R7 5800X3D vs R9 5950X for 4K gaming.

[Zekkah]

I've read that the 5800X3D is better for gaming, but the 5950X is superior at higher resolutions (4K), which I play at most of the time.

Pairing it with a RX 7800 XT, which one is better? The CPU I have now is the R7 5800X.

I'd like to know your opinions.

[C1REX]

Origineel geplaatst door De Hollandse Ezel:

thats the only way to properly run 4k. if not.,. stick to 1440p.

Many games have excellent DLSS implementation and look better at 4K+DLSS than native 1440p at a similar performance requirement. Games with dynamic scaling also look best at 4K. Most games are designed for consoles first and to be displayed at 4K.

The OP probably has a 4K screen and I see no reason to ask him to buy a different one now.

[Zekkah]

Origineel geplaatst door De Hollandse Ezel:

thats the only way to properly run 4k. if not.,. stick to 1440p.

Or I'll just stick to 4K with 60 FPS. Most 4K monitors are just 60Hz anyway, with the few ones going over that being very expensive. I had a 1080p 75 Hz one before and wouldn't trade resolution over refresh rate. And like C1REX said, there's upscaling solutions that work very well.

To conclude the topic, while the Ryzen 9 5950X isn't better than the Ryzen 7 5800X3D for gaming at 4K, the differences do become insignificant at that resolution, with the 5950X being clearly superior at everything else other than gaming. I do 3D modeling and watch movies with SVP 4 Pro on my computer, so the final decision is a no-brainer, specially with how expensive the 5800X3D is.

[r.linder]

You'll be so far GPU bottlenecked by a 7800-XT that even going up to AM5 with a 7800X3D would make almost zero difference, most of any differences would come down to loads where cache helps

Getting a better CPU will only really help with other kinds of loads that actually need more processing power, and streaming using CPU encoders

[Zekkah]

Origineel geplaatst door r.linder:

You'll be so far GPU bottlenecked by a 7800-XT that even going up to AM5 with a 7800X3D would make almost zero difference, most of any differences would come down to loads where cache helps

Getting a better CPU will only really help with other kinds of loads that actually need more processing power, and streaming using CPU encoders

An online calculator says it's a 7.1% GPU bottleneck at 4K. At 1800p it's 0.1%. I assume it's not taking overclocking into account.

[C1REX]

Origineel geplaatst door Zekkah:

Origineel geplaatst door r.linder:

You'll be so far GPU bottlenecked by a 7800-XT that even going up to AM5 with a 7800X3D would make almost zero difference, most of any differences would come down to loads where cache helps

Getting a better CPU will only really help with other kinds of loads that actually need more processing power, and streaming using CPU encoders

An online calculator says it's a 7.1% GPU bottleneck at 4K. At 1800p it's 0.1%. I assume it's not taking overclocking into account.

These numbers are mostly made up as bottlenecks change with each game, each graphical setting and even where in the game you are. Towns for example are usually more CPU heavy.

[r.linder]

Origineel geplaatst door Zekkah:

Origineel geplaatst door r.linder:

You'll be so far GPU bottlenecked by a 7800-XT that even going up to AM5 with a 7800X3D would make almost zero difference, most of any differences would come down to loads where cache helps

Getting a better CPU will only really help with other kinds of loads that actually need more processing power, and streaming using CPU encoders

An online calculator says it's a 7.1% GPU bottleneck at 4K. At 1800p it's 0.1%. I assume it's not taking overclocking into account.

Bottleneck calculators are bogus because how much something will bottleneck will completely differ from load to load and you can't quantify bottlenecks as a percentage accurately due to at least thousands of possible system configurations.

The fact of the matter is that the 7800-XT isn't that great for 2160p above 60Hz and it's pretty much the bare minimum for 2160p ultra. The 7900-XTX, 4080 SUPER, etc. can go much higher in many cases and those cards would be more likely to incur a real bottleneck from the CPU, whereas a 4070 or 7800-XT will pretty much always be the limiting factor for gaming.

[Zekkah]

Origineel geplaatst door C1REX:

These numbers are mostly made up as bottlenecks change with each game, each graphical setting and even where in the game you are. Towns for example are usually more CPU heavy.

Origineel geplaatst door r.linder:

Bottleneck calculators are bogus because how much something will bottleneck will completely differ from load to load and you can't quantify bottlenecks as a percentage accurately due to at least thousands of possible system configurations.

It's an average. It can be used as a reference while still taking all that into consideration.

[76561199793899658]

ok i use my other account cause moderating here sucks (you actually say a good thing than somebody else responces weird and goes full antisemetic) and you get banned by proxy for "prone to argument"

but ok.. you said 60 fps 4k is enough for you.
i hate upscaling myself i turn that crapp off.

to do 60fps in 4k you need an 6900xt or better 7800xt or better 3080 or better or 4070s or better

3080 + 5800x3d is fine
4070s + 5800x3d is fine
6900xt + 5800x3d is fine
7800xt + 5800x3d is fine


7800xt + 5950x gives a sizable bottleneck of over 11%

6900xt + 5950x gives a sizable bottleneck of 9%

3080 + 5950x gives a bottleneck of 9%

4070s + 5950x gives a bottleneck of 4%

this does not linearly scale with more potent card it seems radeon cards bottleneck a bit sooner.
these percentahes roughly

as the 4070s actually performs the best (closer to 65fps) the 4% loss still should keep you above 60 fps you desired.

with a 7800xt + 5950x your performance will drop back to around 52fps

ofcourse these numbers are without overclock and generalised..
(meaning averaged for various game titles)

so no.. the 5950x is not better in gaming it bottlenecks earlier.
but it is better in rendeeing/tasks thar benefit from the extra cores.. if you do such tasks ot coulf be worth taking the hit but than really go with 4070s..

if not.. go 5800x3d

[r.linder]

An average based on specific games isn't relevant when taking into consideration all current and future games. It's pointless to worry about specific numbers when it's literally impossible to accurately quantify it as a number because you're talking about calculating it based on tens of thousands of games, including new games as they come out to keep the data relevant.

It's not a reference, it's hocus pocus nonsense.

[76561199793899658]

Origineel geplaatst door r.linder:

An average based on specific games isn't relevant when taking into consideration all current and future games. It's pointless to worry about specific numbers when it's literally impossible to accurately quantify it as a number because you're talking about calculating it based on tens of thousands of games, including new games as they come out to keep the data relevant.

It's not a reference, it's hocus pocus nonsense.

you full well know that no hardware is ever tested like that.. cant be done they buy various cpu and gpu.. combine yhem in all variations run 8-20 game titles with them generally the latesr most demanding and from a broad spectrum of genres.. and than publish the data..

they ofcourse use stock models and not faster clocked varianta of these gpus nor overclock the cpu due that being waay to many variables to test.

and the data usually is a very good reference point. 5950x will bottleneck sooner tham 5800x3d and 4070s will bottleneck later/less than the 7800xt 6900xt and 3080. despite performing slighly better. thats the trend.. wheter than means in a particulair game 50 60 or 70 fps.. especially future titles and if you do a cpu overclock.. yeah that cant be tested for.

[r.linder]

Origineel geplaatst door Outcast82:

Origineel geplaatst door r.linder:

An average based on specific games isn't relevant when taking into consideration all current and future games. It's pointless to worry about specific numbers when it's literally impossible to accurately quantify it as a number because you're talking about calculating it based on tens of thousands of games, including new games as they come out to keep the data relevant.

It's not a reference, it's hocus pocus nonsense.

you full well know that no hardware is ever tested like that.. cant be done they buy various cpu and gpu.. combine yhem in all variations run 8-20 game titles with them generally the latesr most demanding and from a broad spectrum of genres.. and than publish the data..

they ofcourse use stock models and not faster clocked varianta of these gpus nor overclock the cpu due that being waay to many variables to test.

and the data usually is a very good reference point. 5950x will bottleneck sooner tham 5800x3d and 4070s will bottleneck later/less than the 7800xt 6900xt and 3080. despite performing slighly better. thats the trend.. wheter than means in a particulair game 50 60 or 70 fps.. especially future titles and if you do a cpu overclock.. yeah that cant be tested for.

"Bottleneck calculators" are a waste of time at best, and misinformation at worst. That's a fact, nobody that actually knows what they're talking about recommends using them because it's impossible to accurately calculate down to a percentage how much a particular component is going to bottleneck.

It's not even worth worrying about bottlenecks because it's inevitable, unavoidable. Just enjoy the hardware as it performs and stop worrying about imaginary numbers that mean absolutely nothing.

[C1REX]

Bottleneck calculators are such a BS is beyond words.
As soon as you change from ultra to high or even better use optimised settings and add upscaling, what almost everyone does at 4K, the bottlenecks calculations are completely worthless.

With GPU bottlenecks there is always a lot of room for improvement. Changing upscaling or resolution can make a GPU even 4x “faster”.

When CPU bound there is close to nothing you can do in most games. If a CPU can’t hold 60fps then it’s over really. You may as well lock fps close to 1% lows - probably at 30fps- for more stable performance but not much more.

[76561199793899658]

Origineel geplaatst door C1REX:

Bottleneck calculators are such a BS is beyond words.
As soon as you change from ultra to high or even better use optimised settings and add upscaling, what almost everyone does at 4K, the bottlenecks calculations are completely worthless.

With GPU bottlenecks there is always a lot of room for improvement. Changing upscaling or resolution can make a GPU even 4x “faster”.

When CPU bound there is close to nothing you can do in most games. If a CPU can’t hold 60fps then it’s over really. You may as well lock fps close to 1% lows - probably at 30fps- for more stable performance but not much more.

in gaming

if getting a faster cpu with same gpu.. your bottlenecked

if getting a faster gpu on same cpu.. gains you fps you have an unbalnced build.

the 5950x is just bottlenecking all the listed 60fps 4k gpus

and that bottleneck. if you lower settings etc only gets worse.
so the point still stays.

only overclocking the cpu will lower the bottleneck a little.

btw
i and everybody around me hates dlss and upscaling deeply and leaves it OFF.
it looks uglier than less fps but real rendering of 4k.
and turning settings lower than maxed out ugh thats even worse.

better max settings real rendered 1080p 100fps than upscaled mid settings 4k 60fps

ofcourse thats if your a pauper for 1440p 100fps max settings no upscaling.. is quite afordable... only 4k 100fps no upscaling is still pricey..

[C1REX]

Origineel geplaatst door Outcast82:

i and everybody around me hates dlss and upscaling deeply and leaves it OFF.
it looks uglier than less fps but real rendering of 4k.
and turning settings lower than maxed out ugh thats even worse.

better max settings real rendered 1080p 100fps than upscaled mid settings 4k 60fps

ofcourse thats if your a pauper for 1440p 100fps max settings no upscaling.. is quite afordable... only 4k 100fps no upscaling is still pricey..

I think we will disagree on this one.
DLSS at 1080 does indeed look bad.
DLSS quality at 1440p looks fine depending of implementation. And it's easier to run than native 1080p.
DLSS at 4K looks fantastic.

After trying 4K I will never go back to 1440p.
1440p upscaled to 4K looks much better than native 1440p in most new games. Even FSR looks fine at 4K.

[Damian]

Origineel geplaatst door Zekkah:

I've read that the 5800X3D is better for gaming, but the 5950X is superior at higher resolutions (4K), which I play at most of the time.

Pairing it with a RX 7800 XT, which one is better? The CPU I have now is the R7 5800X.

I'd like to know your opinions.

I have 9950X and RTX 7900XTX... the higher resolution the less CPU matters...
At 4K doesnt matter if you have 9800x3D or 9950X etc.
Because CPU matters only once you reach 144fps+

And at 4K gaming 144FPS is a lot. Also remember to disable Anti Aliasing at 4k:)