Question Is my Asus RTX 5080 Prime OC holding back..?

[sharpshotter]

Hey!

Im a pretty dedicated gamer, and was lucky enough to get my hands on a 5080 (while actually trying to buy the 5070ti), and im wondering if this is a normal score for the gpu.. I´ve seen countless posts about the score being closer to 35k on the graphics side without OC or anything.

Also, im aware that the ryzen 7600 bottleneck 10-15% while gaming in 1440p, but that shouldnt really be visable in this test, - right? (With the separated gpu performance)

Anyway. Im hoping someone can give me some guidance here cuz im not really sure what to do.

EDIT: I own the ASUS RTX 5080 Prime OC

https://imgur.com/a/b8GYB2m

 

[helpstar]

looks fine
https://www.notebookcheck.net/Asus-...ed-pricing-over-the-RTX-5080-FE.962110.0.html

 

[sharpshotter]

looks fine
https://www.notebookcheck.net/Asus-...ed-pricing-over-the-RTX-5080-FE.962110.0.html

Hmm! thats weird. Most benchmarks ive seen (fr. swedish reviewers) seem to point at a higher graphical score. This card was sold as msrp over here as well, so I guess we got lucky!

 

[helpstar]

BIOS up to date of your motherboard?
drivers from nvidia as well?

 

[sharpshotter]

BIOS up to date of your motherboard?
drivers from nvidia as well?

BIOS is up to date, and the nvidia drivers as well. Fan-wise I have a decent setup with 3 front fans, 2 top, 1 exhaust all aligned correctly. The cpu never exceeds 65 degrees celsius, and the gpu had a top at like 75 degrees for a moment, otherwise at around 70.

Im wondering if armorycrate and icue might interfere? Or something else running in the background?

 

[USAFRet]

Also, im aware that the ryzen 7600 bottleneck 10-15% while gaming in 1440p

And this 10-15% number comes from...where?

 

[sharpshotter]

And this 10-15% number comes from...where?

Well, thats pretty much what people are writing in other forums. Im not so sure. But I´ve also benchmarked according to these tests just to try to make sense of it. Link to FE performance (which should be lower than the prime OC) but paired with a AMD Ryzen 7 9800X3D instead. The rest is also way better than my current setup, but shouldnt give that much of an impact.

Cyberpunk benchmark is with the ultra (out of the box preset), horizon and starfield are the same "routes" as the reviewers are running. Same graphics settings.

https://imgur.com/a/sWP1vNs

View: https://imgur.com/a/sWP1vNs
(ME)
AVG 137
LOW 100

https://imgur.com/a/MVSTM6P

View: https://imgur.com/a/MVSTM6P
(ME)
AVG 126
LOW 95

https://imgur.com/a/IiD8IR5

View: https://imgur.com/a/IiD8IR5
(ME)
AVG 110
LOW 91

The rest of my specs:
PSU: Asus tuf gaming 850W
Mobo: Asus B650-M PLUS
RAM: Corsair vengance 5200 mhz (CL40) - / CL30 was out of stock and still is -,-
SSD: nVME m2 (Dont remember speeds but it shouldnt cause any problems)
Cpu fan: Deepcool AK620 digital pro

Other: 6x Corsair RS120 argb

 

[beyondlogic]

Well, thats pretty much what people are writing in other forums. Im not so sure. But I´ve also benchmarked according to these tests just to try to make sense of it. Link to FE performance (which should be lower than the prime OC) but paired with a AMD Ryzen 7 9800X3D instead. The rest is also way better than my current setup, but shouldnt give that much of an impact.

Cyberpunk benchmark is with the ultra (out of the box preset), horizon and starfield are the same "routes" as the reviewers are running. Same graphics settings.

https://imgur.com/a/sWP1vNs

View: https://imgur.com/a/sWP1vNs
(ME)
AVG 137
LOW 100

https://imgur.com/a/MVSTM6P

View: https://imgur.com/a/MVSTM6P
(ME)
AVG 126
LOW 95

https://imgur.com/a/IiD8IR5

View: https://imgur.com/a/IiD8IR5
(ME)
AVG 110
LOW 91

The rest of my specs:
PSU: Asus tuf gaming 850W
Mobo: Asus B650-M PLUS
RAM: Corsair vengance 5200 mhz (CL40) - / CL30 was out of stock and still is -,-
SSD: nVME m2 (Dont remember speeds but it shouldnt cause any problems)
Cpu fan: Deepcool AK620 digital pro

Other: 6x Corsair RS120 argb

what was the hardware of the review.

also if you have armour crate installed you may need to implement game mode. as it can throttle hardware.

 

[sharpshotter]

what was the hardware of the review.

also if you have armour crate installed you may need to implement game mode. as it can throttle hardware.

Ahh, gamemode you say.. Gonna look it up! Thanks!

Specs of the benchmarking PC:
Seasonic Prime-TX Noctua Edition, 1 600 W
Asus ROG Crosshair X870E Hero
AMD Ryzen 7 9800X3D
Noctua NH-D15 G2 LBC (cpu cooler)
2× 16 GB G Skill Trident Z5 NEO RGB 6 000 MHz, 28-36-36-96
Kingston Fury Renegade 2 TB

Dont know if the rest matters but here they are:
Streacom BC1 Open Benchtable (Case)
AOC U27G3X (Monitor)

Out of the box clocks, no oc or anything.

 

[beyondlogic]

Ahh, gamemode you say.. Gonna look it up! Thanks!

Specs of the benchmarking PC:
Seasonic Prime-TX Noctua Edition, 1 600 W
Asus ROG Crosshair X870E Hero
AMD Ryzen 7 9800X3D
Noctua NH-D15 G2 LBC (cpu cooler)
2× 16 GB G Skill Trident Z5 NEO RGB 6 000 MHz, 28-36-36-96
Kingston Fury Renegade 2 TB

Dont know if the rest matters but here they are:
Streacom BC1 Open Benchtable (Case)
AOC U27G3X (Monitor)

Out of the box clocks, no oc or anything.

the thing to keep in mind at 1080p and 1440p the cpu is still important only when going to 4k does the cpu more or less start to put its feet up.

the score is including your cpu you can compare similar scores on that test. with someone elses.

other ways to check if cpu is holding you back in gaming for example is to see the gpu utilisation if its at 100 percent or close to it your practically got the gpu as far as it can go. some games will also swing at cpu performance more then gpu.

cyberpunk or bug punk i call likes a faster cpu and more cores at least from my testing i get less hitching.

 

[Gaidax]

You are really going into a rabbit hole with all this.

9800X3D vs 7600 of course matters a lot, especially in 1440p, where CPU is still plenty relevant.

Gotta compare apples to apples, not apples to oranges.

---

Besides, people randomly yapping on forums how they get all these crazy scores is often misleading. Which resolution, what settings, what hardware so on and so forth.

Like the person above said, if you want to try to compare actual scores - you compare against same hardware, as in this part here:

Which seems to me par the course given your GPU is practically a stock model, the "OC" part on Prime is a joke, it's like 30Mhz OC, just so they could put "OC" on the box.

So given that, the fact that pretty much all non-reference cards aside Prime/FE have at least somewhat more of OC and the fact that many people on 3DMark slam overclocks to their 5080s, that score seems just about right.

If you want to juice more out of it - hit 5080 OC guides, the GPU has a lot of juice there untapped.

I managed + 470ish on core and +2000 on memory for example ezpz.

This is mine BTW, a simple PNY 5080 OC. A simple 3 slot cooler and a budget model for a budget vendor.

---

P.S. Do keep in mind that too aggressive of an overclock may shorten your GPU life. But if you would do half of what I did, it should not really matter and that already would be a nice few extra more % overall.

My OC was all I needed to be just at a shy more than ~60FPS average in CP77 Path Tracing with DLSS-Q, and 1440p Ultrawide.

 

[sharpshotter]

Yeah I guess so.. Well, from the tests ive been running myself, the 7600 feels a bit sluggish being paired with this card. So I guess I know where my money goes next. Thanks!
You are really going into a rabbit hole with all this.


Thank you so much for a well explained and detailed answer. I guess you´re write about that whole "OC" thing with this card. However, I didnt believe that it was almost on par with the FE!! Didnt have time to do any research cause all the cards in sweden sold out the first couple of seconds xD But yeah, that explains it I suppose.

I think that a light oc combined with a better cpu would do the trick. Especially since I want to futureproof this build as much as I can.

Thanks a bunch!

 

[Gaidax]

However, I didnt believe that it was almost on par with the FE!!

Yeah, PRIME for Asus is their most stock budget line. Everything PRIME for Asus is practically the basics.

On the flipside, for 5080 it does not terribly matter unless you intend to OC hard. You might as well be happy you saved yourself a stack of greenbacks.

Give it a mild OC of +120 on core and +1000 on VRAM with MSI Afterburner and be happy. That kind of OC is safe, and it will give you feels good ~4% extra for free.

 

[VizzieTheViz]

Looks pretty much like you’d expect. You could check if you got stuck with one of the 50 series where they “forgot” to put in a few ROP’s.

 

[beyondlogic]

Yeah I guess so.. Well, from the tests ive been running myself, the 7600 feels a bit sluggish being paired with this card. So I guess I know where my money goes next. Thanks!
You are really going into a rabbit hole with all this.


Thank you so much for a well explained and detailed answer. I guess you´re write about that whole "OC" thing with this card. However, I didnt believe that it was almost on par with the FE!! Didnt have time to do any research cause all the cards in sweden sold out the first couple of seconds xD But yeah, that explains it I suppose.

I think that a light oc combined with a better cpu would do the trick. Especially since I want to futureproof this build as much as I can.

Thanks a bunch!

this is my hot take with cpu wise ( and my own experiance and opinion).

AMD Ryzen 9 7900 would be a good choice 12 cores main reason I'm using a 5900x for 1440p and while i was using 8 cores i still got hitching and slow downs. on 12 cores I'm experiencing almost zero .​

the x3d parts to me are overpriced. on am5. and while 12 cores arent fully being used yet im seeing some games like hogwarts legacy using them more.

frame rate to me means ah heck all if its a complete mess.

i had way less stutter going from 8-12 cores. my framerate didnt improve much at all but the overall smoothness of the gameplay was better.