Rtx 5090 interesting behavior *help*

[Wulfy83]

Ok so me and my buddy have very similar systems.  We both just got identical MSI RTX 5090 Gaming Trio OC cards.  In my system my card on Port Royal is around 37,600-37,800 points or so, while my buddy is getting 38,400+
I took my pc up to his place and we ran some baselines with stock settings (fan speed, core clock, mem clock) and with the same driver 581.08.
My card in my pc: 37744
his card in his pc: 38498
His card in my pc: 38316
My card in his pc: 38423

He has a 1300watt psu while i have a 1000watt psu.   He kind of figured maybe his card is more power efficient.
I have a new 1600 watt psu and just jumped it to power only the video card in my system, and still getting low numbers. Just got a 37946,   I am confused, frustrated, and a bit disappointed to be honest.   Not sure what to do. 
Anyone have any ideas?

 

 

 

[emosun]

did you want to tell us anything important like what the rest of the system parts are or what the config was previously or what the os is or how old the install is or the diagnostics you've already done that don't include randomly plugging another power supply in...

[Affe]

Post your complete System Specs along with his(cpu, ram, coolers, case, etc). Port Royal score is also affected by FPS, thermals and clock stability. 

You should also post both your results and his results from the 3DMark website. 

Here's mine for example: NVIDIA GeForce RTX 5090 video card benchmark result - AMD Ryzen 7 9800X3D,Micro-Star International Co., Ltd. PRO X870-P WIFI (MS-7E47)
My stuff is overclocked too.

[Arkh82]

Hey, My friend posted this for me since I hadn't yet made an account here.

Systems are as follows.

Mine:                                                                                    His:  

Ryzen 9 9950X3D                                                               Ryzen 9 9800X3D

Asus TUF Gaming B850M-Plus Wii                                   Gigabyte X870E Aorus Master

32GB G.Skill Trident Z Neo RGB                                        96GB G.Skill FlareX 5

MSI RTX 5090 Gaming Trio OC                                         MSI RTX 5090 Gaming Trio OC

 

 

Both CPU's are currently liquid cooled in a custom loop.

Both Win11 23H2 (24H2 can burn)

Reults in order of testing as per 1st post.

 

Each score is the link

 

My card in my pc: 37744
his card in his pc:  38498
His card in my pc: 38316
My card in his pc: 38423

 

Before i was running an MSI RTX 4090 Gaming Slim and my friend was on a Radeon 7900XTX.

We both ran DDU to remove old drivers and installed a fresh copy of the latest nVidia drivers.

 

I used Ryzen master and set to game mode to essentially make my 9950 a 9800 to see if that made any difference. 

It did not seem to.

 

Just a little while ago i remebered i used a different DP port on my card vs his,  and he used HDMI.     I swapped my DP to the one i used when his card was in my system and scored a 38100.   kind of interesting.

 

Appreciate the replies, and sorry for not posting system specs to begin with.

[PDifolco]

That's only a 2% difference, it's really tiny...

Maybe his PC is a bit cooler, his CPU or RAM a tiny bit faster

[MURFING]

1 hour ago, Arkh82 said:

Hey, My friend posted this for me since I hadn't yet made an account here.

Systems are as follows.

Mine:                                                                                    His:  

Ryzen 9 9950X3D                                                               Ryzen 9 9800X3D

Asus TUF Gaming B850M-Plus Wii                                   Gigabyte X870E Aorus Master

32GB G.Skill Trident Z Neo RGB                                        96GB G.Skill FlareX 5

MSI RTX 5090 Gaming Trio OC                                         MSI RTX 5090 Gaming Trio OC

 

 

Both CPU's are currently liquid cooled in a custom loop.

Both Win11 23H2 (24H2 can burn)

Reults in order of testing as per 1st post.

 

Each score is the link

 

My card in my pc: 37744
his card in his pc:  38498
His card in my pc: 38316
My card in his pc: 38423

 

Before i was running an MSI RTX 4090 Gaming Slim and my friend was on a Radeon 7900XTX.

We both ran DDU to remove old drivers and installed a fresh copy of the latest nVidia drivers.

 

I used Ryzen master and set to game mode to essentially make my 9950 a 9800 to see if that made any difference. 

It did not seem to.

 

Just a little while ago i remebered i used a different DP port on my card vs his,  and he used HDMI.     I swapped my DP to the one i used when his card was in my system and scored a 38100.   kind of interesting.

 

Appreciate the replies, and sorry for not posting system specs to begin with.

I am no expert, but I would say it could be a silicon loettry, ya know.

[Arkh82]

53 minutes ago, MURFING said:

I am no expert, but I would say it could be a silicon loettry, ya know.

Yes, but does not explain the reason his card in my pc works just fine.

[Arkh82]

Well definitely is a configuration issue.  I did a fresh install of Win11 24H2 for testing.  Was still getting in the 38100 range. 

 

I remembered we disabled 1 ccd up at my buddies house. So I did that and got 38300.

 

So pretty close now. I then proceeded to try some overclock runs where my previous best was 40163. Best I achieved so far now is 41475.  My buddy has hit 41600 something. That like half a percent diff.  Just need to figure out why my score is worse when both ccd's are enabled .

 

Definitely open to suggestions on that.

 

Thanks.