all this shit is going down because they're too prideful to use cooling fans i'm going to die of laughter

>>103608949
do ppl really run their monitors at full brightness? i have mine at around half and that's in a fully lit room.

File: ooGDL4VeawZiZpZ93oRWM8.png (175 KB, 2172x1712)

175 KB PNG

>>103608706
Between the two, the AOC is the better monitor overall, considering features and price. Basically you're looking for more zones, faster response times for VAs, and less input lag in the HDR mode as upgrades over the AOC. Input lag probably isn't going to change substantially, and the AOC already has way faster response times than most VAs (although there are better on the market), and more zones are on the horizon but who knows when. If you need a monitor, for me the AOC is the no brainer. It's pretty cheap anyway, so you can always upgrade it when something better comes along.

>>103608720
I would expect the B380 (if it exists) be be the SFF card

>>103608288
>amd trying to be more like nvidia
The funny thing is, and this is going to blow your mind, that _RTG_ is actually the group with no competition. Nvidia competes with itself, but RTG is just able to phone in its generations and features every generation and it will still remain in second place. At least, that was the case until recently, when Intel teleported into the GPU race. RTG has let itself fall so far behind Nvidia that there's a massive gap between first and second place in the GPU race (remember: historically RTG hasn't had to do much of anything and it will take second by default), so Intel has been able to wedge itself in between RTG and Nvidia and contest second place almost immediately. Right now Radeon is trying harder than ever to pretend it's still most like Nvidia (e.g., a name change even closer to Nvidia's nomenclature), but unfortunately for RTG, Intel has 200+ XeSS games, while FSR4 is in absolutely zero (0) games, and it's not even known whether FSR4 will be any good. If Intel can get its drivers sorted, Intel will have consolidated its place as second in the GPU race. RX 9000(?) and UNDA1 need to be decisively good generations if RTG is to have any chance

File: dlss unreal marvel rivals.png (306 KB, 850x1486)

306 KB PNG

>>103608074
Just want to remind /pcbg/ that this OP is an VRAMemer and a DLSS denier. In practical usage, if image quality is held constant, the 4070 is generally as good as a 7900 XTX because of DLSS alone, while if we were to add in RT (no RT in pic related), the 7900 XTX would fall another tier, to the 4060 Ti. DLSS2+ is in probably 500+ games, while Radeon does not have access to high quality upsampling from its own brand, and instead relies on the lower quality XeSS kernel (i.e., not the higher image quality, faster kernel that Arc cards use) to maintain any semblance of image quality in new games.

File: dlss vs fsr 1440p wukong2.png (547 KB, 1110x1606)

547 KB PNG

>>103608074
>>103609119 (me)
Cards like the 7800 XT and 7900 GRE are already too weak for 1440p even without full RT; you'd have to be a moron to buy one of them for muh VRAM

>>103609119
>>103609126
You got that wrong sister

File: dlss vs vram.png (1.34 MB, 1298x1254)

1.34 MB PNG

>>103609137
>another triggered /pcbg/ amdrone
You made your bed: lie in it.

File: unreal_dlss.webm (3.5 MB, 1234x862)

3.5 MB WEBM

>DLSS denier

File: 1734608912564061.webm (1.67 MB, 1272x1080)

1.67 MB WEBM

>>103609146
such quality! better than native!

>>103608288
why do companies hate the number 8?

File: 2023-07-21 at 01-04-47 Is(...).png (526 KB, 1301x867)

526 KB PNG

>>103609146
Prove that's DLSS Quality at 1440p or 4K

>>103609154
Your cherrypicks of AMD sponsored games like Avatar don't matter. A substandard implementation of DLSS doesn't mean that the tech itself is bad
Your cherrypicks will never outweigh the averages

My 2025 dream PC
AMD 9950X
NVIDIA 5090
128GB RAM
4TB SSD
What am I looking at $4000?

>>103609146
>>103609154
>Darn Laggy Super Sampler
kek

File: Capture.png (13 KB, 561x239)

13 KB PNG

Planning on buying a 40 series after the 50 series drops, expecting a price drop. I also got myself a new Odyssey G6 OLED; wondering what else I need to game at the new 1440p.
Also any recommendations to make the form factor smaller? Wouldn't mind transferring this and the new GPU to a smaller case, currently using a Lian Li mid-tower

>>103609187
What do you want to do with your pc? Gaming?

File: latency retards destroyed.png (273 KB, 1209x553)

273 KB PNG

>>103609187
>AMD 9950X
Is that for gaming or some other workload? If it's for gaming, the 9800X3D will be faster, and the dual CCD chips like the R9 have some pretty bad lows sometimes, probably for latency reasons. If you're building a workstation, the 9950X is fine, although of course consider benefits of an Intel iGP for Quicksync

>>103609191
DLSS also has less latency than 'muh native'

>>103609194
The 5600X is still fine for average gaming, but it's getting a bit long in the tooth. The absolute max GPU you should be pairing that with is the 5070. If you're looking at a stronger graphics card, you should do a full rebuild when the 9900X3D is in stock for MSRP

>>103609230
>DLSS also has less latency than 'muh native'
You can get the same result without the ghosting by reducing resolution to 960p

>>103609230
Probably won't be getting the 50 series at those price points, would you still recommend getting a 9900X3D with a 40 Super?

File: 1734541948751089.mp4 (3.79 MB, 1920x1080)

3.79 MB MP4

>DLSS is better than native

>>103609119
>>103609126
is DLSS mostly for 4K? i play on 1440p and DLSS doesn't seem to be recommended on reddit

>>103609230
if i'm gonna stay on B450/5800X3D for 3-4 years should I just get a 4070 Ti Super? I can get one for $800 locally

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

2.51 MB WEBM

>>103609244
First of all, you have to use Nvidia Reflex to get those results (unsupported AMD Anti-Lag 2 need not be mentioned), but secondly the image quality of DLSS (yes, including ghosting) will be far higher than the 960p TAA that you normally run games at with your Radeon card

>>103609255
The 5070 will be better value than the 4070 or 4070 Super. New gens are always better value, if only slightly. You should wait a couple of months for the 5060 Ti if your budget for a graphics card is below $600

>>103609284
That's not DLSS, that's software Lumen, retard

>>103609291
>is DLSS mostly for 4K?
It's for turning whatever resolution you can run into a higher faux resolution. If all you can run is 1080p, using DLSS Quality to turn it into 1440p or DLSS Performance to turn it into 4K will be better than native 1080p. According to Little Timmy at AMDumboxed >>103609176 and my own experience, DLSS Quality targeting 1440p is better than or equivalent to full 1440p in terms of image quality on average; the performance is also much higher than native. Particularly for Unreal Engine games, DLSS Quality at 1440p is usually better than native, because Unreal's TAA is so blurry. And considering that Unreal has by far the most games of any engine coming out in the near future (around 2/3s of all demanding games by my estimation), it's a fact worth noting.

>>103609306
Not right now, no. New gpus are launching next month and in february as well so you'd have a far better gpu for the same amount if you buy 1-2 months later

>>103609323
Reflex is marketing nonsense. You've been able to get the same benefit by capping your fps with rtss for years.

>>103609291
dlss at 1440p is 1707x960, literally less resolution than 1080p 1920x1080.

>>103609119
>fake frames and upscaling is good for you
You're the reason we have shit game optimizations. Kys

>>103609323
would ideally keep the cost for the gpu below 600, yeah.
So it's looking like wait a bit for 5060Ti > 9900X3D > MOBO?

>>103609357
also, the B780 could be very compelling if Intel decides to pressure the 5070 and 8800XT by pricing in the $500-600 range, personally $550 is an ideal price target, this could be very good if the B780 is 16GB when the base 5070 is 12GB per Videocardz rumors, by the time B780 releases Intel will have at least 2 major driver updates

>>103609357
5060ti isn't likely to be released anytime soon as a strategy for nvidia to clear out older gpus. The only leaked gpus so far range from the 5070ti to 5090

>>103608949
>We give you functionality but you can't actually use it
This reminds me Nvidia low/mid cards with raytracing.

File: Screenshot 2024-07-01 at (...).png (858 KB, 1258x1356)

858 KB PNG

>>103609230
>>103609255
>>103609357
>9900X3D
I meant 9800X3D*, sorry. The 9800X3D is single CCD, the 9900X3D is dual CCD, and the cache is only one one CCD, both of which are bad for gaming.

I think if you're getting a 5060 Ti tier card, you can keep the 5600X. Depends on the type of games you play, but for general gaming (some singleplayer, some mulitiplayer), the CPU and GPU paring is just fine. You could probably even get away with the 5070, although there would be a lot more games where you're CPU limited. I'd only upgrade the CPU if you're going for a 5070 Ti or above, although a better CPU would be nice to have even with less graphics cards

>>103609306
I can't tell you about your local pricing, but I would guess that the 5070 Ti will be $800 and be closer to 4080 performance, possibly with new DLSS features. Unless you see a massive flash sale on a current gen Geforce card, probably not worth buying this close to the launch of the next generation (we're literally weeks away)

>>103609339
No, because that by default means that your framerate is lower and therefore latency is higher. Reflex optimizes the rendering buffer for you; you can get close, but there's no way you can manually optimize your system to get low latency of Reflex

>>103609353
You are a retard who doesn't understand how DLSS works, despite having been taught (by me) multiple times in this thread

>>103609354
No, Nvidia is the reason you're not using checkerboard rendering or FSR like conslows

>>103609402
>muh vram
Shut the fuck up, retard. 12 GB vs 16 GB is literally the last concern >>103609126

File: 1730566360379198.webm (3.13 MB, 1280x720)

3.13 MB WEBM

>>103609448
DLSS looks like this

gtx 1660 or arc a380?

File: IMG_3534.jpg (3.67 MB, 4032x3024)

3.67 MB JPG

>>103608415
Ap201 is based, I just really wish it had 140mm bottom fan mounts from the factory, I had to zip tie mine on there
Not that it matters in a case with an opaque side panel but I had to use the little rubber rings that came with an old aio for this to avoid the fans rubbing on the bottom plate.

>>103609448
>Shut the fuck up, retard. 12 GB vs 16 GB
i enjoy a 3080 10GB, that doesn't deny the fact that the market optics are full of VRAM obsessed redditors, learn to distinguish difference between observing something as is vs. supporting it retard

>>103609448
>>103609423
So what's the deal with these intel cards? I'm seeing the B580 is very cheap, and I'm presuming it's an upgrade from 1660S. Even the B780 I'm seeing is supposed to be ~450.

Could be the perfect thing, when the fuck did Nvidia GPUs start being 4 figures to get anything but the bottom tier or 2 gens old?

>planning out my new build
>its ballooning in cost fast

h-haha the 5090 is in my budget

>>103609539
every fucking time.
seems impossible to upgrade anything for less than 1,000

File: Screenshot 2024-12-05 at (...).png (80 KB, 1024x788)

80 KB PNG

>>103609442
Really?
2060 6 GB >>>>> 5700 XT
I wonder if you Radeontards will finally admit that Geforce cards last longer than Radeon cards. It has been obvious to the wise for about four years at this point (DLSS, the best fine wine to ever come to gaming), but now with "Radeon literally can't play games" I think it's an unavoidable conclusion even for you AMDelusionals

>>103609497
Why join the mindless masses? Who fucking cares? Talk about what actually matters instead

>>103609486
At the same price, the 1660. Probably not worth buying anything less than a 2060 6 GB, as it has DLSS and RT.

>>103609501
>what's the deal with these intel cards?
They are interesting, but they're still for beta testers. Drivers aren't quite good enough to recommended to the average retarded gaymer. For example, low test teacher (not going to promote his youtube channel) had problems with 4/11 games he tested in his Arc B580 review. He said that all but one was fixed by drivers during the review period (the remaining problem was the new Indiana Jones, which did not start), but that just shows how pervasive the problems still are. Additionally, there are problems with frame times on Arc in a lot of games. I think Intel has a shot at taking the number two shot in the GPU race, as I wrote here >>103609061

>>103609539
It's reasonable to think about the cost of the graphics card against the cost of the total build. If spending 30% more actually gets you 30% more performance, not a horrible idea to buy faster at that linear price/perf

>>103609551
>>103609563
I have the money to be able to build whatever I want but with each click on partpicker i see a piece of my soul go away

>>103609230
>>103609212
Local machine learning and gaming

File: 1723881945855268.jpg (193 KB, 2048x1152)

193 KB JPG

>Nvidia driver overhead still unfixed after many years

>>103608415
im so torn, i got the sama im01 pro since it was only CAD$60 free shipping, the AP201 was a close second, it seems to be more refined not 100% a fan of the black thick border on top but it does look more refined,. Still almost 2x the cost vs the SAMA IM01 Pro here
Lian Li A3 would be the best imo if it had a metal front don't care if it's solid or mesh it just has to be NOT plastic, dont want cheap wood glued on top of plastic either

Also I bought the mesh version of the SAMA IM01 PRO someone please help me feel better about not paying $20 more for the glass version

>>103609717
>Lian Li A3 would be the best imo if it had a metal front
Bet you can buy it on Etsy or somewhere else. The custom wood panels for this case are selling there since forever.
Personally I got an AP201 because I catched it on sale for the price of an A3 (in my country it costs twice more than A3 usually). A3 was the second option I considered and if I bought it I would probably tried to make my own front panel, first what comes in mind is using 3D printed plastic parts or laser cut metal.
https://youtu.be/A6w-e10h2pI

>>103609717
>Also I bought the mesh version of the SAMA IM01 PRO someone please help me feel better about not paying $20 more for the glass version
Have you noticed how more and more popular cases are moving away from glass side panels, they offer modern alternatives for people who don’t want a flashy pc. There was a span of time where every good conventional case had a window, and it just became the default. But now you have way more choices for modern cases with current tech and features but without a window.
There is no concern in simply not wanting the window option, because it means you’re still looking at good cases. As opposed to a while there where if you didn’t want a window you either had to get an old case design or an expensive case like a fractal define.

File: min-fps-2560-1440 hogwarts.png (54 KB, 560x810)

54 KB PNG

>>103609659
I question those results
Regardless, with a 5600X supposedly performing better than a 7800X3D in minimums, and there being barely any increase in average framerate, he probably fucked up the benchmark pass in some way or another, or it's simply a Zen4 problem, as we know that Zen4 has caused problems in other games such as Cyberpunk

File: Screenshot 2022-01-14 at (...).jpg (315 KB, 1654x737)

315 KB JPG

>>103609659
But since we're talking about driver overhead, Radeon has had fucked up drivers since the launch of DX11 over a decade ago, and they're still fucked up today

Should I get Acer X32Q FS
>32" 3840x2160
>160hz IPS with QD filters
>1152 zone mini-led
Seemingly no reviews other than anecdotes on leddit.

>>103609903
Not worth buying an expensive monitor without thorough reviews. It could have major problems with HDR + VRR for example

File: fuckinnvidiadriverissues.png (204 KB, 789x700)

204 KB PNG

>>103609659
nvidia drivers suck ass dawg

I have never ever noticed or felt a difference better or worse after updating my video drivers

Anyone that claims to notice a difference is lying and people that bitch about drivers are just coping for having shit hardware or are playing an unoptimized game that has to be fixed by the devs of the game, not drivers

>>103610003
I did with intel Arc early on but obviously that’s a different scenario

>>103610003
https://en.wikipedia.org/wiki/Windows_Display_Driver_Model

what driver you get is 100% dependent on what build number your OS is running.

how do you decide what kind of cooling you need? I need to replace my 8y old laptop soon, ill probably build a small matx pc and i want to be able to use it during summer heatwaves

File: Screenshot 2024-09-04 at (...).png (838 KB, 1292x828)

838 KB PNG

>>103609978
>muh steam forum posts
OK, but Radeon can't even display HDR correctly because of Freesync on a lot of the actual HDR monitors available today

>>103610076
Doesn’t matter that much unless you’re playing outside in the sun in Australia or something. The temperature delta of a comfy 70f to a hot 100f wouldn’t impact a cpu cooler much.
CPUs operate at 80c+ without any issues, that’s 176f a difference of +/- 15c depending on ambient temperatures isn’t a concern for the cpu.

>>103610076
the cpu you choose has a known amount of heat output, the case you choose has a known amount of clearance

>>103610110
I don’t know what HDR or freesync do, I don’t use either of them.
Where’s the product for me

>>103610126
>he doesn't use vrr
Not the thread for you buddy

>>103610126
17" tn panel from 2007

>>103610110
that seems to be more of a panel issue than an amd issue. the monitor maker needs to release a firmware update for their freesync panel.

>>103609645
Get the 9950x3d in a month's time then

File: Screenshot 2024-02-16 at (...).png (365 KB, 1303x823)

365 KB PNG

>>103610185
It's a problem with AMD's Freesync Premium Pro, which is what AMD wants monitor manufacturers to use. You can f-f-f-fix the problem by disabling Freesync Premium Pro and using normal Freesync.

>>103610110
>60w PD charger is ESSENTIAL to a monitor
HUB are stupid.

File: 1705600183285.png (37 KB, 638x498)

37 KB PNG

>>103608288
This is why Amd can't compete. They keep wasting resources changing the names of shit that don't need a name change.

>>103608074
Is there a good way to find ATX 3.0 power supplies on pcpartpicker? There doesn't seem to be a filter for it.

>>103608106
b580 might be the most overhyped piece of shit hardware for years

>>103610242
Not hard to list all Seasonic PSUs.

>>103610219
For someone it might be
Its not like they're saying EVERY SINGLE FEATURE HERE IS ESSENTIAL, I guarantee their highest rated monitor doesn't check every box, it just a useful reference for anyone looking for certain features

>>103605188
oh, thats pretty good. i shouldve mentioned that im running zen5 gaming enhancements at level 2, so that likely hurts the score. but yeah, its also at +150 with -40 CO, 1:1 CL28

>>103609194
>expecting a price drop
won't happen
get a malkans sf8g, 4060ti 16gb, and 5700x3d

>>103610204
again, that seems to be more of a panel issue than an amd issue. the monitor maker needs to release a firmware update for their freesync panel. or just stop buying garbage panels who cant update their own display.

>>103610220
>5090 or 9090
>well 9090 is bigger so get that one

Do you guys leave programs like Hwinfo64 & RTSS running all the time?

>>103608949
I can't even run my C4 at full brightness unless I'm watching movies, for desktop and gaming it's at 50%.
Remember that human eyes react to brightness logarithmically, the difference between 50% and 100% is like 4x higher panel wear at 100% over 50%.

decent brand for external SSDs? or any brands to avoid? I own an old WD external HDD and its still working, considering looking for another of those
purpose is a spare archival backup, so HDD could be better, I don't know shelf life of current SSDs
thanks for the advice

File: IMG_3533.jpg (171 KB, 750x546)

171 KB JPG

These are cheap as dirt now, I got one for under $300 new
These have 2tb of ssd storage on them anyway and that’s useful in its own right. And I think I might actually have a use case where having an application and its cache on the gpu may provide me some minor performance benefits.
But what is the driver situation like for these Vega Radeon pro cards?
I know underneath is like a downclocked Vega 64 which is fine performance for my uses. But I ran into a lot of problems with my Vega 56 years ago just not fuckin working right with windows 10.

Anyone using a similar Radeon pro card?

>>103610338
I run the sensors in the background from hwinfo always just to keep an eye on anything weird that may be happening. It barely uses any system resources and lets me track if anything is getting too hot in situations where I may not be paying direct attention.

>>103610338
no, they have performance impact

>>103610347
get a m.2 drive with dram and an enclosure with a known good controller

>>103610338
>RTSS
only for games where I want to monitor shit on the fly, though most of the time I run Special K which is much better with other features
>Hwinfo64
only in my overclocking phases, rest of the time it stays off.
right now I'm running it because I keep an eye on my RAM temps while gaming

>>103610349
Wait, how do applications make use of the 2tb storage for the GPU?

>>103610220
>>103610334
Are we sure the guy making this claim isn't just dyslexic and confused 9700 for 9070

>>103610294
>im running zen5 gaming enhancements at level 2
Is this some form of latency killer thing like MSI boards have? Only level 2 thing I know is from ASUS' Core Tuning.
>+150 with -40 CO, 1:1 CL28
That's awfully close to stock performance for those settings, are you sure that CO is stable and not clock stretching?
Even getting CO -30 is pretty rare. It only works on my CPU because it's binned like garbage and drinks fucktons of VID to push clocks at stock, so the CO merely brings it back to regular performance.

Have any of you guys used fingerprint ID to log in into your desktop, if so, how do you do it?

>>103610390
Supposedly it works with some video editing programs where it allows for real time playback of raw 4k/8k footage while editing because the gpu has its own PCH bus and direct access to four nvme ssds in raid, there’s an LTT video about it from years ago that demonstrates it working.
But it also barely works with anything, specific application support is limited to a select few tools.

However windows just sees it as a raid array and you can do whatever you want with the drive that way. I play a game which is heavily dependent on cache to vram transfer and having the game and its cache on the gpu might result in better render times.
I don’t actually know, it’s all hypothetical, there’s barely anything done with these GPUs because even their intended market didn’t like them much, that’s why you can get $6000 video cards NOS for sub $300
Nobody really took into consideration the potential benefits of having games on that ssd setup because any potential benefit was counteracted by it being a $6000 Vega 64.

I may have fucked up a little bit anons
>building new pc
>parts arrive
>open case
>acrylic window because cheaper than glass and won't break easily, if ever
>remember all those videos of people saving the plastic pull for last and having the static fuck up their rig day 1
>pull plastic off entire case
>proceed to build the rest of pc
>go to grab something beside the acrylic window laying down, end up rubbing it a little bit
>immediately builds a static charge, get shocked touching my toolbox
>gently rub the window again
>another small charge + shock
the correct thing to do is to leave the window off and order a new case, right? All my other pc builds have been completely metal for my personal build, or had a tempered glass window in builds I did for others. Never crossed my mind to look into acrylic generating static.

>>103609717
Can't say about the A3, but I have both the IM01 Pro and the AP201, you're not losing much quality wise, the only thing the 201 has over the IM01 Pro is being slightly longer. Also it could be just mine, but my 201 has horrible fit and finish, the side panel behind the motherboard always sticks out some 2-3mm.
Don't worry about the glass one, some versions have a bold white SAMA lettering on the psu cover which looks like shit. I got the mesh one too because of that.
Also it shouldn't hurt airflow too much staying stock, but if you want you can cut all the extra metal on the bottom and top brackets off, if you'll always have the same fan sizes mounted.

>>103610489
so the pc is completely built and you just are worried about getting shocked when near it?

>>103610511
My ap201, the one above with the 7350k and Titan, seems built pretty well but I will say the internal layout for cable routing is abysmal. Dumping all the cables from the front mounted psu through one hole and then also crossing all the front panel cables back over them is messy.
I think if the psu was flipped and the cables went up the top it would be way more manageable but I see why they don’t do that, they’d need to have a taller base to allow for a bottom mounted power cable to the psu

File: Screenshot 2024-05-26 at (...).png (835 KB, 1303x832)

835 KB PNG

>>103610309
You are confused
Understandable, as you're a Radeon defender and they're usually confused
How many monitors do I have to post?

>>103610242
Filter by 12VHPWR

>>103610338
Definitely not HWiNFO
RTSS is on a lot of the time, though not always

>>103610489
There's a standoff between the mobo and the case for a reason
Anything touching the case is isolated

>>103609448
>No, nvidia is the reason
>defends framegen and upscaling slop
Kys retard

File: Starfield 1080p taa dlss fsr.webm (3.94 MB, 2048x1152)

3.94 MB WEBM

>>103610572
Take your pick
Keep in mind that DLSS not only looks the best, it's also substantially faster than full res TAA.
If you can use full res DLAA, great, but you can't by definition if you don't have a Geforce card

>>103610515
Yes. Everything is on, completely assembled, and fully operational.
>>103610566
>There's a standoff between the mobo and the case for a reason
>Anything touching the case is isolated
I'm still concerned because it's an older model case so it's pretty thin. When I put the window back on it's going to be touching the power cables going into my GPU. So I'm concerned about that as well. Cable is braided so I guess it would be fine but idk man. The whole thing has me spooked a bit.

>>103610613
I don't think you realize I don't give a shit about it and you defending it is why we have shit performance. The more idiots like you we have the shittier performance we have because imbeciles like you keep buying this shit.
Kill yourself you piece of shit

>>103610624
then shocks shouldn't matter, i mean ya its annoying i suppose but it wont endanger your rig

File: input vs output2.jpg (426 KB, 2007x1202)

426 KB JPG

>>103610649
>you defending it is why we have shit performance
Wrong. See my first point: you'd be using checkerboard rendering or be simply stuck on 1080p if not for DLSS

>>103610665
Yeah, you fuck, you praise dlss because optimization is shit and optimization is shit because you praise dlss
Retard

>>103610694
>random /g/ posters are responsible for lazy overpaid dei devs
meds

>>103610649
>>103610694

He is just a symptom of where the industry has gone over the last 10~ years. Nothing will change unless corpo execs fuck off or give power back to people with vision for their games.

File: JPcE.png (21 KB, 1420x259)

21 KB PNG

>>103610454
>Is this some form of latency killer thing like MSI boards have?
not sure exactly, but it came with the latest bios update via agesa. level 2 is the highest level of "optimizing," and i assume its both the reason i can hit -40 in a stable way (havent done deep tests yet but i havent crashed yet either), but also why my cinebench score is kinda low. as for clock stretching, i dont think so. i watched the clocks like a hawk while doing all my cinebench testing even at -40, and they never did any significant dipping. the processor happily churns through cinebench rendering while at clocks like 5340+ mhz, again, likely because of the zen5 optimizations gutting certain things to benefit gaming. i should probably just turn them off really quick just for curiosity, but i dont really feel like spending all day retweaking everything as id likely need to reduce the CO offset, find that new sweet spot between it and PBO +, etc. i guess i also forgot to mention that my x870e mobo has a "pbo enhancement" feature, which goes up to "90C / Level 5" that i also dont know what its really doing. that might also affect my score lol

Why is this thread now always about frame gen?
Ever since this tech came out it’s been these constant arguments over frame gen.
I do not give a fuck about any of that technology, it’s just dick waving “my image is less blurry than your image because brand”. Oh my god, play your fucking video games.

File: 1 fps.jpg (150 KB, 872x820)

150 KB JPG

>>103610742
im tired of dlss

>>103610524
That's the nature of the beast with these types of cases, the IM01 Pro doesn't have much space to tuck the cables at the back either, they also dangle from under the PSU.
I will say though that I found the IM01 Pro a lot easier to build in, the top and bottom fan backets being removable is a godsend, and it supports SFX PSUs natively, if you have one of those for some reason. Honestly I think that unless you really need the extra length of the AP201 like I did, the IM01 Pro is the better buy I'd say.

File: Jedi Survivor PS5.png (895 KB, 1125x633)

895 KB PNG

>>103610694
Often games are built with conslows in mind. PS4 was using checkerboard rendering to massively reduce render costs, and it was ugly as hell. You're always going to have that "optimization" problem. Since the days of the 360/PS3, games have been targeting 30 FPS and had lower resolutions than what the consoles were theoretically capable of. Most conslow games are designed to run with an internal resolution of 720p 30 FPS these days. To run a conslow game at 1080p native resolution, you'd have to have a GPU 4x as fast as what's in the PS5, so about 4x 6700 non-XT. And that's just 1080p.

I don't give a fuck about your ideological bents: right now, practically speaking DLSS and RT sets PCs apart from conslows

>>103610742
Frame Gen has barely been mentioned in this thread. The discussion is mostly about Super Resolution. /pcbg/ has to have this discussion every day because it's filled with retarded luddites who deny reality

>>103610851
>To run a conslow game at 1080p native resolution
1080p 60 FPS*

>>103610566
>Motherboard standoffs (with PSU 20 pin connected)
>7 ground wires on the 20-pins power connector going back to PSU
>PSU mounting screws on metal
>Household/outlet earthing (depends on country)
None of these things prevent direct static discharge from destroying your components, only you getting electrocuted when a live power source touches the case/components. There is no true isolation going on, there's plenty of ground loops.

>>103610707
>consuners have no effect on products
Said every single consumer who's dissatisfied with the product but still consumes it.
Retard
>>103610716
I hate this defeatist mentality. Somehow industry made idiot consumers think it's OK to have shit optimization because now we can have shit looking frames to combat it!
I refuse to buy any GPU with upscaling and framegen and fuck all triple A games. I literally would rather buy older console than support this shit.

>>103610511
it's been my experience in this cases like this that bottom fans are not needed since normally GPUs are close enough to the side/bottom mesh that they can intake on their own, but I am curious about the IM01 PRO if using the expansion bracket on the bottom will help with GPU thermals
I also heard the mesh on the IM01 Pro was a bit more restrictive for airflow compared to the AP201 or A3. Better for dust I guess

File: arch8.jpg (298 KB, 1778x1000)

298 KB JPG

>>103610888
>PSU mounting screws on metal
The PSU's internals are isolated. That's why your PC doesn't kill you when you touch it

>None of these things prevent direct static discharge from destroying your components
Basically you're thinking about arcing? It's possible but pretty unlikely. Wouldn't happen under normal conditions of an assembled PC sitting on your desk, even if you shuffle your stockings along a carpeted floor.

>>103610894
>I refuse to buy any GPU with upscaling and framegen
You'll never buy a new GPU then. Even Radeon has relied on the garbage that is FSR1 since RX 5000, and neither reconstruction nor frame generation is planning to fuck off and die somewhere. Even RTX 7000 has dedicated ML hardware intended for ML upscaling among other things

Do I need to fresh install my OS if I'm upgrading my motherboard? (yes, a board upgrade)

>>103610995
No but you probably should anyway. At the least you'll have to reactivate Windows even if you don't reinstall
>>>/g/fwt

>>103610900
I still like to put fans at the bottom, because not only if I force a positive pressure the intake is always done through filtered fan mounts, which is the only one with a factory fan on both IM01 Pro and AP201, but also it puts some air on graphics cards with fan stop during low loads, helps keep noise down while watching videos or something. No idea about the expansion since I don't use it.
Also, these types of cases love rear intake too, I have mine set up like that with a 120mm magnetic filter and obviously with the cpu cooler fan orientation reversed. I'm using both the rear and the two rearmost bottom fan mounts as intake, and the two forward top mounts as exhaust, since that's where the warm air from the cpu, gpu and psu ends up at.

>>103611075
>factory fan
*factory filter.
fuck.

>>103610894
>refuse to buy any GPU with upscaling and framegen
just say you're retarded, it's ok

>>103608706
if your main use case is gaming, then the Xiaomi is the better choice as it's an IPS while the AOC is VA
also, if you live in Europe it's actually easier to get the Xiaomi cheaper than the AOC (a fact that the aussies at Monitors Unboxed conveniently left out in that review)

>>103608996
>I don't mind paying $600-800 for a good miniLED monitor
that Xiaomi anon was talking about costs 300

Is a 5700X3D worth undervolting? My motherboard doesn't support curve optimizer for it.

I'm really wanting to build one of those tiny ITX boxes but the supply of those short 4060 cards seems to be drying up, think the B580 will get a short ITX variant?

Why are the games in benchmark videos always games nobody plays?

File: image.png (21 KB, 357x533)

21 KB PNG

>>103611316
Just run pbo2 tuner

>>103611338
because all the games that people play run well on a 5 year old budget laptop

>>103611340
Is there a linux version for this?

>>103611316
No

>>103610727
>level 2 is the highest level of "optimizing,"
Just read on it a bit and it doesn't work like that, this is Gigabyte's version of ASUS Core Tuning (same thing).
AGESA 1.2.0.2a changed some parameters that caused a performance regression compared to the previous version, most notably regarding memory latency.
https://www.reddit.com/r/Amd/comments/1h8siwi/asus_intros_core_tuning_config_for_gaming_feature/m0wxcfd/

"Level 2" means that you're running with the same parameters as 1.2.0.2a, for gaming you will most likely want to run "Legacy" which restores the performance to AGESA 1.2.0.2 levels.
See this for the performance comparison : https://www.youtube.com/watch?v=H89xHs9eLmI
You can measure the latency difference in AIDA64's memory "benchmark", the results should consistently be 5 to 7ns lower on Legacy compared to Level 1 and 2.

>as for clock stretching, i dont think so. i watched the clocks like a hawk
You were looking for the "Core Effective Clocks" and not just the "Core Clocks", right? The effective clocks have to closely match the reported core clocks, that's how you confirm clock stretching.

>x870e mobo has a "pbo enhancement" feature, which goes up to "90C / Level 5"
I also have no idea what this is, but I think you should probably disable any kind of motherboard-specific tweaks like these and only use the regular PBO settings .
Motherboard limit, +200Mhz Fmax, low Scalar and whichever CO is stable as your baseline for PBO should definitely improve your results.

>motherboard doesn't support curve optimizer for it.
wait a minute.
how can I check if the mobo I want to buy has that?

Can a b650e run 128 gb at 6000 30?

File: when your gpu lacks vram-1.webm (3.93 MB, 1280x724)

3.93 MB WEBM

>>103609061
This doesn't show the full picture. Remember in Hogwarts Legacy how even with high texture settings, they would still be fucked on low vram Nvidia cards?

>>103609119
>>103609176
>>103609126
GameGPU Nvidia shill needs a life.

File: Screenshot 2024-12-22 142725.png (161 KB, 720x828)

161 KB PNG

>>103611306
>that Xiaomi anon was talking about costs 300
About $370 in the US, which is not cheap for a unit with so many compromises.
Plenty of owners are claiming the local dimming algorithm's underwhelming and worsening blooming in gaming.
The dimming zone is effectively unusable for more general use-cases.

>>103611388
Only an issue with Zen 3 CPUs on very extremely old and cheap B350/B450 mobos, if you're getting Zen 4/5 it's there by default in the AGESA settings.
>how can I check if the mobo I want to buy has that?
Google it, I know most B450s from MSI were updated with it but no idea about other vendors. Just use some software or Ryzen Master to adjust it if the mobo doesn't support CO.

>>103611359
there's CoreCtrl and a "Ryzen-5800x3d-linux-undervolting" python script on github but I have not actually tried neither.

>>103611420
>Ryzen Master
Can't adjust shit on X3D, just monitor sensors

>>103611420
ok good thanks.

>>103611338
Games like cyberpunk are useful because they run awful on any system so you can stress even high end parts . It doesn't matter that no one plays it

>>103611460
Cyberpunk has 700,000 steam reviews and Skyrim numbers of active players at any time.

>>103611420
I own 3 AM4 boards (X470, B550 and X570S) and only one of them supports CO for X3D Zen 3. The issue is that AMD themselves said they shouldn't get it (supposedly cache is particularly sensitive to VCore) and that gave an excuse to board manufacturers to segment the market.

>>103611333
>think the B580 will get a short ITX variant
Maxsun and Acer seems like they'd do something like that.

>>103611469
>Skyrim numbers
And it's just as average.

>>103611469
>Same active number of players as a game from 2011

>>103611333
it's a 190W gpu using 4070 levels of silicon, doubt it

>50 series
DOA

>>103611316
Absolutely. Here's my Cinebench 2024 numbers:
>Stock
Score 744
Clock 3958
Hotspot 90.1
Temp 89.4
Core VID 1.3
Power 104
>-30 magnitude undervolt
Score 802
Clock 4033
Hotspot 83.8
Temp 83.4
Core VID 1.188
Power 88
Using a Noctua L12S in a SFF case.

File: 1713642198830550.png (75 KB, 758x335)

75 KB PNG

>AMD copies Intel's chipset naming scheme
>AMD copies Intel's CPU naming scheme
>AMD copies Nvidia's GPU naming scheme
What is there left for AMD to copy?

>>103611591
nvidia's performance

File: spectre_pro_230_black.png (3.99 MB, 2342x2291)

3.99 MB PNG

>>103608074
Why did 200mm and 230mm fans fall out of favour?

>>103611600
too big, most cases are 160mm wide maximum now so 140 is the standard

>>103609187
Make it 192GB RAM, 2xRTX5090, 4x4TB SSD and 4x20TB HDD.
Just for the lulz.

File: 61ROCI1GK9L._AC_SL1200_.jpg (70 KB, 730x1022)

70 KB JPG

>>103611609
>160mm wide maximum
But why would they do that when most high end air coolers are at 165mm or higher when you have to adjust for ram?
Pic related got 180mm CPU height while only costing 80 bucks.

File: 1729331273940919.jpg (250 KB, 1152x2048)

250 KB JPG

what do people use to test curve optimizer stability?

>>103611631
>But why would they do that when most high end air coolers are at 165mm or higher when you have to adjust for ram?
Jews are trying to condition you to buy AIOs instead because they can upsell normalfags more with RGB pumps and fans and also with screens on the pump

File: 1724418237161543.png (34 KB, 579x483)

34 KB PNG

The cases you order new look like the ones I'm throwing on the trash, kek

Hot air fucking rises on its own, no fans needed. Physics class is important, make sure your kids get a proper physics education.

>>103609496
Mini-anything is never based it's just cute and gay.

File: 1720572289565249.jpg (73 KB, 720x924)

73 KB JPG

And yes I may be laughing over that, but I'm not joking. I literally scrap the worthless cases.

>>103611631
even with 165mm coolers you get a 170-180mm case
an extra 20mm of material costs + the front panel would have to fit 400mm vertically + the need for companies to keep making 200mm fans + the rise of the sff trend

File: comparison.jpg (262 KB, 2408x989)

262 KB JPG

>>103608074


Been a few years since I built my last rig and have not really been paying too much attention to new stuff.

I have seen that there is a ton of debate if it is better to go intel or AMD so I watched a few days worth of youtube vids from fairly reasonable sources and put together a partpicker list of each.

Which would you guys reccomend?

Which will last longer before being obselete?

Last one I built was a i5 with a 1060 which was already 6 months behind THEN, so figured worth asking.

>>103611600
Some smartphones have more RAM and CPU cores than some of those 'rigs' you witness ITT. And those come with 0 fans, Einstain.

Urban Rescue Branch rakes in millions using YT and he cuts all the content on his fucking iPhone. Haha! Bwahahaha!! HA!

File: file.jpg (72 KB, 659x1107)

72 KB JPG

>>103611711
>And those come with 0 fans, Einstain.

>>103611591
Where did the RX 8000s go?

>>103611609
> most cases are 160mm wide
We are talking gaming machines here, not thin clients

>>103611657
There are per-core tests available on utilities like OCCT.

>>103611748
Its some dumb twitter rumour . don't listen too it

>>103611697
Go AMD for better overall efficiency, no point buying Intel 14th gen for new builders.
Arrow Lake is getting the fixes it needs, I would get that instead if you insist on an Intel platform.

File: Screenshot 2023-05-27 220200.png (74 KB, 1179x386)

74 KB PNG

>>103611697
4x DDR5 modules don't play very nicely with XMP/EXPO.
And buy a single set of four DDR5 sticks if you need that much system memory, these sets have been validated to run those specs with all four memory slots populated.
https://pcpartpicker.com/product/HKn9TW/corsair-vengeance-128-gb-4-x-32-gb-ddr5-5600-cl40-memory-cmk128gx5m4b5600c40
https://pcpartpicker.com/product/BXP8TW/corsair-vengeance-192-gb-4-x-48-gb-ddr5-5200-cl38-memory-cmk192gx5m4b5200c38

File: arrow.jpg (52 KB, 1578x273)

52 KB JPG

>>103611785

ok.... so, this one?

I see it is 20 core vs 24 ... and similar output.

why would you advise this one over the i9?

just price or is there more to it?

I can swap my boot drive sata ssd from my old pc to my new one without issues right? Windows will just work?

>>103611851

ah... ok.... i was wondering about that.

I noticed that the set of 4 was a lower speed and cost more.

Thought I was winning by finding pairs of higher speed cheaper... but that makes more sense.

Give me my 5090 leather jacket man

>>103611851

infact.... that is probably what happened with my current computer too now that i think about it.

I started with 2x8, then added a second pair of same brand and speed months later.

It worked like a beast for a few months, then one pair stopped working and was locking the machine up. I had to remove the pair for it to boot and function.

File: aDQb3YCEMpSJdkCtbcUd9B.png (181 KB, 2188x1812)

181 KB PNG

>>103611407
So what you're saying is that you'd rather have performance go into the toilet when you haven't configured your settings correctly. Regardless, in most of these games, you won't see a difference in presentation on the Geforce card. It's just better optimizing its use of memory.

wait I thought AMD was done trying to compete? why am I seeing shit about 9000 series like 9070 etc?

>>103611899
there may be some issues but it will probably work

File: Recco1.jpg (139 KB, 1255x905)

139 KB JPG

>>103611785
>>103611851

updated build with the suggestions....

this look better?

File: 192GB_6000CL32.jpg (239 KB, 1001x967)

239 KB JPG

>>103611851
> 128GB ddr5-5600-cl40
> 192GB ddr5-5200-cl38
This is what noobs have to make do with.
Just get 2 identical kits of 2x48GB, dial in the settings manually et voila! 192GB DDR5-6000CL32!
Inb4 but this requires the will to do some research and patience.

File: ohnonono.png (100 KB, 881x1503)

100 KB PNG

>>103612050
Better vram than native, KEK

>>103612069
what does a 70 imply they are competing with

>>103612081

so you are saying I could in theory get the other two working again on the old rig?

is this an adjustment I can make BEFORE putting them back in, because if i try to boot with them in it will not boot?

>>103612069
They're trying to leech off of Nvidia's branding
>RX 9070 XT
vs
>RTX 5070
And hoping normalfags don't see AMD Radeon in the name and think they're getting an Nvidia card

>>103612123
midrange market

I WANT MY 5090 AND I WANT IT NOOOOOWWWW

File: 1626185535184.png (364 KB, 738x1187)

364 KB PNG

>>103611460
Cyberpunk 2077 was and is massively popular, especially considering the Phantom Liberty expansion

>>103612121
User error

File: K9nhzXsoY2x6jtwW2RYoY6.png (188 KB, 2244x1790)

188 KB PNG

>>103612121
If you even tickle Radeon's VRAM it crumbles.

File: power-multithread.png (81 KB, 560x1210)

81 KB PNG

>>103611893
Intel 12/13/14th gen CPUs aren't very power efficient because they are made on inferior 10nm nodes compared to Ryzen 7950X's more power efficient 5nm compute chips.
https://www.techpowerup.com/review/intel-core-ultra-7-265k/24.html

>>103612081
I'd get a ThermalRight Phantom Spirit or Scythe Mugen 6 dual fan CPU cooler instead of the CoolerMaster 212, both of these will work on the new Intel LGA 1851 socket.
https://pcpartpicker.com/product/zkWJ7P/thermalright-phantom-spirit-6617-cfm-cpu-cooler-ps120
https://pcpartpicker.com/product/rV9wrH/scythe-mugen-6-dual-fan-6029-cfm-cpu-cooler-scmg-6000dbe
https://www.thermalright.com/product/phantom-spirit-120/
https://www.scytheus.com/mugen6dbe

File: DLSSbutMissingVRAM.webm (1.34 MB, 1280x720)

1.34 MB WEBM

>>103612157
>>103612186
> user error
Nah, it’s insufficient hardware for the resolution

File: DtkmeTCuyRPtSVFfSEdRu4.png (180 KB, 2076x1812)

180 KB PNG

>>103612227
User error. Regardless, better hope your VRAM doesn't get tickled
Like for like (GB for GB) it seems Radeon needs to give its cards more VRAM because of its poor memory management. What do you think would be fair, Radeon needs 50% more VRAM? So 8 GB on Geforce = 12 GB on Radeon

File: nvliarVram.jpg (21 KB, 387x168)

21 KB JPG

>>103612186

>>103611600
I just bought one with 200mm intakes
Super comfy having them spin at low rpm

>>103612050
>Regardless, in most of these games, you won't see a difference in presentation on the Geforce card.
I'm saying I don't know if that's true.

File: DLSSandFGstolen.webm (3.49 MB, 1280x720)

3.49 MB WEBM

>>103612260
> User error
No such thing, just insufficient hardware, KEK

>>103612216


Ok... so.... the ultra 7 is almost as good as the i9 and is more power efficient.... and cheaper.

The ryzen is better than both but costs a little more.


the only advantage to going intel is the ability to add more ram , using 4x48 instead of being capped at 4x32... which brings the cost back up to about even with the ryzen set-up.

I honestly expected the i9 to be the best option out there, but this has me squinting now.

File: 3RNqkVzbrG3wmi3g9TCYd7.png (169 KB, 2142x1664)

169 KB PNG

>>103612322
Complain all you want, but at least the 8 GB Geforce card is usable in most cases, even with all that fancy Nvidia tech (unrivaled, which is why you see Nvidia competing with Nvidia in your webm), while if you had an 8 GB Radeon it would probably be completely crippled even before trying Frame Generation (which it doesn't have access to at all in HZD)

>>103612322
oh my god both those pictures are the exact same shit who gives a shit

>>103612076
Small adjustments https://pcpartpicker.com/list/WNR78Q

SN770 needs a firmware update before using it on newer W11 versions.
I'd just pay the extra $20 for a 1TB SN850X for peace of mind.
https://www.windowslatest.com/2024/10/15/windows-11-24h2-causing-bsods-microsoft-will-reportedly-halt-update-for-more-pcs/
MSI Gaming motherboard isn't better than the Tomahawk models, which is discounted atm.
ASUS Dual video cards have bigger & better cooler than dual-fan Ventus cards.
Super Flower 850W ATX 3.x units should be better than the MSI.

If you want to try manually tuning two packs of 2x 48GB kits like >>103612081 you can try.
But you'll have to be comfortable with tuning the memory controller & memory module-related voltages, as well as adjusting primary memory timings at the very least.

File: 1688029954545666.jpg (104 KB, 775x719)

104 KB JPG

Hardware Buster's PSU picks updated for 2025!
https://hwbusters.com/best_picks/best-atxv3-pcie5-ready-psus-picks-hardware-busters/

File: nvliarX2.png (335 KB, 1024x769)

335 KB PNG

>>103612372
> but at least the 8 GB Geforce card is usable
> in most cases
KEK, even here he struggles to cope
> which is why you see Nvidia competing with Nvidia in your webm
Nah, it just shows that the same chip has different performance for some reason, I wonder why

>>103612358
Yes they're basically neck-and-neck performance wise. You should really be deciding based on software specifics or platform considerations. AMD gives you an in-socket upgrade path while Intel can theoretically push 2DPC memory configs to higher speeds. Also Quicksync if you use it.

Gonna buy an Intel B580 or B770 soon... should I get an Intel CPU too? Currently have an AMD (AM4). I've heard Intel GPUs run more efficiently at idle when combined with an Intel CPU.

File: Untitled.png (1.2 MB, 1280x833)

1.2 MB PNG

>>103611075
>SAMA IM01 PRO
I have a very stupid idea
I can make it happen because the mesh and the glass panel are CAD$50 and CAD$60 each
Total of CAD$110 is still cheaper than a lot of cases, and only slightly more expensive that a single A3 or AP201. The Lian Li A3 is CAD$99 here, and the ASUS AP201 is CAD$99 too (actually upwards of CAD$120 off amazon, it's just $99 from a local store on sale

only problem is I waited too fucking long now only the white IM01 PROs are in stock now, the black ones sold out already FUCKKKK
so if I go for the IM01 Pro I have to be extra patient and wait for sales on white GPUs specifcially (I just missed a sale for the white merc310 7900XT for CAD$750 basically around US$520)
Also have to wait for a sale on a white PSU, thankfully the white tax on PSUs are usualyl just like $10-$20 more, sometimes none depending on model and availability

File: T3qDDk2Lu4fgF9XpGs3J65.png (114 KB, 1920x1441)

114 KB PNG

>>103612375
He's saying that the 16 GB card is faster, which it is under those exact conditions. It also costs more, food for thought

>>103612449
Those aren't the most recent Tom's benchmarks. All of these benchmarks are coming out the recent A580 review, and your chart doesn't have the A580. Looks to me like even Intel has better memory management in its first gen Arc.

>>103612460
Even with a B770 I wouldn't think you'd need to move off AM4 as long as you have at least an R5 5600

File: Screenshot 2024-12-14 at (...).jpg (1.16 MB, 2560x1440)

1.16 MB JPG

>>103612554
>A580 review, and your chart doesn't have the A580
B580*

File: file.png (408 KB, 1085x791)

408 KB PNG

good or meme?

File: VRAMbutDLSSorFGblocked.webm (3.49 MB, 1280x720)

3.49 MB WEBM

>>103612554
> Those aren't the most recent Tom's benchmarks
> December 6, 2024
KEK, Deep Learning Super Coping

>>103612585
good but depends what you do

>>103612426
>>103612459


awesome.... one more question.....

Every review of the 4060 TI 16gb I see says it is awesome IF you can't afford a 4070.

and yet... every 4070 review I have seen says that the gddr6x sucks and you want gddr6 with no x.

Every 4070 I look at has the X? lol and also only 12 gb.


is the 4060 ti better than the 4070 with the GDDR6X?

and IS there a 4070 that is JUST GDDR6?


I know I want something in that range, but which is the best option?

>>103612616
they recently released a gddr6 4070 and performance is bascially the same
wouldn't touch a 4070 though you'll be vram cucked and it'll feel bad not being able to use the full potential of the card

>>103611591
Nvidia copied ATi's (now owned by AMD) naming scheme first.

>>103612585
meme
a 990 pro is half the cost and real world use you will never notice a difference unless you are doing VERY specific tasks

>>103612608
Music production and gaming casually.
(I just realized that my MB supports gen5)

>>103611697
>AMD build has AMD GPU
>Intel build has Nvidia GPU
Stop going full retard, it's AMD CPU + Nvidia GPU.

>>103612460
>Intel CPU
don't touch intel cpus even for free

>>103612616
4060ti 16GB could be better than 4070 12GB in some scenarios at 1440p since 12GB is primarily for 1080p right now, but I would buy 4070TiS 16GB or 7900GRE if I was going for 1440p instead

>>103612636
intresting

>>103611591
real answer, intel/nvidia own ~80-90% of the market, so consumers know how to interpret their schemes

coming up with their own thing would just confuse a certain number of of consumers they can't afford to lose and there's little to no downside of just using the popular thing

>>103612652

Honestly... did not think it was an option *shrugs*

why would I do this ?

>>103612674
amd have the best cpus
nvidia have the best gpus
it's that simple

File: Capture.jpg (240 KB, 2304x1477)

240 KB JPG

>>103612585
is it faster?
yes, of course

is the extra money worth spending over 1 second of loading time? not to me

>>103612659
Amd has had their current scheme for long enough for normies to catch on. If this 9070 meme is actually real, it's retarded. I hope it's a troll

>>103612656

I see...

Monitor is limited to 1080p unless I hook up my TV ( which I have considered )

So beyond resolution , not really too much difference then?

>>103612709
1080p you can get away with something cheap and shit like a 3060

File: Screenshot 2024-12-22 163902.png (35 KB, 552x395)

35 KB PNG

>>103612616
>and yet... every 4070 review I have seen says that the gddr6x sucks and you want gddr6 with no x.
>Every 4070 I look at has the X? lol and also only 12 gb.
GDDR6X is the better chip, it's custom-designed by Micron for NVIDIA only.
GDDR6X just runs a bit hotter but it's not a problem with RTX 40 series because video card manufacturers learned how to cool it with the RTX 30 series products.

4070/Super has a stronger GPU chip. but new 5070 cards will launch in the next 1~3 months.
It's not a bad idea to wait for the newer cards unless you just want something right now.
I would wait for the new cards, and save a bit more money until then.

>>103612707
>for normies
but what about the drooling retards?
they have money too, and AMD literally needs every dollar

File: 1080p-high.png (106 KB, 1342x1090)

106 KB PNG

>>103612595
>Intel Arc B580 review
>By Jarred Walton December 13, 2024
If you think the 16 GB 4060 Ti is worth another $75 or $100 over the 8 GB version to use FG in HZD, it's your money

>>103612616
>is the 4060 ti better than the 4070 with the GDDR6X?
No

>which is the best option?
4070 is substantially more powerful than a 4060 Ti. The 16 GB 4060 Ti isn't usually recommended for gaming because it doesn't make sense next to the 4070.

>>103612656
Maybe in some manufactured scenario

>>103612691

interesting concept.

NGL.... first person I have ever heard suggest this. I assumed you had to pair AMD stuff as a mean to make thier stuff proprietary like apple.

>>103612741
they have 0 relation
you may see some things like 'Smart Access Memory' for AMD GPU/CPU combos but it's all marketing, you can do the same with an Nvidia GPU

>>103612734
4060 Ti 16GB is the cheapest, smallest, and most efficient GPU worth recommending
There's an argument for the 7700 XT but I don't have faith in that GPU long term

>>103612741
amd doesn't have the install base to force proprietary shit like that
their technology like FSR and freesync are open for anyone to use with any hardware unlike how nvidia handles their stuff

>>103612793
4060Ti isn't cheap enough, it's a $350 card at best (the 16GB version).

>>103612808
>cheapest
out of the gpus worth recommending, it's the cheapest

>>103612724
if Radeon is resorting to "they'll think it's an Nvidia card" to get by maybe they really are on the brink of collapse

>>103612820
Still not cheap enough, especially at the tail end of the product generation. End of story.

>>103612832
never said it was cheap

>>103612741
>NGL.... first person I have ever heard suggest this
It's literally the meta right now and has been for couple years now.

>>103612720
>>103612734


OK..... so the 4070 actually is better than the 4060 TI. I really only NEED a 3070 for 1080p, but may as well get something as new as possible if reasonable.

If the new hotness comes out soon and these go down then should deffo get the 4070 over the 4060ti.

*noted*

File: Screenshot 2024-12-07 114946.png (416 KB, 631x1059)

416 KB PNG

>>103612854
Both NVIDIA and AMD have been winding down the production of current-generation video cards.
Do not expect more discounts on the 4070 & 4070 Super.

>>103612829
it's not that at all
it's knowing that a XX70 is mid class

kinda like a bmw 3 series and audi A3 are the same 'class'

it makes it easier for people to glance at a product and set expectations more than anything
nvidia has the ability to change shit around at their leisure because they are the market leaders and people look to them to set the standard right now

>>103612854
a 3060 is enough for 1080p
4060 Ti 16GB would be a better buy than the 3070
I would go 1440p though

>>103612448
The best is the 700 watts fanless one from Seasonic right now. No other one (fanless or with a fan, doesn't matter) comes close in terms of efficiency.

Which allows them to charge ridiculous prices. 260 euros in germany today

File: Screenshot 2024-12-22 164727.png (91 KB, 1203x441)

91 KB PNG

Just made a new build with a SN850x. Are these speeds in line with what it should be doing? The first row for sequential write is in line with the reported 7,300 read / 6,600 write on the product page, but I'm wondering about the other rows. I understand those tests are supposed to be "harder" I just don't know if the speeds I'm getting are what should be expected.

>>103612876
>4050
have people finally started calling the 4060 a 4050 and 4060 Ti a 4060?

File: VRAMbutMissing1PrecentLMAO.webm (2.06 MB, 1280x720)

2.06 MB WEBM

>>103612734
> >By Jarred Walton December 13, 2024
7 days, that’s recent, KEK
>If you think the 100 fps version is worth another $75 over the 30 fps version
Yes, the difference is literally between usable and unusable here, lmao

>>103611413
370$ for all those features is still quite difficult to come across bellow 450
and if anon is indeed living in Europe then he won't find the AOC with any significant price differences to the Xiaomi
on this last black friday I was in a similiar position to anon between the AOC and the Xiaomi, in my country of Spain the cheapest offer for the AOC was 320€ while the Xiami was 310€
and if anon is also looking for a gaming monitor, then the Xiaomi an extra edge due to the simple fact it's IPS and not VA
because let's face it, for gaming VA is shit
until microLED becomes a thing and magically solves every issue, the only sensible choices for gaming monitors are either IPS or, if you have the money for it, OLED
>local dimming
it's more than good enough for gaming and watching media, I can attest to that
the small blooming effect around the cursor is only visible outide of those use cases, but honestly why would anyone even want it activated outside of those?

>>103612880
Maybe I am the drooling retard here because I had no idea there was this sort of association lol

>>103612854
>If the new hotness comes out soon and these go down
how long until people finally realise that this a thing from the past?

File: performance-rt-1920-1080 (...).png (49 KB, 560x690)

49 KB PNG

>>103612794
>FSR and freesync are open for anyone to use
Funny to me that there's never been a Vulkan game shipped with FSR according to MachineGames

>>103612793
There are arguments for the base 4060 8 GB and the base 4070 over either variant of 4060 Ti. In fact, the gen's almost over, so here's a rating of the RTX 40 series cards for gaming:

4070 > 4070 Super > 4060 > 4090 > 4070 Ti > 4080 Super > 4060 Ti 8 GB > 4060 Ti 16 GB > 4080

>>103612854
At this point, you want to wait for next gen cards if you have the budget for the 4070. The 5070 will be out likely within a month (or shortly thereafter)

>>103612911
Again, you're showing user error settings. The RT in REm4ke is so bad that it damages the image quality because of its low resolution. It's a technical abomination, and far worse than the RT in previous REmake titles that actually solved problems with SSR. But let's see what TPU has to say on the topic.

What have I been telling you: tickle Radeon's VRAM and it becomes worthless

No Headphones thread currently up so I thought i might ask here since its one of the most commonly used headsets. Anyone got the Logitech G Pro X or any other Logitech headset for that matter, and knows if you need to have the Logitech G Hub software running all the time in the background for the equalizer settings to work? I tweaked some settings and it immediately sounded better but i noticed after restarting my PC its back to bad quality again and only after i start G Hub it instantly gets better. Do I really have to have that run in the background all the time now or is there a way to save the settings?

>>103612922
really?
you haven't noticed nvidia puts out:
low end X060
mid X070
high X080
top X090
and then refreshes with super/TI
theyve done it since the 900 series like 10 years ago

File: VRAMvsDLSS.webm (2.31 MB, 1280x720)

2.31 MB WEBM

>>103612958
>resident evil 4 RT
>3070 8gb case
OH NO NO NO
The same issue was happening on the 3070 as well, KEK https://youtu.be/3IBKRnOxoRY?feature=shared&t=46
> you're showing user error settings
Nah, I’m showing insufficient VRAM size, user error would cripple both of them
> The RT in REm4ke is so bad that it damages the image quality
In every game pretty much, I agree

>>103610759
>100ms
Kek

>>103603581
low color gamut check srgb% rating, bad brightness (nits) and bad contrast
all result in bad image

>>103612990
yeah I think I'm retarded

B780 when? C980 when? I want those mid to high end graphics cards (for video decoding and encoding) Shilltell.

File: Screenshot 2022-07-18 at (...).png (286 KB, 1227x452)

286 KB PNG

>>103613050
>cyberpunk rt overdrive on low end cards
>more user error settings
There's no way that a card that's too weak for 1080p Quality DLSS (66 FPS with Frame Gen, in other words about 30 FPS without) should be even trying for 1440p Balanced. Of course if you throw manufactured examples in, I guess you'll always be able to find broken cards

Wasn't that the basic principle relating to turning off RT and DLSS so that Radeon could look competitive in bar charts, not broken?

>>103612069
Amd can definitely compete with nu xx80 class cards which are super cut down should be named xx70, it's just xx90 is far out of reach

>>103613103
Are the Arc cards actually that good about video encoding or is this just a buzzword intel shills like to throw out?

File: Ah.jpg (197 KB, 2560x1440)

197 KB JPG

5090 is how much?!

File: p1.jpg (158 KB, 1280x1828)

158 KB JPG

>>103608415
>>103609717
>>103609496
>>103608602
I couldn't find the "perfect" case and ended up going with the antec performance 1 silent
seems like the best combination of solid/mesh side panel, no rgb, good airflow, and fits just about everything I want.

anyone tried it? I don't really feel like I can trust reviews since they seem to have done a ton of sponsored videos for it.

>>103613214
I was looking at maybe a new case and I ended up thinking I would get the Asus ProArt PA602. Cost is more than yours though.

File: Cyberbug2077NoReflections(...).webm (2.36 MB, 1280x720)

2.36 MB WEBM

>>103613115
> >more user error settings
> should be even trying for 1440p Balanced
Using RT on an RTX GPU is user error? Based
>rtx 3070 crash with other 8GB gpus forgotten
Outtickled I guess, KEK
> relating to turning off RT and DLSS so that Radeon could look competitive
I mean, RT in cyberpunk is pretty similar or at times worse compared to raster

>>103613234
I heavily considered that case, as well, but it doesn't have solid side panel, the price isn't great as you mentioned, and I saw thermal testing somewhere that made me figure it wasn't worth the price.

It does look super nice though

File: Cyberpunk 2077 RT Overdri(...).png (1.97 MB, 2560x1440)

1.97 MB PNG

>>103613239
RTX Overdrive is only usable with the 4080 and 4090, not for a 3070 or 4060. That's not because of the VRAM obviously: that because of the performance of the core. Who cares about these user error settings? I'd love to see a Radeon card even try to use PT. Get out the slide projector!

File: 1717954809104607.jpg (71 KB, 1025x1043)

71 KB JPG

>>103613257
>but it doesn't have solid side panel
Yes it does.

>>103613191
AV1 is fucking amazing, there's a reason why there's a big shift from Nvidia from HEVC to AV1 because it smokes the other encoders. Better quality for lower file sizes and adjustable bitrate that still looks good on lower bitrates. And Intel has AV1 under the bag when it comes to their GPUs. Even the Arc A770 provides a better performance in video encoding than the RTX 4090 all across all types of bitrates and resolutions which is absolutely nuts.

>>103613214
be Quiet Silent Base 802 is an alternative, it has a solid side panel variant and it ships with two separate front panels (closed & meshed).
https://www.bequiet.com/en/case/2049
https://www.techpowerup.com/review/be-quiet-silent-base-802-atx-chassis/2.html
Cost & availability is the issue.
https://www.avadirect.com/Silent-Base-802-No-PSU-E-ATX-Black-Mid-Tower-Case/Product/13745842

>5080 comes out first
Why? Who the fuck cares about a discount 4090?

>>103613312
>A770 provides a better performance in video encoding than the RTX 4090 all across all types of bitrates and resolutions
This sounds fake as fuck and I don't believe it

>>103613318
be aware if you get that 802 you gotta replace the fans if you care about cooling, the included fans are good for quiet operation but don't run at more than 900rpm.

>>103613191
That's the Alchemist cards' claim for the 1 minute of fame they got. It's the sole reason why people even bought the A380, and A580 cards. They're budget ways of getting access go AV1.

File: DLSSbutItKilledByVRAM.webm (3.82 MB, 1280x720)

3.82 MB WEBM

>>103613276
>paid for an rtx gpu
> is only usable with the 4080 and 4090
Based
> Who cares about these user error settings?
Who cares about good FPS?
KEK
> I'd love to see a Radeon card even try to use PT
Radeon can just use raster and get the same quality

>>103613341
There's gonna be a lot of people replacing 30 series cards (it's been 4+ years after all) and I would guess a lot of WAIT people finally moving from 20 series.

File: 1705044786863246.webm (2.13 MB, 1280x720)

2.13 MB WEBM

>>103608288
based antisemitic micro devices.

File: 1693659672992479.png (766 KB, 1247x1200)

766 KB PNG

>>103613374
>dlss performance at 4K
I guess is you want to run 4K, you should have sprung for the 16 GB version in that case There's still no competition from Radeon in that case. Don't even try to bring up FSR in Performance mode after destroying your image quality, the Radeon will just crash because its VRAM can't handle a little tickling

>>103613391
They're skipping the 88 series
That makes AMD DEI

The new Radeon GPU seems to be on part with the 7800XT in shader units etc, I wonder how much of a generational improvement they're going to bring over it.

File: 1712533227045642.jpg (21 KB, 369x407)

21 KB JPG

>>103613437
>They're skipping the 88 series
yeah, missed opportunity, njudea shill.

File: DLSSbutYouArentAllowedToFly.webm (307 KB, 994x762)

307 KB WEBM

>>103609323
At first I was
> Performance to turn it into 4K will be better than native 1080p.
> the performance is also much higher than native.
>>103613437
But then I saw DLSS Performance at 4k and then
> I guess is you want to run 4K, you should have sprung for the 16 GB version
KEK
>pic
Already got debunked

>>103613437
I want a full comparison of DLSS and FSR with XeSS

Is the 14900 worth it over a 13900

>>103613437
>calls them amdunboxed
>uses their data from three years ago because it agrees with your shilling

>biggest slight against AMD is them supposedly copying Nvidia's naming scheme
Honestly by this time we usually get "leaks" about all the "problems" AMD GPUs are going to have so this being the biggest problem is kinda giving me hope for them matching the 4080 like rumors say.

>>103613503
they are literally the same cpu, both defunct if you are a gaymer and not disabling 80% of the cpu and locking it at 1.3v though

File: Screenshot 2023-04-03 at (...).jpg (247 KB, 994x668)

247 KB JPG

>>103613495
>look mom, i cherry picked it again!
Why are you complaining about the 4060 Ti 8 GB not being able to run 4K Performance DLSS >>103613374
and then immediately backtracking about wanting to run that feature? Performance is generally a lesser option, but Quality at 4K is usually better than or equivalent to 4K """native""", while being faster >>103609176

>>103613523
It's amazing that AMDumboxed admits something but then refuses to apply it to their benchmarking methodology. Why are they now starting to test with upsampling? Because RTG might actually be competitive with FSR4, five years after DLSS2. Why are they going to start testing with RT? Because RDNA4 is supposed to have a relatively larger jump in performance for RT over raster

>>103613526
>kinda giving me hope
Same, 9070 XT might be the 5070 Ti false equivalent. Certainly possible because I think AMD will be able hit hit 4080 false equivalence with their best card this gen, which would likely be close to the 5070 Ti. The names would line up, and it shows they think they can compete in the future with very similar naming and therefore at a very similar tier

>>103613611
>AMD Shillware has to literally post a disclaimer because of how much shilling they do for AMD
lmao

>>103613347
Performance as in frames decoded? I think Intel loses without a more powerful GPU than the A770. For video quality? Intel Quicksync and Nvidia NVENC usually trades blows, and that is also true for AV1.
https://youtu.be/TYOkJFOL5jY?t=660
I think Intel still has a slight edge in AV1 because other players came in late and didn't start integrating it until after Intel did with Alchemist. I expect AMD and Nvidia to both start pulling ahead given the above video with their dedicated Alveo card and being able to put that on a GPU and Nvidia to pour resources into improving it to maintain their lead. Given that Battlemage didn't tout any improvements on that front and they only added VVC to Lunar Lake, Intel is probably not as focused right now as tackling other downsides of their GPUs.

>>103613341
Because they made the mistake of selling 4090 and 4080 together and no one bought 4080, so this time they will try scam people and get some sales

File: dlssbutyouarentallowedtoc(...).webm (803 KB, 1222x1078)

803 KB WEBM

>>103613611
> >look mom, i cherry picked it again
KEK
> Why are you complaining about the 4060 Ti 8 GB not being able to run 4K Performance DLSS
1080p internal
> and then immediately backtracking about wanting to run that feature?
I never said that, I enjoy the best image quality- native without any temporal AA or upscale
> Quality at 4K is usually better than or equivalent to 4K """native"""
KEK

>>103613664
They already announced that the 5090 is for "professionals" so I fully expect them to price it at $2k+ and for the 5080 to take the top gaming GPU spot at previous top gaming GPU price, $1k

File: performance-3840-2160 sta(...).png (99 KB, 560x1490)

99 KB PNG

>>103613674
>I never said that
That's literally the setup of the webm you posted in #103613374. Most of your webms are crying about some RTX exclusive feature you can't use

>I enjoy the best image quality- native without any temporal AA or upscale
Sure you do