>>103040270
>progressively goes from most optimized games to unreal 5

File: 1729782098404741.jpg (38 KB, 613x677)

38 KB JPG

>9800x3d
>$479
>not including tax and shipping
sisters... you said it would be under $450, what happened?

File: 0_E1eNateTiDThGcYI.jpg (180 KB, 1056x594)

180 KB JPG

>>103040309

>>103040309
Intel shat themselves on the street

>>103040309
inb4 the shills call you a poorfag for not wanting to pay that much

>>103040290
they're doing the opposite of cherry picking by showing a spectrum of both extremes, i don't see how anyone could complain about that.

Also 8% is not bad. We need to get used to the idea of incremental improvements someday, the massive leaps can't go on forever. As long as we can consistently get >5% i will be satisfied. Obviously this isn't worth upgrading to if you have a 7800x3d, why should it be? CPU's are meant to be used for several generations at least; sometimes much longer (lots of 5800x3d chads are going to be holding on to it like the 2500k chads of old)

File: intLOL.jpg (64 KB, 709x412)

64 KB JPG

>>103040309
Oh no, 30 dollars more for the best processor you can buy -_-
Meanwhile at intel:

File: widertree.png (55 KB, 288x96)

55 KB PNG

>>103040270
>>GPU
>1080p: RTX 4070 Super, RX 7700 XT
lol. lmao.

>>103040339
>the massive leaps can't go on forever
There has to be an incentive. If your competition is shit, you can get away with incremental improvements

>>103040365
I'm not paying $500+ after tax for an overshilled cpu, if you want to waste your money that's fine

>>103040309
they already forgot zen 3

>>103040375
it's also not possible to just make massive leaps forever through iterative design. AMD have gotten a lot of mileage out of the framework laid down by Keller and his team. We are also nearing the limits of silicon.
>If your competition is shit, you can get away with incremental improvements
Intel finally had competition with Ryzen and all they were able to manage over a period of 7 (seven) years was one (1) actually impressive generational improvement, alder lake, it's just not that easy.
And don't say AMD are incremental in the same way Intel was incremental in the Quad Core Purgatory Era. 0-3% improvements is just sad. 5-10% we should be happy.

cpus wont be getting much faster
technology has hit a permanent wall fellas

>>103040465
If amd can consistently deliver 5-10% it means you can upgrade your cpu on a 6-8 years cycle as god intended and enjoy huge gains every time. I only fear stagnation if they fail at that and turn into post-sandy bridge Intel.

>>103040493
they will do the exact same thing intel did if they have no competition

do not ever for a second think that corporations are your friends

How’s the noise on the p12 max? I need to get a pair of case fans for my case, I’m considering if I should get the p12 max or just splurge on 2 noctuas for $60

>>103040503
I don't, but also it's in their best interest to deliver improvements that encourage upgrading even with no competition. That means settling into a sustainable cadence of gains if possible. People disappointed with 8% must be the kind of mentally ill people upgrading every single generation. if you're considering zen 5 it should be a lot bigger than 8% for you.

>>103040365
Intel processors are for adults who are too busy having sex and smoking crystal meth to care about benchmarks or stupid virgin nerd shit like cost to performance. What even is a 7zip compression? I use winrar because my penis goes in and out of vaginas weekly. Intel is the brand name associated with being normal and having sex with women, AMD conjures images of neckbeards screaming at 14 year old Russians on Rust.

>>103040309
I said it'd be 500, since listings popped up for 480-520$
Sorry you're retarded

>>103040270
> Lian Li O11 Air Mini
This seems to be the best case in its size. Is there anything else I should consider?

>>103040368
12GB VRAM, weak ass bus

Building. Budget $1500-$2000CAD, would prefer lower end. What I have right now:
>CPU
Ryzen 5 7600 6-core/12-thread 3.8GHz [Turbo 5.1GHz]

>Cooler
Corsair iCUE H100i 240mm Liquid CPU Cooling System w/ Copper Cold Plate

>Motherboard
MSI PRO B650-P WIFI6E

>RAM
32GB (16GBx2) DDR5/6000MHz Dual Channel Memory CORSAIR

>SSD
1TB WD Blue SN580 Series (PCIe Gen4)

>GPU
RTX 4060 or 4070 Super, undecided

>CASE
undecided

>PSU
800 Watts - Standard 80 Plus Gold Certified Power Supply

>MONITOR
undecided

>OS
Windows 11

>CASE FANS
+3 basic fans

>>103040640
wait for 50xx to be announced in ces
buy 2tb nvme
you should be able to get a x3d cpu for 2k build

File: 4k_cpu_bottleneck_4090_as(...).png (697 KB, 1920x1080)

697 KB PNG

>>103040465
CPU progress was alright this year, just not for gaming. AMD was prioritizing servers, and Intel thermals.
Still, CPUs have been advancing slower than high-end GPUs. A few more generations like these, and 4k CPU bottlenecks will become much more of a thing than they're now.

File: FB_IMG_1689583760046.jpg (52 KB, 915x681)

52 KB JPG

>>103040652
>wait
Classic pcbg

File: 1701920062762793.png (30 KB, 1620x152)

30 KB PNG

At what point is it false advertising?
>Compares 109,500TBW to 700TBW of other SSDs
>TBW of the gigashit SSD is based on arbitrary "AI workloads" moniker while everyone else uses JEDEC standards for accessing TBW

>>103040652
Why wait? For all we know the new cards will be more expensive and require a bigger PSU, besides the 70 series is expected to arrive in february. There are also rumors of NVIDIA wanting to discontinue the 40 series asap as well, with a big sale happening this black friday.

>>103040270
source of the graph?

>>103040640
4060 is ass. Get a super, an AMD card, a used card or wait for 50 series

By the way, if you can, check out canadacomputers or Canadiancomputers or canadaelectronics I forget what they're called
They offer cpu+ram+mobo bundles and when I checked it was actually like a hundred bucks cheaper than buying them separately from the cheapest Canada stores
Make sure your PSU is good, everything you said about it is basically irrelevant

File: 1385799350491.gif (832 KB, 500x283)

832 KB GIF

>>103040493
God intended for us to get 20% every generation. shill

fuck this new spam filter having to wait 10mins just to shit post. this website is cooked bruh on god fr fr

File: 1669466690844143.png (494 KB, 607x604)

494 KB PNG

>finally built an ITX system
>PC crashes while installing windows
>Apparently its a known problem with Core Ultra 265k

How the fuck can something like this even happen? I literally cant even install fucking Windows.

File: blur.png (408 KB, 1000x871)

408 KB PNG

>>103040584
>my penis goes in and out of vaginas weekly
Tell us more.
What's it like?

>>103040671
>2 years into a gen and new gpus 1 quarter out with improved efficiency, rt and whatever software features they came up with
>lmao You wana wait?!?!

File: 1633305214082.png (163 KB, 512x468)

163 KB PNG

>>103040290
Flu-Kong is UE5

>>103040738
Turn off cookies, then only phone posters struggle (like me when I'm in bath like rn)

muh cpu bottlenecks
muh gpu bottlenecks
being a ram bottlenecked faggot (i multibox in eve online) is the worst feeling yet
>build a pc in 2020 with 128gb of ram running at 3600mhz on lga 1200
>add 13 more accounts to my already 21 and now need more cpu power and ram
>it's been like 5 fucking generations of cpus and an entire new memory standard which even has time to matur appears in the meantime so im sure it'll be fine!
>best ram options on ddr5
>...128gb, maybe 192gb ddr5
>......at 3600 mhz (same as my ddr4)
>with significantly more stability and performance issues compared to 4 sticks of ddr4 running at the EXACT same speeds on my much older tech for some reason
im going to kill myself holy shit
I don't want to go threadripper my budget for a new pc is only $3500.

>>103040758
I guess being linear forest with sparse npc's being scattered at the end of a tunnel makes it easier on the cpu

>>103040745
>Core Ultra 265k
>He bought.

>>103040536
>if you're considering zen 5 it should be a lot bigger than 8% for you
What if I were to go from a 1600? I hope it's going to be worth it since $479 is pricey for an 8 core in current year even if it's the best for gaming.

>>103040745
https://www.youtube.com/watch?v=6QoCCFXD0xc

He tried to warn you

>>103040745
Next time don't be a mindless consumer blindly trusting a failing company that has made 2 literally broken generations before this one which was widely reported on and check out reviews before buying something. Idk. This post is probably just bait nobody can have their head this buried in the sand and also be posting here.

File: 1730267962849398.jpg (198 KB, 1200x1192)

198 KB JPG

my 3060 12 gb i got for 250 € years ago dips below 90 fps sometimes in skyrim vr on my vive pro at 2000x200 even after performance tweaks.
what's a sensible upgrade considering the price of my current card

>>103040768
If you ran 128GB DDR5 at 3600Mhz it would be stable
The issues come in when your trying to run 128gb of DDR5 at 6000Mhz.

The high speeds get lost in translation when people talk about DDR5 instability at high capacities. Slamming even 192GB in a system will be fine, just don't expect to enable XMP.

>>103040784
The gain coming from a 1600 would be astronomical yes

>>103040784
>$479
cpus always go way below msrp, only gpus do the opposite. it'll be 300 after a couple months.

>>103040796
a 5070

File: file.png (279 KB, 1590x1478)

279 KB PNG

r8 my build and suggest if anything amiss

>>103040798
>The high speeds get lost in translation when people talk about DDR5 instability at high capacities. Slamming even 192GB in a system will be fine, just don't expect to enable XMP.
Except cpus are far more performant sensitive to ram speeds than they were for the era of cpu im currently using thanks to AMD's FUCKYUCLK. I would also be going for a dual ccd naturally as i need more than 8 cores for this and the reduced ifabric speed will hurt significantly
also
>Slamming even 192GB in a system will be fine
My friend who also multiboxes at my scale has a 7950x, his motherboard supposedly supported it but when he tried to run 192gb even without xmp/expo it was unstable as shit no matter what he tried. Found out it's because Asus's 4-dimm support is garbage. Nothing is guaranteed with 4 sticks on this dog shit ddr5 platform it seems.
Am I gonna have the issue if i get something like an MSI Mag x870 and a 9950x3d? Maybe not, I hear it's a pretty Asus specific problem, but I'll still have pretty degraded performance only being able to run at 3600 or 4000 or whatever when the cpu wants ram running at 6000-6400, an issue you didn't have on my current cpu because 3600 was fine for comet lake.
IDK it's all just so tiresome, AMD is so bad at memory and Intel is just so bad at making a working cpu now.

>>103040793
Most of the problems are thanks to Windows retarded scheduler. Pretty sure once thats fixed its a perfectly fine midrange CPU. Just sucks that right now I cant even use my PC at all.

>>103040844
And the fact they had two generations (13th and 14th) just beforehand which are literal ticking time bombs that oxidize themselves after 3-6 months of load? Is that Window's fault too? Intel have never made a mistake is that what you're saying? You get what you deserve to being a mindless consumer who doesn't read reviews. ESPECIALLY after their recent fuck-ups.

>>103040833
nice b8

>>103040874
It's not bait, anon. Genuinely asking. Moving from a shitty Macbook to a real PC.

>>103040844
intel literally had to strongarm OS developers into writing custom scheduler for their retarded e-core/p-core bullshit just so they could increase the peak power without looking bad on power consumptuon

if they actually had an architecture that could compete with zen they wouldn't have to do this

>>103040833
Good bait, but some improvements for next time would be to add small HDDs (like 1tb) and a very cheap case but a ton of non PWM case fans

>>103040754
NTA, but my cock is weird, I can't feel much until I go into that primal monkey mode where I'm about to cum, where you're basically dragging your cock from a tight and wet hole. Shame I don't enjoy it more honestly.

>5060 is a rebranded 4060 Ti 8GB
KEK

>>103040784
If you're still using 1600, could just get a 5700x3d and call it a day. The uplift would still be hueg.
Depending on your GPU, 9800x3d may not be necessary. Unless you somehow bought a 4090 to go with that 1600, are waiting for 5090, or maybe wanna play shooters at the most ridiculous FPS available.

>>103040789
youtube has been showing me his vid in the recommended section but I never clicked but now that you've posted it I really like this guy.

>>103040290
Unreal can be optimized if you know what you are doing and remember the origins of the engine

File: 1712673042889229.png (1.51 MB, 1500x721)

1.51 MB PNG

anybody ever use these? or are they just a meme

>>103040886
What is your use case?

>>103040931
No need to waste money on making new entry-level dies that make sub-60% margin.

Should I hold off til Black Friday? Halloween sales seem decent enough

>>103040756
The super refreshes are not even a year old and they bring some decent value over the originals, as I mentioned chances are all of the new cards will require a new PSU for the workload and will be at least $100+ during launch and scalping season. You're asking anon to wait for at least half a year, I know you guys might not feel the passage of time but life is short and the clock is ticking.

>>103040768
Wait (I know I know) for CUDIMM RAM which is supposed to fix stability issues and be super fast, supposedly a corsair vengeance one comes out this november but it might be expensive.

>>103040798
3600 on DDR5 is much slower than 3600 on DDR4
DDR5 memory controllers literally can't go to CL18 or CL16, the registers don't hold numbers that small

File: 1708310647255230.png (30 KB, 200x193)

30 KB PNG

>>103040745
mine installed fine

>>103041002
Developer. Have a lot of shit running in my dev environment. I cannot endure lag and a sluggish PC.

>>103040999
All benchmarks showed that its just as good as termal paste (sometimes even slightly better). You pay way more for basically a sidegrade because it isnt a fucking mess should you ever need to remove your cooler. Also no degradation over the years, so actually I would say it might even be an upgrade. Cleaner and no maintenance needed. Just more expensive and slightly riskier as its conductive, so watch out when installing your cooler and doublecheck that it fits.

give it be honest.

is it still okay to buy an intel chip? I'm doing video editing 4k. VR. gaming etc.

>>103040465
all gaymers want is fatter cache
meanwhile intel gives us 36MB for $600

>>103041245
It's like buying an amd gpu when nvidia cards exist

>>103041245
id rather buy an nvidia arm cpu

is newegg trustworthy?

>>103041245
12th gen and under is ok but that's old

who are the retards still buying nvidia gpus with all these driver issues worse than amd lately

>>103041305
>Intel Core i7-12700K

these are on sale $299 CAD.

are the new intel chips out?

>>103041360
yes but they're horrible for the price and who knows what problems are lurking like with 13th/14th gen
just go AMD

>>103041360
the 12700k is faster than the U9 285k in most games anon. Intel is losing the plot.

>>103041295
It's big enough to still be there next year, and it has a physical headquarters that an enraged customer could sent a lawsuit and/or drive to to hack them to death with an axe (in minecraft), that's about as trustworthy as it gets.

>>103040999
1 word: air bubbles

File: PTM.png (1.63 MB, 790x1831)

1.63 MB PNG

>>103040999
graphite pads are a meme. phase change material pads are legit. just the science behind the latter is so cool.

File: 299.jpg (247 KB, 1920x1080)

247 KB JPG

there's no reason why the sram wouldn't have 200MB if it's +2x as big
https://www.youtube.com/watch?v=ZTMBqNbHUrs

>>103041377
>>103041378

what a better AMD choice then the intel 12700k ?

>>103041471
5800x3d and beyond

>>103040838
Fwiw asus' 600 series am5 boards have fucked memory traces so that's not unexpected
Their x870 boards are good as are msi's and gigabyte's

>>103041460
Staggered for thermals and/or done on less dense node and/or fully functioning ones get binned to server CPUs?

when is amd getting cudimm

>>103040290
most of the unreal gains are probably just from 24h2 too since its shipping a later dx agility.

>>103041508
Zen6

>>103041215
Only developer stuff? Get 9950X then

>>103041525
>9950X
I-I don't trust AMD. Never used AMD shit in my whole life.

>>103041537
you can still get a 285k just know that it's slower in games at 1080p. if you play at 1440p or 4k it matters significantly less

File: file.png (218 KB, 1299x901)

218 KB PNG

>>103041537
ogey

>>103041537
I'm not sure where you've been this whole time but AMD is the default choice for reliability now. Especially after the 13th and 14th crashing fiasco and very recently, the core ultra 200 launch instability problems

>>103041537
>I-I don't trust AMD
Then you will have to enjoy your 3 gen old Intel chip, or the housefire Intel chip that electrocutes itself, or the Early Access Intel chip that you will be beta testing for a year before it gets enough patches to stop crashing.

>>103041557
>>103041549
Thanks, anons. Let me check it. My current build is from around 2013. 3750k, 840 pro, 660ti. Can you use Nvidia GPUs with AMD?

File: 1720226634889222.jpg (102 KB, 1252x971)

102 KB JPG

>>103041549
>>103041557
My 14900k is fine tho

>>103041582
>Can you use Nvidia GPUs with AMD?
obviously yes

File: file.png (481 KB, 1920x1080)

481 KB PNG

>>103041583
>My 14900k is fin-ACK

>>103041583
prove it
run the OCCT CPU+RAM test for an hour and post the result

I am extremely power conscious, between let's say the Ryzen 4500 and the i3 12100f, which one should have lower idle consumption?
Under load they seem to be about the same, maybe the 4500 is a bit lower as far as I have seen.

>>103041633
4500 will definitely be lower if you don't turn on XMP
with XMP on the motherboard might do some questionable things with the infinity fabric

>>103041583
Not all (>99%) are fine though

>>103040365
imagine actually buying a 245k

>>103041688
This, even Intcels own 14600K would be a better deal.

>>103041245
The newest chip they released is good in theory for all of these things, but I'd wait until the clusterfuck clears out and it becomes more stable.

AMD has the 9950x which can compete with the 14900k and 285k but it's expensive, apparently the new X3D chips won't come with a productivity penalty so I'd wait for the 9900X3D lineup if you wanna go AMD.

File: 18489563_1458706537485008(...).png (1.01 MB, 1920x1080)

1.01 MB PNG

>>103041583
You uploaded the wrong picture, you butt hurt chump!

Also, you're a filthy liar who doesn't provide proof to back up your claims!

I'm so glad that I'm an AMD Ryzen/NVIDIA master race member!

I really hope the 5090 is only 1699...

>>103041740
>only 1699
never combine these two terms again

>>103041748
dont reply to me again ramesh

>>103041754
shut up ramen

>>103041728
thankyou for the reply

is it okay to just grab a 12700k then? they are on sale in CAD.

>>103040666
Isn't resolution more of a GPU thing? 4k is just not a viable resolution for consumers anymore unless you're like watching movies on a monitor.

>>103041737
>NVIDIA
>master race
nvidia is just intel for GPUs
absolute kikes overcharging for shit

i like my 295k

>>103041779
Yes BUT Nvidia actually develops new tech, Intel are lazy as shit.

https://videocardz.com/newz/amd-ryzen-7-9800x3d-now-official-oc-support-20-faster-than-core-ultra-9-285k
$480 for the 9800X3D
thoughts?

>>103041779
>turd worldy doesn't understand capitalism
many such cases

>>103041781
>295k
nigga has access to unreleased CPUs

>>103041768
E-core scheduling can fuck things up if your program loses focus where all p-cores would go to sleep and only the 4 slowass e-cores are working on the program so you'll have to keep that in mind

>>103041809
capitalism is mutually exclusive with innovation
otherwise NVIDIA would be the first on every node, but they aren't
they place all their eggs into architecture and wait for a node to be mature so yields are higher

>>103041809
>it's okay they're kikes because CAPITALISM
ok kike

>>103041833
if you want nvidia to lower prices
you should be mad at amd for not competing
simple as

>>103041843
AMD competes & beats nvidia

>>103041843
uh oh amdrones arent gonna like this one

>>103041779
NVIDIA is actually competent and develops good tech that usually doesn't blow up in your face, they're "overpriced" when compared to less jewed times but their features are very good for both gaming (DLSS is objectively better than FSR) and development with AI and similar machine learning stuff.

>>103041833
Capitalism provides the financial incentives to innovate, that's the AI boom in a nutshell.
The financial incentives to be on the latest node aren't high enough for NVIDIA.

>>103041843
AMD could release a 4080-tier GPU for 200 bucks and still no one would buy it, NVIDIA brain rot is too strong.

>>103041888
uh huh totally, just like everyone is running out to buy the 285k that is actually shit

make a good product and the people will buy it
dont make a good product and they won't
simple
as

>>103040270
Upgradefag here. Bought a ryzen 5700x to upgrade from my 1600af, 2 tb nvme drive, and 32 gig ram kid (2x16). Stupid thoughtt crossed my mind to use the new and old ram combined to get to a total of 48 gig across the sticks. I did make sure clock and CL match, though I was forced into different brands due to my existing not being available anymore.

My question is, combining them it is said you can lose dual channel memory or get PARTIAL dual channel where you get dualchannel until 8 or 16 gig is occujpied, then single channel the rest of the way. Has anyone experimented with this. My board is a b450 for reference.

Last note about ram, gamersnexus had an interesting benchmark video where he found a really weird performance anomally on zen3, where 32gb setup as 4x8 did BETTER than 2x8 for 16 gig. His tests are not memory bound either, but there is a noticeable increase in performance across his tests. Supposedly 2x16 is even faster than that.

In short, do I just do the 2x16 and maybe sell my old 2x8? Can I do both by putting 16 gig on slots 2 and 4 for the second banks to have 32 on second and 16 on first, or should I configure to have 24 gig across BOTH banks to make a total 48 and remain dual channel across the board? How will this impact performance of the cpu?

>>103041900
>just like everyone is running out to buy the 285k that is actually shit
Correct, it is selling out everywhere.
You just proved my point.

>>103041900
>dont make a good product and they won't
weird that so many people bought 40 series gpus before super then

File: 1720000233872882.png (225 KB, 2271x1330)

225 KB PNG

>>103041737
According to the tards on here I have to wait an hour to "prove" it which I don't have atm but it's otherwise been perfect

>>103041900
For CPUs? Sure but that's not how it works in the GPU market where brand loyalty is everything

>>103041912
a paperlaunch =/= good sales
their stock market prices are also continuing to plummet

>>103041917
>it which I don't have atm
Concession accepted, buy AMD next time.

>>103041800
not really winning the best value for money award when 5700x3d exist

>>103041912
https://www.tomshardware.com/pc-components/cpus/intel-hasnt-sold-a-single-arrow-lake-cpu-at-germanys-largest-retailer-core-ultra-200s-sales-stagnate-after-just-one-week

you were saying

>>103041875
>Capitalism provides the financial incentives to innovate
only as long as there is the threat of someone else taking money away for themselves in the form of a competitor, though this isn't always a guarantee of innovation as everyone could play it safe and get lazy to maximize profits as we see with gpus
mobos have also done some profit margin discovery and have hiked prices accordingly
>the financial incentives to be on the latest node aren't high enough for NVIDIA.
it's more because they are willing to let others take the risk and problems first, as ATI did for example
>>103041912
nintendo strategy where they have limited supply on launch

My new (but very old) monitor should be 144Hz but in Windows it's detected as 143.999, and it's annoying because some games running in fullscreen cause a long blackscreen when quitting. Is there an easy fix?

>>103041998
No and it has nothing to do with the hz number

File: file.png (3.82 MB, 1922x1119)

3.82 MB PNG

Intel should release an e-core only version of ARL

>>103041917
lol i respect you for actually going through with it
i'm praying for that 14900K

>>103042005
What is it, then? If I overclock the display to 145Hz the problem goes away, but I'd rather not do that because it probably fucks VRR.

File: 1720647844374599.png (227 KB, 2271x1288)

227 KB PNG

>>103041949
>>103042016
I could be here for an hour, depends
I need to repaste it seems

9800X3D is good?

>>103042065
It's a flop

so no dual 3d cache on the 9800/9950x3d?

Thinking of getting a RTX 4070 super, which brands should I get/avoid?

>>103042146
EVGA

>>103042146
EVGA

>>103042146
BFG

>>103041819
>>103041819

so im going to be editing, 4k gaming and VR.

will this fuck the performance?

also if I'm rendering. will this cripple it?

File: 1714415239696276.png (268 KB, 347x494)

268 KB PNG

>managed to fit just about everything onto my 14tb drive with 300-400gb to spare after pruning old stuff from the other drives
1) what's a good drive cloning utility
2) what are the best value SSDs these days for gaymz libraries
3) are there any SLC SSDs being sold these days

So I'm considering an OLED monitor. I use my pc maybe about 5 hours a day on average. Assuming I don't have it at full brightness, how long should the monitor last?

>>103041981
>only as long as there is the threat of someone else taking money away for themselves in the form of a competitor, though
Not at all, Apple didn't need to design their own chips for example.
They did it to put themselves in a position to make even more money.

>>103040821
> cpus always go way below msrp
Ever heard of the problem of induction?

>>103042173
Editing - maybe
Gaming and vr - nope

>>103042175
1) macrium reflect free trial is great
2) not sure
3) nope, SLC hasn't been available for years. But you could technically buy a WD Black SSD (or other brands with aggressive HMB) and only partition 1/3 of the drives total capacity and the firmware will use HMB and effectively give you SLC performance and durability

>>103041215
> cannot endure lag and a sluggish PC
In this case you need more than just one drive.
Install dedicated drives for different processes that run concurrently and access the fs.
Storage has always been the nastiest of bottlenecks.

>>103041537
> I don't trust AMD
If you trust Intel you trust the silicon jew.

>KOORUI 24E3
anyone here own this?

>>103042302
I'd like to know wtf you could be doing to actully saturate a 7GB/s NVMe

File: 1705523352724928.png (251 KB, 2272x1376)

251 KB PNG

>>103042056
I guess I did have an hour

>>103042065
I'm sure it will be, it's basically a Zen5 7800x3d with improved thermals and thus clocks.
Also the best CPU to be released in 2024, cause the others were kinda mid (for gaming).

It is a shame that ARL suck ass, Z890 has a shit ton of IO

File: norefunds.png (101 KB, 920x511)

101 KB PNG

>>103041583
no refunds

>>103041917
>5GHz
that's not how you test for instability

Will the 3090 become a 1080p GPU when the 5000 series launches?

>>103042496
thats not how it works but we already know you want to argue in bad faith

>>103042396
17W idle for a 24 core "space heater"? that's not bad.
meanwhile my 8 core >>103035899

>>103042366
Imagine downloading some torrent in the background, converting images/videos and playing some game (that continuously reads data from storage ) to kill the waiting time. You will experience stutter if all processes access the same drive no matter what it's theoretical troughput might be (which is only achieved if a single process has exclusive access). All file system operations, in particular searches will be much faster if they have exclusive access.

>>103041432
Phase change is a meme.
>costs a premium
>pain in the ass to work with
>insignificant performance improvement over regular paste

>>103042534
1080 Ti was a 2160p GPU when it launched, now it is a low end 1080p GPU. The 3060/3070 were considered 1440p capable when they launched, now they are also low end 1080p. 4070's are on that cusp.

>>103042065
It will bring back overclocking on the menu and the childish fun we once had with it.
If 7800X3D is good, 9800X3D will be even better!

will desktop computing see silicon carbide and gallium nitride chips

C'mon Intel get your shit together, how come you have these "Stability issues" with your iGPU

Fuck, I'll have to wait until Panther Lake or something, I should've listened to /g/ and avoid jews

>>103042151
>>103042157
EVGA is no longer on the menu you trolls.

>>103042604
I'm waiting for diamond chips and diamond coolers.

>>103042577
t. paste maker

File: white build.jpg (81 KB, 663x624)

81 KB JPG

Thinking of doing a full huwhite build. Will I love it more because it's huwhite?

File: boost-clock-analysis.png (60 KB, 939x802)

60 KB PNG

>>103042396
good
now post that image when someone tells you your 14900K isn't stable
>>103042441
it's normal, the E-cores drag down the average

>>103042441
Anon said run OCCT CPU+RAM and that's what it runs at
I run MC with some very extreme mods, has no issue lighting up the favored P-cores to 6ghz. If you ain't crashing with modded MC your not crashing period.
My VID is 1.404V at 6GHz so the degrade meme is all moot.

File: boost-clock-analysis.png (60 KB, 939x802)

60 KB PNG

>>103042672
wrong image lmao

CONSOOOOOOOOOOOOOOM

i will pair my 9800x3d with my 3070

>>103042648
>all these black parts.
is it even possible to make a actuall FULL white build?

>>103042175
forget about cloning, use freefilesync and just mirror the drive
also crucial stuff is cheap and good

>>103040270
>only 1% increase in cyberpunk
what kind of fuck ass game did cd projkect make tf

>>103042741
enjoy your 20 fps at higher resolutions

I have an opportunity to buy a used 3090 for around ~$500. It was mined with for about a year and the guy said he undervolted the card. Would it be worth to pick it up say versus something like a 4070? I mostly do Homelab/AI shit with some gaming on the side.

>>103042800
surely i'll get at LEAST 30fps increase since i'd be upgrading to a am5 build and from a 5600x

>>103042803
well depending on how serious you are about work stuff a 5090 is gonna be much better

>>103042812
after 5600x nearly all games become gpu bottlenecked at 1440p

at most you'll notice fewer stutters in games like silent hill remake but the average fps stays the same

>>103042746
>all these black parts.
the only black part in that photo is the motherboard and yes, white motherboards do exist

>>103042818
I mean a 3090 would be a sizable boost for me (I'm currently on a 3060), and I upgraded my machine with the hopes an opportunity like this would happen (bought a good CPU, PSU, tons of RAM, but transferred my old GPU over). Even though a 5090 would be cool, I'd rather not spend a used-car's worth of money on a GPU.

okay whys amd highest end card only 800? whats the catch

>>103042791
A game that's GPU bound instead of CPU, more than likely.

>>103042877
You don't need a car if you never leave your mom's basement.

>>103042803
no if you game
>I mostly do Homelab/AI shit with some gaming on the side
then yes absolutely

>>103042877
4070 has a bit too little vram for proper ai stuff, but we dont know if theres 24gb 5070 around the corner as was rumoured with 8x3gb modules and that might be value king in few months

niggers tongue my anus, but GN has a pretty interesting vid up about nextgen X3D

B860 motherboards when?

>>103042917
>nvidia
>vram
lol

>>103042621
Just today I bought a "diamond coated" non-stick frying pan from Crofton™, 28cm in diameter for 26,99. High-tech in my kitchen!

How can I tell if my CPU supports AMD IOMMU V2?
Ryzen 3 2200G

File: topkek.jpg (32 KB, 450x300)

32 KB JPG

>>103042917
> 24gb 5070

>>103042917
>>103042948
Yeah, I'm not confident that nvidia will deliver anything worth the time for the mid-range market. I'd rather just get what I can now.

>>103042908
I'll have you know, I microwave my tendies and honey mussy in my own house like a big boy.

how bad would a 5600 bottle neck a 7900xt

>>103042955
Reading a bit about it, it may only be available for EPYC (Rome,) but that's about all I could find, so the question stands.

Is the thermal paste that came with my cpu fan likely to be any good?

>>103042955
Do you have it enabled in BIOS?

>>103043035
obsessing about thermal paste is a sign of autism
you either use the preinstalled paste or PTM so you never need to replace it again

>>103043046
just regular IOMMU.
not sure seems like this V2 stuff is actually "hardware-assisted IOMMU" for new server-grade hardware

File: starcraft-brood-war.gif (696 KB, 220x124)

696 KB GIF

How many monitors can I fit on wall mounts until it gets stupid?
I was thinking either:
>32:9 main monitor
>2 4k monitors over the 32:9
>1 monitor set in portrait to scroll the internet or chats
or
>Just do all of it in 4k but with like 5 monitors

File: 1534282956377.gif (688 KB, 286x310)

688 KB GIF

>$30 more for pretty much the same perf
how long til amd jews stop selling 5700X3Ds and 7800X3Ds?

>>103040738
>God intended for us to get 20% every generation
We're literally nearing the area where your loliporn will get fucked by quantum effects. Silicon is over.

>>103043113
you now remember during zen 1 when amd was the wholesome chungus reddit friendly brand

where do you guys sell old parts?

>>103043191
i dont

>>103043191
on tori.fi

>>103043113
> pretty much the same perf
It'll be 30% faster in productive apps.

>>103043191
the craigslist equivalent of my country

7950x3d keeps getting more expensive... At least the 9900x is getting cheaper, if it drops another 50-100 it might be an attractive proposition.

>>103042917
>if theres 18gb 5070
Fixed, and that if we are lucky, knowing jewidia they will probably release it with 128-bit 12GB

If you are just going to run non-overlock with a stock HSF is there really anything worth buying over Intel i5-13400F ? Single core speed really seems to fall off at this price point.

>>103042746
yes, you just have to go intel or get a better amd mobo

>>103043191
ebay/ leddit

>>103043191
nowhere ever since they started requiring covid pass for that sort of shit
so it just collects dust in my closet now

so if i just use my shit to game, the 7800x3d and 9800x3d are basically the same, right?

>>103043352
its basically the same until you play a hyper specific game that benefits from x3d like factorio or pso2.

>>103043352
for now, depends on the game
I imagine something insanely CPU limited like flight simulator will benefit a lot more than gpu limited raytracing games like cyberpunk and avatar

>>103043368
wait but i play pso2

>>103043352
yes, if you can get a 7800x3d for sub $400. then go for it

>>103043385
you're gpu bound mostly in combat especially at 4k. but if you want to hit that 180fps client cap in town then the fastest cpu is the way to go.

>>103043352
maybe in the future a console comes out that has a 12 core x3d amd cpu and games will utilize that shit more

File: Screenshot 2024-10-31 at (...).png (1.53 MB, 1307x842)

1.53 MB PNG

>>103040270
Mass replier here, the Sheriff and Prophet of /pcbg/. It's possible that the RX 8000 series might have a quality and value relatively close to Geforce in the price classes in which it releases RX 8000. (To any delusionals reading this: the previous sentence coming to pass would be an improvement over the current situation.)

How do I surmise this? The answer is AMDumboxed coverage. AMDumboxed doesn't seriously acknowledge any Radeon flaws until there's some fix in the works (e.g., RDNA1 issues and DX11 driver overhead issues, both of which received patches that partially mitigated the problem, neither of which AMDumboxed acknowledged until a "fix" was on the way that they could also report on, read: shill). AMDumboxed certainly doesn't test in a way that would show Geforce as having the massive lead in image quality and performance that it does, i.e., they have historically tested with RT and DLSS disabled. This would obviously be antithetical to AMDumboxed's agenda, as they wouldn't be able to shill Radeon to their rabid audience.

Recently, AMDumboxed started GPU testing with upsampling enabled (in advance of FSR4), and they're currently about to add an RT suite / more focus on RT to their graphics card reviews (in advance of an improvement in Radeon RT performance), likely because they're privy to solid insider information. The midrange might actually be competitive this gen (and no, Radeon is not currently competitive in any price class except the $200 class), but who knows whether pricing will be good enough to make the upcoming gen one of those unequivocal "buy" generations. Maybe with Intel in the midrange mix as well, prices will end up depressed from the excessive MSRPs that these corporations would like to charge because consumers don't have any other options. I feel it's unlikely that the Geforce cards will end up with massive price cuts, but especially the Radeon RX 8000 cards _after price cuts_ could be very attractive.

would it be dumb to buy a used psu

>>103043273
A-at least you'll have DDR7!

>>103043408
if it's one of those with long warranty not really

>>103043408
a new ones price hovers around 100€ and will prob last a decade so
as far as pc parts annual expense goes theres little reason to buy a used one

Does lga1700 or AM4 bave better idles

>>103043441
my 5700x3d idles at 30w

>>103043403
you have autism go take a shower

File: 1715272212951984.png (236 KB, 456x592)

236 KB PNG

>>103043433
oh shit part warranty still works if you aren't the one who bought it?

>>103043455
yes, just be sure to get the original receipt and warranty insert along with it

>>103043441
my old 5600 idled at 8-9w

>>103041642
Isn't XMP turned on basically standard for 99% of systems? Why would it affect the 4500s idle consumption more than the 12100f?

>>103043403
why do they always show something that would have been easily done with a planar reflection?

notice how there's never anything like the bean in chicago that actually requires raytracing

is it abnormal to have to bump your RAM's voltage up to 1.4v from 1.35? 6000CL30 on Zen 4

>>103043113
All consumer X3D are Epyc rejects.
how long til amd jews stop selling 7003X Series?

>>103043681
>CPU's are binned
what a revelation bro

>>103043352
Not same but close enough to not feel the urge to upgrade if you have a 78.

File: 1729917048143477.jpg (68 KB, 592x1144)

68 KB JPG

So apparently you can build a brand new computer with a ryzen 5900x and an rx 6800 for a thousand dollars and change... It's going to last 10 years easily, regardless of your use case. Why haven't you done that yet anon?
>b-but it's 2 gens behind i need to consoom
That gen was the last good gen and we're not getting any good and accessible hardware from AMD anyways because of the intel disaster

Im on a zen 4 motherboard rn, gonna upgrade from 5800x3d to 9800x3d, whats a decent midrange motherboard for this?

>>103043654
its safe up to 1.45v.

>>103043621
>Isn't XMP turned on basically standard for 99% of systems
most people don't even know the option exists
>Why would it affect the 4500s idle consumption more than the 12100f
you can't mess with the system agent voltage on non-K intel CPUs
on AMD motherboard manufacturers always pump the voltage up to get XMP stable through brute force

>>103043745
based, thanks anon. wasn't seeing stability with 1.35v and was worried 1.4v might be hurting the memory

>>103043739
because i only game so i prefer x3d and amd gpus crash on a game i play

>>103043739
>It's going to last 10 years easily,
>building "brand new" on AM4 in 2024
accurate image

>on a ryzen 5 3600
Can I wait a bit getting a 5700x3d or are the prices going up like the 7800x3d

>>103043793
just get one from aliexpress

>>103043747
https://hattedsquirrel.net/2020/12/power-consumption-of-ryzen-5000-series-cpus/
How accurate is this? From what I understand this would also apply to 4000 series CPUs. Although I am not sure at all if XMP turned off would affect it.

>>103043793
wait for 11.11

>>103043681
More accurately, they are yields that need more volts but scale high. Binning for Epyc is based on power efficiency.

>>103043817
the 4500 is monolithic
same with the 5500 and the APUs
chiplets use more power

>>103043856
So no hidden power draw? The 4500 is looking all the more attractive then.
Although at this point, with basically every other upgrade path for that socket offering the normal chiplet design for AMD, I am not sure if I should just go with the 12100f to be safer...

>$480

oh fugggg what's the 9900X3D gonna be

Any help on how you decide which brand you get for a RTX 4070 Super ? I know you need to consider overall length. But other than that I heard conflicting info on choosing 2 fan vs 3 fan etc.

>>103043793
Just wait two weeks, when Trump wins, they will stop pretending there's no recession and the economy will collapse overnight, after the civil war you'll be able to buy a whole supercomputer for pack of canned tuna from starving semi-zombified techies squatting in bombed-out datacenters inside the California Containment Cordon.

>>103044025
Never understood the retarded American concept of brand loyalty, just look at reviews of the particular model and buy the best according to budget.

>>103042302
>>103042560
Fuck. This is a very good point. Thank you, anon. I will get 2x 512gb.

>>103044057
some brands are dick

>>103044008
600 with the 9950X3D at 750

>>103044079
All brands are dicks, corporations always strive to maximize profits while fucking the consumer.

>>103044090
no i meant dick as in worse than others

What y'all think about cloud gaming? I have been using [redacted] for a month now and playing from my work Macbook. It's fucking good. I thought it would be lagging and unplayable but it's way better than expected. It's the reason I'll be skipping GPU in my build.

>>103044057

They all have fine reviews. Point being when you are spending $600 and have multiple choices might as well pick the best one.

>>103044098
But that's a variable between models not between brands.
Still, the amount of fans isn't really important imo after experiencing two cards with the same power draw using the same diameter fans yet the 2fan cooled the same as the 3fan variant. That's why I insist looking at reviews.

File: image.png (621 KB, 1440x1749)

621 KB PNG

>>103043902
>So no hidden power draw? The 4500 is looking all the more attractive then.
there's no good tests for this unfortunately
the 3400G should be the closest approximation on this chart
>with basically every other upgrade path for that socket offering the normal chiplet design for AMD
it's really not that bad, but your choice
what you should do is avoid X570 because that chipset uses ~15 W compared to ~5 W for B450/B550

i need an ssd for gaming GOOD. recommends?

>>103044025
take it from me anon, it's not about the length, it's about how you use it

>>103044057
If you're in the USA, the best option at MSRP is probably the ASUS Dual Evo. If you want to spend $20 more, the Gigabyte Gaming with the Quiet BIOS switch flipped to On is technically better, but on a card that only uses about 200W, you don't need more than the ASUS

...can I cool a 7950X3D with a a fan?

>>103044313
gaming is one area where whatever ssd works fine even if it's dramless trash or even sata

>>103044323
You can cool any cpu with a fan

>>103044332
...should I cool a 7950X3D with a a fan?

>>103044316
>but on a card that only uses about 200W
My asus dual evo draws 240W out of the box in gaming and the little shit even asks for more since hwinfo shows power throttling.

File: Senua's Saga Hellblade I(...).png (6 KB, 378x115)

6 KB PNG

>>103044344
The spec for the 4070 Super is 220W. Definitely undervolt your card. I have a 4070 non-Super that usually runs at around 150W

>>103044337
That CPU's TDP is around 120W. Any decent CPU fan cooler will handle that without much fuss.

>>103044230
Thanks for the info, the board I have lined up for the 4500 is a b550 so that's good to know. But would you also happen to know the best chipset choices for lga 1700 as well? Concerning power consumption again of course.

>>103044404
>But would you also happen to know the best chipset choices for lga 1700 as well? Concerning power consumption again of course
intel chipsets all have the same power draw (~6 W)
AMD chipsets kinda suck in comparison, they're outsourced to asmedia, but as long as B550 is enough for you the power draw is comparable

>>103044427
>5W for b550, 6W for Intel
So they are basically the same, nice.
>>103044230
Also this is for the whole system at idle and not just the CPU right?
If you were to guess, where would the 12100f land in that comparison?
And thanks again for the help, this info is hard to come by even after hours of searching.

>>103044427
>AMD chipsets kinda suck in comparison
Epyc is SoC so AMD doesn't care and goysumer chipsets get shunted off to the lowest bidder and we end up with glued-together abortions like the X-series motherboards.

>intel pumped after earnings call
apologize to Pat

>>103044472
>So they are basically the same, nice
yeah some sources say 7W for B550 so, basically the same
the issue with AMD is only when you're going high end, with say X670E or X870E they can get the same I/O as intel but they do it by putting two chipsets on the same board (which obviously increases manufacturing costs and power draw)
intel designs their own chipsets which is absolutely the way to go
>Also this is for the whole system at idle and not just the CPU right?
yes
like you said there's no good info out there, most reviewers don't test idle power at all
>If you were to guess, where would the 12100f land in that comparison?
very close but maybe a little higher, lower end chips are always more efficient since they sit lower on the voltage-efficiency curve
the 4500 is originally a laptop design which really helps with idle power, it has more C-states than the regular ryzen chips

>Gelsinger said Intel would manufacture its next-generation Panther Lake processors at Intel Foundry starting late next year.
lol

>>103044549
isnt the 18a meant to be functional in few months

File: veilguard.jpg (1.41 MB, 2560x1387)

1.41 MB JPG

Veilguard is super CPU-intensive with meme tracing on
CPU bottlenecked hard with a 5800X3D + 4090 4K DLSS quality at around
~80fps
the fps doesn't budge when lowering the resolution or changing to dlss performance

even without ray tracing it's still mostly CPU bound, DLSS quality -> ultra performance only gives around a 10% fps boost from ~100 -> 110

wasn't going to upgrade from the 5800x3d any time soon because thought figured you'd always be gpu bound at 4k even with a 4090 but now tempted to consider the 9800X3D

>>103044501
Ryzen is SoC too, has been zen1
It's just the "chipsets" like X300 use something called the knoll activator which only has the functions of allowing the CPU to activate and defining features enabled on the CPU (like overclocking) it offers no IO otherwise.

File: gelsinger torah.jpg (27 KB, 722x237)

27 KB JPG

>>103044556
trust the plan

>>103044580
>knoll activator
I'm not googling that.

>>103044576
well if you can affford a 4090 def for go the 9000 series
you get more fps from ddr5 ram as well

>>103044580
iirc at least zen1 had separate usb controllers in both cpu and chipset, maybe sata too, not sure about it.

>>103044522
Well that pretty much answers basically every single question I had, it's definitely going to make deciding between the two a lot easier.
One last thing though, and this is probably a pretty stupid question, but from what I have read, without XMP, your standard 3200mhz ram is going to be running at a lot lower speeds.
Could you increase your ram speed to its stock levels without using XMP or anything that affects idle draw?
>>103041642
As I understand it, XMP on amd enables infinity fabric, which is where the power draw problem comes from. So I am guessing XMP on Intel doesn't have a similar issue?

>>103044364
Did some OC+undervolt and now draws the same power but maintaining 2730 effective clock instead of 2610.
Wonder if I should keep the extra performance with same power draw or just stay at stock performance with less power draw.

>>103043191
sell?

>>103044633
you can manually lower the SOC voltage with XMP to reduce the power draw back to stock levels
the SOC scales a little differently on the monolithic chips so i can't comment on an optimal voltage, but at 3200 you should be able to set it right back to stock (since AMD already guarantees that will work OOTB)
>So I am guessing XMP on Intel doesn't have a similar issue
it doesn't have the same issue on non-K intel chips because motherboard manufacturers aren't allowed to overvolt the chip
the consequence of that is that XMP pretty much doesn't work on a 12100F above 3200-3400 MHz
remember that XMP is overclocking and the non-K chips aren't "intended" for overclocking

>>103043739
>Why haven't you done that yet anon?
I already did, but with an RX7600 XT instead(was on sale, bite me)

>>103044605
yeah well not that it's completely trivial sum but sure I can afford it just thought it was pointless at 4k
apparently not quite so
this one game may not be enough to pull the trigger but it does make one think
genuinely surprised to see a cpu bottleneck with a 5800X3D in a AAA game at 4K even if dlss quality is applied

Which pc case would you recommend for a first time build under 200?

>>103044687
Okay thanks, so I guess with a little tuning XMP between AMD and intel is about the same in power draw.
On a slightly different note though, I also assume that the 4500 and 4100 are interchangeable in the idle power comparisons since it's also a monolithic laptop design as far as I understand (like the 5500). With the 4100 probably being slightly lower in power usage because of it being an inherently weaker CPU.

I'm not paying $700 for a CPU and 1k for a GPU with less than 16GB VRAM in Canada to game at 1200p while sucking the 24v6pin. Never mind the rip off for all the rest other than ram which is tolerable still for some reason. When is healthy competition coming back bros.
>6700k, 16gb ram, and 1060 6gb for another ~2 years until Windows 10 takes me out of service
It was an honor playing games with you. I'm not paying $3000 for cluttered M2 slots wedged under a giant tower cooler or fucking with AIOs/radiators for increasingly shit TDP, sick of space heaters and coil whine. I'm going retro.

>>103044760
https://www.techpowerup.com/cpu-specs/ryzen-3-4100.c2757
>Zen 2 (Renoir)
https://www.techpowerup.com/cpu-specs/ryzen-5-4500.c2758
>Zen 2 (Renoir)
https://www.techpowerup.com/cpu-specs/ryzen-5-5500.c2756
>Zen 3 (Cezanne)
>https://www.techpowerup.com/cpu-specs/ryzen-5-5600.c2743
>Zen 3 (Vermeer)
renoir and cezanne are monolithic laptop designs
vermeer is a chiplet desktop design

the architecture is what matters, not the exact model number
any differences between say a 4100 and 4500 are probably too small to notice and i don't think anyone has bothered testing it

>>103044576
holy fuck I don't think I've ever seen a game nuke a cpu this hard before

>>103041791
New tech or new problems? 4090 is nothing special compared to 3090ti outside of being a die shrink and hueg as fuck transistor wise. Their "software innovations" like dlss only exist because raytracing is so garbage on current hardware yet is peddled as a must have.
At least AMD tries to innovate with their hardware with the chiplet design approach even if it didn't work out.

>>103041909
I did the same thing, you need to have the same sized DIMM's in a channel. 8+8 in Channel A and 16+16 in Channel B. Whether you are capable of hitting XMP speeds with this configuration depends on the quality of your CPUs IMC and motherboard trace quality. You should definitely do memory stress testing as well as this set up can be unstable.

>>103044801
gotta love to see it
proper utilization and spreading the work around, not just 1 or 2 threads maxed at 100% while the rest of the cpu is sitting close to idle

>>103044787
I see, so the difference is completely negligible.
But the 5500 (although different architecture), is still as you said a laptop design, so it should still have similar idle draw to the 4500/4100, and probably the 12100f for also being monolithic?
With the 12100f and 5500 probably having ever so slightly higher idle power draw than the 4500/4100 because they are stronger chips (something like 8W for the 12100f/5500 vs 5W for the 4500/4100 I assume), but still comparable because of their monolithic design and much lower than chiplet CPUs (which idle at around 30W from what I understand).
Did I get anything wrong?

>>103044576
Just gives you an idea what a piece of shit the messed up version of the frostbite game engine is that they used for that game. It's not fair to blame the 5800X3D. You can choke any CPU/GPU with shitty enough code.

>>103044922
the experience in game is good though, the frame rate is high enough and very smooth without major stutters like you see in every Unreal game and the visuals are good if you can stomach the controversial art style
if they manage all of that I don't really care how they get there, on the technical side this is one of the best AAA games I've played in a long time

>>103044911
i really can't give you any specific guarantees on the numbers but yeah, pretty much
any of them would be a good choice, i would rather make the choice based on other factors like performance and motherboard availability/pricing
>>103041909
>gamersnexus had an interesting benchmark video where he found a really weird performance anomally on zen3, where 32gb setup as 4x8 did BETTER than 2x8 for 16 gig
it's the difference between single and dual rank
dual rank is more "efficient"
>Supposedly 2x16 is even faster than that
it depends on what chips are inside
if you get 16gbit chips it will perform the same as 2x8
if you get dual 8gbit chips it will perform the same as 4x8

>>103044910
You can max out any CPU doing meaningless work, 100% utilization doesn't mean that the game is making good use of all that work.

>>103044507
>be Intel, lose $3.6 billion
stock goes up
>be AMD, make more money than expected by like $500million
stock goes down

>>103044576
Streamers are already crying about lag/dropped frames/stuttering

>>103045010
it's lucky I'm not a streamer then

>>103045025
Just means you're gonna need a good CPU for this game, saw similar shit when dragon's dogma 2 released.

>>103044967
>any of them would be a good choice, i would rather make the choice based on other factors like performance and motherboard availability/pricing
That's what I am going to do now that I know where both stand.
And I guess upgradability is also something to consider since there are no higher end monolithic CPUs for AMD. Whereas an i7 12700 should still idle around or under 10W.
Unfortunately you can't really go much more beyond that because of the defective 13/14th gens.

Is it possible to build a desktop machine adequate for office and shitposting tasks and watching videos that consumes equal to or less power than a lightbulb?

Hey anons, I'm in the market for a monitor mount now that I have a desk I can actually safely attach one to. I've got three 1080p monitors I want to mount. Any recommendations?

>>103045228
all of them are the same metal tubes from china, probably the same factory as well

>>103045220
Your monitor is basically a big fancy lightbulb so no.

>>103045220
100w (bright old-school incandescent lightbulb) is enough for an Intel T-based system with low-powered peripherals
but uh if efficiency is your goal you may just be looking for a mini-pc or even a laptop instead of a traditional desktop

>>103045220
Yes.

I wanted to get an AsRock Z790 Pro RS WiFi motherboard but the only issue is that even after visiting their site its not clear what RAM sticks are going to work out of the box with no BIOS update. Is there a way to figure this out or is there a better brand where I can easily tell which RAM will work out of the box ?

>>103045220
Buy? sure, Intel N100 mini PCs draw like 10w on load.
Build? Probably with underclocked/undervolted components.

>>103045292
excluding the monitor ofc

>>103045299
>Intel T-based system with low-powered peripherals
the right mobo would also need to be a factor for sure

File: 1712318837138685.png (25 KB, 626x287)

25 KB PNG

do i get this and be free from my 1080 ti shackles

>>103045345
yes