3DMark - Fire Strike Challenge

[The Mother Fucker 26]

Show me what the best you got!  

 

 

[kaz 2,439]

mine would be a negative score

[DynamiteCop 2,085]

Remij is going to eat your ass alive. 

[DynamiteCop 2,085]

Why is my Vega 64 taking a dump all over your Vega 56's?

 

Edited April 19, 2019 by Nya`

[Remij 4,670]

Crossfire ain't working breh.. you should be killing that score. 

 

Actually, looking at results on 3dmark, I see all the Threadripper 2990WX scores are garbage and Crossfire doesn't seem to work..

[Remij 4,670]

Here's mine.  Would be a lot better overall if I had a better CPU.  

 

[DynamiteCop 2,085]

9 minutes ago, Remij_ said:

Crossfire ain't working breh.. you should be killing that score.

No it is but there's something else going on, his individual graphics score should be like 16,000 something but it's 20,931 for both. Could that Threadripper be bottlenecking his GPU's?

Edited April 19, 2019 by Nya`

[Remij 4,670]

1 minute ago, Nya` said:

No it is but there's something else going on, his individual graphics score should be like 16,000 something but it's 20,931 for both.

There's Vega 56 2x crossfire gfx scores over 40,000 with Intel CPUs..

[DynamiteCop 2,085]

5 minutes ago, Remij_ said:

There's Vega 56 2x crossfire gfx scores over 40,000 with Intel CPUs..

Could that Threadripper be bottlenecking his GPU's?

[Remij 4,670]

10 minutes ago, Nya` said:

Could that Threadripper be bottlenecking his GPU's?

I just think Crossfire isn't working.  His score is 20,000... yours for the 64 is 24,500... that seems about right for a single card.  So I think the 2nd card just isn't being utilized.

 

@The Mother Fucker use MSI Afterburner or whatever and see if your 2nd card is being utilized at all.  Regardless, there's definitely something wrong.  But as I said Dynamite, there's no instances of Crossfire working at all on 3dmarks results with the 2990WX.

 

Must be the CPU.

[DynamiteCop 2,085]

Just now, Remij_ said:

I just think Crossfire isn't working.  His score is 20,000... yours for the 64 is 24,500... that seems about right.  So I think the 2nd card just isn't being utilized.

 

@The Mother Fucker use MSI Afterburner or whatever and see if your 2nd card is being utilized at all.  Regardless, there's definitely something wrong.  But as I said Dynamite, there's no instances of Crossfire working at all on 3dmarks results with the 2990WX.

 

Strange.

Vega 56's post in the 16k's so it's parsing the other card to a degree, this is kind of weird. 

 

 

[Remij 4,670]

2 minutes ago, Nya` said:

Vega 56's post in the 16k's so it's parsing the other card to a degree, this is kind of weird. 

 

 

He's presumably OC'd his cards like everyone else... an OC'd Vega 56 isn't doing 16k.. 

 

I'm pretty much guaranteeing that only a single card is being utilized.

 

It could be a driver issue (try reinstalling fresh drivers, or try the previous drivers) or his PSU might not have the juice for all that?

[DynamiteCop 2,085]

4 minutes ago, Remij_ said:

He's presumably OC'd his cards like everyone else... an OC'd Vega 56 isn't doing 16k.. 

 

I'm pretty much guaranteeing that only a single card is being utilized.

 

It could be a driver issue (try reinstalling fresh drivers, or try the previous drivers) or his PSU might not have the juice for all that?

Maybe he only has one card but it's showing up as two?

[Remij 4,670]

8 minutes ago, Nya` said:

Maybe he only has one card but it's showing up as two?

I'm not too sure about that.  I checked his results in detailed view and his hardware shows 2 physical gpus with slightly different clock speeds (not overclocked terribly high), so I definitely think there's 2 physical gpus there.  He's got 64GB of RAM too.  Pretty beastly system.  Definitely smokes mine in any applications that aren't games.

 

Wonder what's going on there?  I think it could be some driver issue, or his PSU isn't up to the task.

[DynamiteCop 2,085]

Updated my score, I forgot I had my CPU 200mhz down because of that other PSU that was failing. Gave me a little boost. 

Edited April 19, 2019 by Nya`

[The Mother Fucker 26]

@Remij_ and @Nya` I think you're both right, I think my issues could be the lack of Crossfire support with this benchmark coupled with known performance regression issues with the 2990wx Threadripper.  

Those are all stock settings using the AMD Ryzen Master Precision Boost OverDrive on that benchmark.  The VEGA 56 with the higher clock is a OC Air Boost from MSI and the other with low clock is a XFX reference VEGA 56.   XFX is the primary display, there's nothing plugged into the MSI VEGA.  

 

When I get home I'm going to play around with Ryzen Master some more, there's a gaming mode and few other modes where I believe  will let me turn off  2 of the CPU dies bringing to the 2950X TR level which apparently performs better in gaming.  Stay tooned.

[DynamiteCop 2,085]

Just now, The Mother Fucker said:

@Remij_ and @Nya` I think you're both right, I think my issues could be the lack of Crossfire support with this benchmark coupled with known performance regression issues with the 2990wx Threadripper.  

Those are all stock settings using the AMD Ryzen Master Precision Boost OverDrive on that benchmark.  The VEGA 56 with the higher clock is a OC Air Boost from MSI and the other with low clock is a XFX reference VEGA 56.   XFX is the primary display, there's nothing plugged into the MSI VEGA.  

 

When I get home I'm going to play around with Ryzen Master some more, there's a gaming mode and few other modes where I believe  will let me turn off  2 of the CPU dies bringing to the 2950X TR level which apparently performs better in gaming.  Stay tooned.

That benchmark definitely supports Crossfire, something else has to be happening. 

[lynux3 2,052]

Did you even put the Threadripper into Game Mode?

[The Mother Fucker 26]

Just now, lynux3 said:

Did you even put the Threadripper into Game Mode?

No, so far I only tweaked my memory with it to get the true 3000Mhz rate speed and not the BS 2933Mhz it was getting.   Hind sight, I probably should have got 3200 rate DDR4.  

[DynamiteCop 2,085]

2 minutes ago, The Mother Fucker said:

No, so far I only tweaked my memory with it to get the true 3000Mhz rate speed and not the BS 2933Mhz it was getting.   Hind sight, I probably should have got 3200 rate DDR4.  

Here's when I was running two 290X's.