• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

DirectX 12 Mixed Multi-GPU: It Works, For Now

btarunr

Editor & Senior Moderator
Staff member
Joined
Oct 9, 2007
Messages
47,668 (7.43/day)
Location
Dublin, Ireland
System Name RBMK-1000
Processor AMD Ryzen 7 5700G
Motherboard Gigabyte B550 AORUS Elite V2
Cooling DeepCool Gammax L240 V2
Memory 2x 16GB DDR4-3200
Video Card(s) Galax RTX 4070 Ti EX
Storage Samsung 990 1TB
Display(s) BenQ 1440p 60 Hz 27-inch
Case Corsair Carbide 100R
Audio Device(s) ASUS SupremeFX S1220A
Power Supply Cooler Master MWE Gold 650W
Mouse ASUS ROG Strix Impact
Keyboard Gamdias Hermes E2
Software Windows 11 Pro
One of biggest features of DirectX 12 is its asymmetric multi-GPU that lets you mix and match GPUs from across brands, as long as they support a consistent feature-level (Direct3D 12_0, in case of "Ashes of the Singularity"). It's not enough that you have two DirectX 12 GPUs, you need DirectX 12 applications to make use of your contraption. Don't expect your older DirectX 11 games to run faster with a DirectX 12 mixed multi-GPU. Anandtech put Microsoft's claims to the test by building a multi-GPU setup using a Radeon R9 Fury X, and a GeForce GTX 980 Ti. Some interesting conclusions were drawn.

To begin with, yes, alternate-frame rendering, the most common multi-GPU method, works. There were genuine >50% performance uplifts, but nowhere of the kind you could expect from proprietary multi-GPU configurations such as SLI or CrossFire. Second, what card you use as the primary card, impacts performance. Anandtech found a configuration in which the R9 Fury X was primary (i.e. the display plugged to it), and the GTX 980 Ti secondary, to be slightly faster than a configuration in which the GTX 980 Ti was the primary card. Mixing and matching different GPUs from the same vendor (eg: a GTX 980 Ti and a GTX TITAN X) also works. The best part? Anandtech found no stability issues in mix-matching an R9 Fury X and a GTX 980 Ti. It also remains to be seen how long this industry-standard utopia lasts, and whether GPU vendors find it at odds with their commercial interests. Multi-GPU optimization is something both AMD and NVIDIA spend a lot of resources on. It remains to be seen how much of those resources they'll be willing to put on a standardized multi-GPU tech, and away from their own SLI/CrossFire fiefdoms. Read the insightful article from the source link below.



View at TechPowerUp Main Site
 
This is the best games news since DX9 now you can put your old cards to use, excellent Microsoft and now that windows 10 is free no one can complain that DX12 is only for windows 10 only.
 
i don't think they will keep this
 
I wonder what the 'mechanics' are? The 980ti is the fastest single option in AoS now but it does not do as well when it is the primary card in the pairing compared to the FuryX as the primary card. It's got nothing to do with sli or XDMA as obviously the pci-e interface is being addressed directly by DX12? (Instead of an sli bridge)
 
i don't think they will keep this

If you mean AMD, I very much doubt they'll block it. If you mean NVIDIA, it would seem abundantly obvious that they will, given their long history of anti-competitive practice and the fact that the mixed setups make it even more clear that AMD scales much better in multi-gpu scenarios, whether AMD only or mixed.
 
I wonder what the 'mechanics' are? The 980ti is the fastest single option in AoS now but it does not do as well when it is the primary card in the pairing compared to the FuryX as the primary card. It's got nothing to do with sli or XDMA as obviously the pci-e interface is being addressed directly by DX12? (Instead of an sli bridge)

Probably something to do with the Fury being the master and the 980Ti card being the slave. It's been the case for a very long time that CF outperforms SLI. The gap appears to have widened rather than closed, particularly vis-a-vis frame times .... which were NVIDIA's one prior advantage, but where AMD now performs much better than NVIDIA.
 
This is the best games news since DX9
this is indeed very good news, after the virtugpu let down, i just hope sides dont break the compatibility and/or hinder performance on purpose from here on out while claiming that "it was never supported"
 
Probably something to do with the Fury being the master and the 980Ti card being the slave. It's been the case for a very long time that CF outperforms SLI. The gap appears to have widened rather than closed, particularly vis-a-vis frame times .... which were NVIDIA's one prior advantage, but where AMD now performs much better than NVIDIA.

Yes but is a lot of this not to do with the XDMA implementation by AMD? The crossfire situation of older cards (Tahiti and beyond) is not so great. So you mean the XDMA pathway Fury uses (as being from Hawaii and upwards) means it communicates better with the API or better with the Nvidia card? I suppose that makes sense if the Nvidia card isn't normally communicating with another card via pci....
 
who cares it works for now if Nvidia doesnt block it this is a great way to get AMD and Nvidia game effects working. Gameworks title no problem you have an Nvidia gpu... DX12 with asynch no problem you got an AMD card lol. What saddens me is they didnt show pure SLI and Xfire for the comparison. Still 7970 + 680 was interesting with the 680 as primary performance was worse than single card. Put the AMD card as primary performance scales wonderfully. Now I am sure Nvidia will block this somehow and it will somewhat depressing when they do but for the time being its quite interesting. Lets see if SLI / Xfire works better than DX12 mixed multi gpu
 
The problem is game developers have to support it and Microsoft isn't going to pay anyone to do that. AMD will pay for Crossfire and NVIDIA will pay for SLI though.

It's a nice experiment but it comes across as impractical for your average title.
 
I wonder what is the incentive for Nvidia or AMD to cooperate with this and try to make it successful. If you're Nvidia then you want people to buy two Nvidia cards. If you're AMD then you want people to buy two AMD cards.
 
except in this case Fury X and Fury are a valid xfire pair. And 980Ti + Fury x is better than AMD Xfire..... so... yeah.
 
curious to know how this would work with lower end cards
 
How much effort did MS put into this novelty feature that hardly anyone will use. Or maybe it was the by-product of other advances. Right, article, I should go read it ...
 
except in this case Fury X and Fury are a valid xfire pair. And 980Ti + Fury x is better than AMD Xfire..... so... yeah.

Obviously...both cards are faster than a regular fury LOL.

And we're talking about an alpha game with never used DX features on immature drivers. I'm pretty sure any testing is for shits and giggles b/c it's interesting. Not b/c it has any new information. We already know that Nvidia cards bog down under heavy load more compared to AMD. There's no new info except that the DX12 craziness actually works.
 
Obviously...both cards are faster than a regular fury LOL.

did you fail to read? FURY X + FURY AMd has no problems allowing the same gpu but different tiers operate in Xfire its been this way for years example 7950 + 7970 . Nvidia does no allow that regardless. DX12 Fury X + 980Ti is faster than 2x AMD gpus in proper X-fire.

Fury X + Fury shows 66% gain over Fury X by itself

Fury X + 980Ti shows 75% gain.

This is proprietary tech vs a DX12 feature aka agnostic and the agnostic option is kicking the crap out of said proprietary solution.
 
did you fail to read? FURY X + FURY AMd has no problems allowing the same gpu but different tiers operate in Xfire its been this way for years example 7950 + 7970 . Nvidia does no allow that regardless. DX12 Fury X + 980Ti is faster than 2x AMD gpus in proper X-fire.

How is that relevant? That's an artificial limitation. It has nothing to do with the performance of the cards.

This crap scales decently well (we already know xfire/sli is all over the place) and the slower cards lost. Big surprise. Two fury Xs would win. Big surprise.
 
I think the fact Lucid got bought and dissappeared is the problem they dropped off the face of the earth basically well as far as anyone is concerned. Even then how did Microsoft get the tech? Did they license it? Did they create their own?
 
I guess people forgot about Lucid Hydra. I wonder what happened to Lucid?
they dropped off the face of the earth
motherboard manufactures stopped including their hydra soc and all the hype started to die down, which isnt so far fetched, after all it required custom hardware to work. now they offer a software solution to use discrete gpus to aid in rendering tasks performed by an integrated gpu, the virtu mvp, which seems to work "ok" on laptops and games that support it
 
Last edited:
did you fail to read? FURY X + FURY AMd has no problems allowing the same gpu but different tiers operate in Xfire its been this way for years example 7950 + 7970 . Nvidia does no allow that regardless. DX12 Fury X + 980Ti is faster than 2x AMD gpus in proper X-fire.

Fury X + Fury shows 66% gain over Fury X by itself

Fury X + 980Ti shows 75% gain.

This is proprietary tech vs a DX12 feature aka agnostic and the agnostic option is kicking the crap out of said proprietary solution.
I don't think they used crossfire for this. You have to enable crossfire in AMD drivers. Which means this is just running on a software layer in dx12 the same as every other combinations.
 
I think the fact Lucid got bought and dissappeared is the problem they dropped off the face of the earth basically well as far as anyone is concerned. Even then how did Microsoft get the tech? Did they license it? Did they create their own?

Microsoft works with AMD since the XBOX One, AMD already has a similar technology that enables CrossFireX without a bridge. Probably Microsoft used based on the AMD technology and developed compatibility.

It has been said that in DX12 would have many functions that were previously Driver (AMD or Nvidia) would be transferred to the API (Dx12), it shows that Microsoft will have more freedom on the use and operation of hardware in DX12, which is simply FANTASTIC, it means more access to the Hardware and less dependence on drivers.
 
Well if this simply means I could keep one of my 460s to run Physx and then get a 390 to run performance stuff...happy there. Course nVidia needs to get off their butts and actually release Win10 drivers for Fermi like they said they would. Drivers which brick the cards do not count.
 
Back
Top