Bondrewd
Veteran
Naaa.
Too expensive.
They do what ISVs tell them to do.
It's been like that since the advent of DX9 and it ain't gonna change.
AMD Execution Thread [2024]
- Thread starter del42sa
- Start date
Too expensive.
It's been like that since the advent of DX9 and it ain't gonna change.
Leadership in real-time graphics encompasses more than just pushing rendering features. What about graphics programming features like bindless, persistently mapped device local upload memory, PSOs, Work Graphs, and just as you mentioned more explicit APIs ?
Besides, what are AMD's competitors willing to do make ray tracing work for virtualized geometry (acceleration structure management) and more complex material (function calls/spilling) ? Are they willing to cement their "leadership" if it means taking a hit in contemporary benchmarks and perf/area ?
Bondrewd
Veteran
That's all on ISVs like MS to solve thru making DXR et al non-dogshit.
Like AMD's gonna bolt a traversal core and expose it thru the amdgcn ISA and...
...well you ain't doing much more than that.
Those things all count. If they actually follow through and deliver a better gaming experience. Work graphs can literally be a game changer.
Depends on when it becomes practical to do so. If AMD believes there is sufficient compute today to trace virtualized micro geometry I would love to see that demo.
I'm more convinced that it's a hardware problem and not a software problem ...
One of AMD's competitor clearly doesn't want to expose traversal shaders (perf loss) or implement fixed function BVH building hardware (more HW complexity) since either of those moves would put them out of their own comfort zone of being able to score easy victories against their rival with their tech demos even if it comes at the cost of leaving out technologies like Nanite ...
Nobody but Apple really wants to make spilling efficient for function calls even though it would make ray tracing play much more nicely with UE5's Substrate material system ...
Bondrewd
Veteran
Probably a bit of both but DXR is ancient and moved about ~nowhere since its inception in 2018.
Apple's always been a special case so they don't count.
DXR hasn't seen any meaningful functionality progression because hardware vendors are more interested in doing their own thing. AMD just wants the "rubber stamp", Nvidia doesn't really want to embrace virtualized geometry since it'll make it harder for them to keep doing their tech demos to differentiate themselves, and who knows how long Intel is going to stay in discrete graphics ...
Is this how it ends for hardware ray tracing ?
DavidGraham
Veteran
The world doesn't revolve around UE5 alone, especially as UE5 is still evolving, UE5 will come around. Epic is working on making HW Lumen much more performant and Nanite Ray Traceable.
UE5 in it's current trajectory is neglecting proper relfections, shadows and advanced global illumination. This is not sustainable if Epic wants engine leadership, which is why they are correcting course.
Bondrewd
Veteran
No.
It does, only Sonylands still run somewhat significant amount of first party engines.
Everyone loves bling but we operate with fixed hardware targets here.
As it is, Nanite is a far sexier offering than anything RTRT brings.
Crispy, crunchy texturework. Majestic really, UE5.2 Fortnite update made it look like a Pixar cartoon from 15 years ago.
DavidGraham
Veteran
Yes they did, AMD was first with a DX11 GPU (the 5870), and they marketed Tessellation hard with Dirt 2, Aliens vs Predator and STALKER.
AMD's Radeon HD 5870: Bringing About the Next Generation Of GPUs
They even made fun of NVIDIA for focusing on PhysX and not talking enthusiastically about Tessellation and DX11 during the 6 months period of Fermi delays.
Once Fermi was released and got revealed to be a Tessellation monster, AMD withdrew from the scene in silence and tried to cheat their way out by setting sepcial driver hacks to reduce Tessellation factors and lessen the performance impact on their hardware.
After that they struggled hard with it for too many generations, they only got good enough with GCN4 and Vega.
UE5 is very important but not sure it should dictate IHV long term priorities. There’s more than one way to skin a high poly cat.
It’s one thing for AMD to do the minimum required to keep Microsoft happy and not believe in a raytraced future. That’s not really the problem. The issue is they’re not sharing their alternative vision of that future. What do they have in store for everyone not using UE5?
It’s one thing for AMD to do the minimum required to keep Microsoft happy and not believe in a raytraced future. That’s not really the problem. The issue is they’re not sharing their alternative vision of that future. What do they have in store for everyone not using UE5?
Bondrewd
Veteran
But what else would?
MS always wants moar so that's quite not really the correct statement.
Moar CU.
Hundreds of them really. Just costs a lot of money.
DavidGraham
Veteran
There is Ubisoft running at least 3 major engines: Snowdrop, Anvil Next, Disrupt and Dunia. EA still has Frostbite, there is the 4A engine, Remedy has their Northlight engine, Rockstar has their engine, Bethesda has the Creation engine, Microsoft has idTech, Void, Flight Simulator and Forza engines. Sony has several important engines (Decima, Naughty Dog, Insomniac, Sucker Punches, Santa Monica, etc). Activision has their own engines, Blizzard too. IO has their own Glacier engine, Square Enix has their own engine for Final Fantasy games, Capcom has their excellent dedicated RE engine, there is also Source 2, Unity, Crytek and a plethora of in house custom engines, there is so much more than I dare to count.
To you maybe, others are captivated by the far more CGI look of Path Tracing.
Last edited:
Bondrewd
Veteran
Only Snowdrop is pushing the technical envelope since Massive are tech darlings ever since the OG Ground Control.
That's a corpse.
One game every 5-6 years is barely worth mentioning as far as IHV targets go.
One game every however many years and not really relevant.
Genuine article yeah.
Granted, they target consoles so what's even the point? lol.
lol
MS tailors their stuff to whatever the DirectX featureset they fancy. They are the ISV so doesn't count.
The only other relevant ones are Sony (it's Sony), and Epic, since UE5 coagulated the middle market AA-ish games at large.
that's console targets too.
COD.
Not really anything else.
Did you miss the recent article of Blizz devs from now shitcanned game begging the mgmt for Unreal? lol.
Dead.
It's all Unreal now since KH3 (FF16 used homebrew probably Crystal Tools derivative but that's a Yoshida one-off).
Point taken, but again, console target. They're Japanese.
Dead.
You know it's bad when Lottes left.
please don't talk like someone from NV marketing. please. do not capitalize things in a non-ironic manner like that.
Right now virtual geometry technology is in it's relative 'infancy' (under 2 years old) in comparison to ray tracing (over 5 years old) so it might not just be UE5 IHVs should be worried about when similar technology could come to other engines ...
DavidGraham
Veteran
Thanks for proving my point, the number of custom and studio specific engines outnumber UE5 may times over, it's naive to think that one engine (still in development) will dictate the trajectory of the whole industry.
UE5 with Nanite was released in early access in 2021. So 3 years now.
Bondrewd
Veteran
They have tiny engine TAM outside of Capcom stuff and Rockstar.
UE5 rules middle market and they're crawling all over the AAA space just as successfully.
That's the point.
It's like Unity on phones. They dictate what's true because phone middle market and big phone releases are all Unity.
Is it a fact that Vega was targeting the 1080ti instead of the 1080? Regardless, I still think RDNA 3 is worse off. DLSS alone makes it hard to consider an AMD GPU without a huge price difference.
Competition for Nanite and Lumen would be fantastic. After all Nanite is still a rasterizer and comes with all the usual limitations of rasterization. The holy grail is high resolution geometry alongside proper light transport. In terms of engine adoption there are far more engines supporting RT today than Nanite level assets. So question is will they all drop RT or will they evolve their RT implementations over time to work with higher resolution geometry.
Either way there’s no indication RDNA holds a natural advantage in UE5 or any potential virtualized geometry solution. It’s all cache and compute right and everyone has that. AMD needs a clear win to steal market and mind share.
DavidGraham
Veteran
No it doesn't rule anything, and it's was never the nature of UE to rule anything. UE3 and UE4 despite being used everywhere in lots of games never dictated anything, they never replaced other engines or pushed them away, and they never dictated how GPUs evolve in major ways. UE4 embraced ray tracing hard (most RT games are UE4 games), yet it never influenced anything. That was never how the industry works, I don't why some people suddenly think this way, but it's not true at all and has no historical precedence.
Similar threads
