• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

DirectX 12 Will Allow Multi-GPU Between GeForce And Radeon

Is this the same as that weird multiGPU thing that was touted a few years back but never got off the ground? I can't even remember it's name.

[Edit] Ah, Lucid Vertu. Remember the promises from that? :D
 
Last edited:
Sin_Chase said:
Never going to happen.

Technical possibility means nothing.
Click to expand...

+1

If you reduce all the matter a person is made of into pure energy, the resulting bang would be enough to take out the USA several times over. It is theoretically possible but it is not going to happen either is it.:D
 
Sin_Chase said:
Never going to happen.

Technical possibility means nothing.
Click to expand...

Yup, it was doable before, it will be doable with DX12, but no one will actually bother doing it properly because there is really no incentive or reason to do so. Even in the best case you'll have different hardware with different strong points and optimisations, it's too much work. You can do asych sli/xfire but again it's just not worth it. sli/xfire bring a lot of benefits but to focus the work on a smaller subset of situations which have greater results makes the work worth the effort. Spreading this work to dozens of different situations just isn't worth while, async xfire/sli is barely a thing(I think xfire does it... a bit... but no one really bothers, not sure SLi does it at all).

The simple assumption that each and every card has the same basic components, capabilities and horsepower simplifies the work done for xfire/sli compatibility so far that it makes the benefit/work ratio worthwhile.

Most things are technically feasible, but this is one that won't see any meaningful work put into it outside of maybe some niche company, kickstarter type nonsense or just some amateurs looking for a new project... nothing will come of it in the end.
 
drunkenmaster said:
Yup, it was doable before, it will be doable with DX12, but no one will actually bother doing it properly because there is really no incentive or reason to do so. Even in the best case you'll have different hardware with different strong points and optimisations, it's too much work. You can do asych sli/xfire but again it's just not worth it. sli/xfire bring a lot of benefits but to focus the work on a smaller subset of situations which have greater results makes the work worth the effort. Spreading this work to dozens of different situations just isn't worth while, async xfire/sli is barely a thing(I think xfire does it... a bit... but no one really bothers, not sure SLi does it at all).

The simple assumption that each and every card has the same basic components, capabilities and horsepower simplifies the work done for xfire/sli compatibility so far that it makes the benefit/work ratio worthwhile.

Most things are technically feasible, but this is one that won't see any meaningful work put into it outside of maybe some niche company, kickstarter type nonsense or just some amateurs looking for a new project... nothing will come of it in the end.
Click to expand...

Exactly, just like VRAM sharing, looks nice on paper, but in reality there isn't all that much data which can effectively be shared.

Maybe some new novel and clever tricks will come one day from the possibility, but I'm not going to hold my breath.
 
SkeeterUK said:
If true then would physx still be disabled cos it detected a amd card in the system? Otherwise whats the point of diff gpus on the same system if not to use the benefits of features of both at the same time? Sharing power for fps and memory is what sli or xfire was for wasnt it?
Click to expand...

nVidia has been moving much of the PhysX functionality into FleX which isn't vendor locked and infact can be run on AMD GPUs.
 
Kaapstad said:
+1

If you reduce all the matter a person is made of into pure energy, the resulting bang would be enough to take out the USA several times over. It is theoretically possible but it is not going to happen either is it.:D
Click to expand...
One fat american and you could take out canada along with the usa.:D
 
aren't amd doing a lot of work with windows to help make dx12? can't see them wanting involvement in that.. all they'll do is allow current amd owners to make their next purchase an nvidia card
 
According to Toms Hardware there is a new thing called Split Frame Rendering... only it's been around for more than 10 years and is related to the origin for the name 'SLI'. Indeed they wrote about it in 2005 though they failed to name the nVidia equivalent (which was actually called Split Frame Rendering back then, while ATI called it Scissor Frame Rendering)

*slow hand-clap*

Edit: So further down the article they suddenly remember that it's nothing new. Weird.
 
Last edited:
Microsoft do have form for this - with Vista (if memory serves) they changed how drivers worked so you could have more than one GPU type in a PC without Windows having a fit. Currently I'm running 3 screens off an AMD and 1 off Intel, so it definitely works. This is also the reason why it's technically possible to run an AMD for display and an Nvidia for hardware PhysX (Nvidia block it, but it's technically possible and I ran it for a year or so with modified drivers), or why it should be possible to run Nvidia for display and AMD for TrueAudio.

I think it's going to be correct that DX12 will allow multi-GPU across different vendors, but developers targeting the 90% of PCs and AMD/Nvidia themselves will ensure that no games actually use this technology in the real world.

Still, fair play for Microsoft to make it possible, just a shame the people they're putting it into the hands of won't make it actually happen.
 
We were also told that DirectX 12 will support all of this across multiple GPU architectures, simultaneously. What this means is that Nvidia GeForce GPUs will be able to work in tandem with AMD Radeon GPUs to render the same game – the same frame, even.
This is especially interesting as it allows you to leverage the technology benefits of both of these hardware platforms if you wish to do so. If you like Nvidia's GeForce Experience software and 3D Vision, but you want to use AMD's TrueAudio and FreeSync, chances are you'll be able to do that when DirectX 12 comes around. What will likely happen is that one card will operate as the master card, while the other will be used for additional power.
Click to expand...

Is this the point where we can run an AMD R9 390X as the main card and a GTX 950 as a PhysX card?
 
David Bisset said:
According to Toms Hardware there is a new thing called Split Frame Rendering... only it's been around for more than 10 years and is related to the origin for the name 'SLI'. Indeed they wrote about it in 2005 though they failed to name the nVidia equivalent (which was actually called Split Frame Rendering back then, while ATI called it Scissor Frame Rendering)

*slow hand-clap*

Edit: So further down the article they suddenly remember that it's nothing new. Weird.
Click to expand...

I am also perplexed as to the point/message of the article. Are they supposed to be saying it's a _NEW_ SFR?
 
You must log in or register to reply here.
Back
Top Bottom