![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
The article isn’t there anymore… it was removed.
Here's an archived capture from the Wayback Machine: https://web.archive.org/web/2019100...force_partner_program_impacts_consumer_choice
Intel finally managed to get one of the bigger ones overturned.
PhysX isn't Open Source only the free SDK is, all the stuff that actually makes PhysX work is proprietary and is part of NVIDIA GameWorks. Also, they're up to 5.1 but the Github is at 4.1. Not every game that uses Unreal engine uses PhysX. Borderlands 3 straight up doesn't have it. In house or Havok pretty much took over. Doing a bit of Google'ing and the last game I found to use PhysX was Mount & Blade II: Bannerlord released in 2020. A game I've never heard of. PhysX was used extensively in the past but it's pretty much dead today.
Not a huge milestone for iOS devices to use it since the most extensive graphics seen on them is Among US. On Mac OSX you don't have many applications and the only 3rd party application I can think of that actively supports it is World of Warcraft which has been bought by Microsoft. "According to Apple, more than 148,000 applications use Metal directly, and 1.7 million use it through high-level frameworks." They mean MoltenVK when they say "high-level frameworks". How's that stopgap working out?
If you find mount and blade 2 on sale grab it, it’s a lot of fun. You want to find a lance and a nice horse, then run around with the lance couched mowing people down. It’s everything I should have done in my first divorce.
Slow learner?
From how I understand it, this prevents any "lock-outs" because the interface from the game engine to the graphics pipeline will have some standard hooks/api's.
Proprietary software works great for the moment you're using it. Give it 3 or 5 years later and none of those features in the game works on newer hardware with newer OS's. All this does is cause headaches for anyone trying to play these games because it turns out you need to download this obscure file from a website that'll probably give you the digital version of HIV while having to edit a registry entry to even get the game to boot. Even then it'll just crash every so often.
If it works, then great. All I know is wayland/mesa devs tried to work with nvidia on a standard API and nvidia wouldn't budge...still hasn't afaik.
It's amazing that you know so many of the finer details on something that's apparently nothing more than a concept and PR piece at this point. There's a number of ways that they could implement something like this and some are far from hardware agnostic.
I forgot that they actually released something* so calling it just a concept was inaccurate but the only other vendor that has said anything about using it hasn't released the hardware or drivers that would use it so the multi-vendor aspect is nothing more than a concept and we have no way of actually knowing if it will be agnostic in practice. I've seen a few articles talking about this and all of them seem to be based on a PR release by Nvidia and are very light on details which makes me doubt that anyone has taken the time to go through all of it and look for predictable issues much less the unpredictable or less obvious issues.
But so far no one else has brought forth any such solution. Games being able support multiple upscaling paths with little extra work will benefit gamers also (If one day 'AMD BetterScaling' is better than 'NV JacketScaling', or vice versa). If it can be done in engine (as said in this thread) then those developers should now know that it's possible to insert their own form of this. But I don't think NV "owns" this tech as people are suggesting anyway as it's just a way to call upon each brands tech. I think Nvidia is worried that FSR will get close enough to DLSS that devs will only need to support it which will render that branding obsolete; this will allow expand DLSS to be available in those "FSR only" games.
I'm not against the concept and I'd be happy to eat my words if this ends up being helpful and doesn't unfairly hinder their competition I'm just skeptical that nvidia would do something like that. I'm not convinced that it is bad but I've also seen nothing that convinces me otherwise and I trust nvidia less than most corporations which I already don't generally trust.
I get the vast skeptical, but I think Nvidia would certainly do something that help people having access to motion vector for their temporal upscaling (or motion blur and what not) for many reasons:
I use NVidia’s stuff and for a number of business reasons I am completely in their camp like it or not. But they deserve every ounce of shade thrown their way. Nvidia are assholes, and Intel may be dicks, but … <Team America rants>… Nvidia may be trying to come off as altruistic here but Intel is forcing their hand here, and that’s the only reason this is a thing.
Ok. And the same can be said for Frgmstr's post on [H], or any article we have yet read, or any post or opinion in this thread.
I never said that they couldn't benefit from a more open approach just that it hasn't been their modus operandi.
and actually, it's probably closer to 15 years or more.
"The first consumer graphics card featuring hardware tessellation that made its way to the market was the ATI Radeon 8500 in 2001."
Despite being designed by ATI together with JEDEC, the first card to use the technology was nVidia's GeForce FX 5700 Ultra in early 2004...
AMD contributed to Mantle, designed to improve upon DX11 by adding console-like API's. Mantle only ever worked on AMD GPU's and is long dead. Vulkan came from that later.
I gave several counter points including several within that decade you're now fixating on, your third point is simply restating what I already said(in the portion you quoted) and we all know where MS got large chunks of dx12 not to mention the motivation to actually release a new 3d api, you're right though I forgot to mention the mantle step in all that.
What? ATI had the first consumer card with hardware tessellation in 2001 as GoodBoy said, and AMD didn't buy ATI until 2006. AMD had fuck all to do with hardware tessellation. You seem to be confusing ATI for AMD; they are not the same.
The 5850 was announced in late sep 2009 which the first card with real *usable tesselation.
It is hard to innovate when your competitorS (both Nvidia and Intel) use dirty tricks that crush your sales, kills your revenue, prevents investing in R&D, that stifles innovation.
** Flame suit on***
I’m not sure about that, the time between AMD announcing Mantle and Microsoft launching DX12 was too short, even Microsoft can’t just shit out a fully functioning and documented API that fast. I’m thinking Microsoft was working on that behind closed doors well before AMD started mentioning Mantle.
Ghostwire: Tokyo uses PhysX, and that just came out.
I believe his claim was Mantle finally forced Microsoft into launching DX12. The fact that Microsoft was working on that behind closed doors and released it only after AMD announced Mantle still can be true.
DX12 was likely in production before Mantle was even a thing, leaked internal documents have Microsoft starting on it back in 2011. Saying Mantle forced anything is just too much of a stretch. If anything Mantle was a desperate attempt to create their own proprietary API to emulate some of Nvidia’s success with them, or some way to keep themselves in the news so people would remember they existed. The DX12 launch was timed pretty well to line up with their OS launch and was happening regardless of anything AMD was doing at the time.
All three worked together on Vulkan, GL5 was WIP long before Mantle/DX12, and once Mantle was announced and preliminary design was released, the DX and OpenGL working group took an interest and talked to AMD about integrating some of the design ideas into their own (at the time wip) APIs. GL5, at the time, was quite different. But it was early enough and designed in such a way that many parts of Mantle/Vulkan could be used there too (although, through a very different API). At some point in all this, AMD started working with the (now) Vulkan Workgroup on designing Vulkan, using Mantle (and maybe DX12 too) as it's inspiration.
I know Kronos announced their plans for a low level API in 2014, at that stage OpenGL4 was still pretty new. And at the time their announcement of wanting a low level API was believed to be more a response to Apple’s Metal than to Microsoft’s announcement that DX12 was coming. But the members of Kronos would have known they DX12 was going to be a low level API so in retrospect that belief doesn’t really hold up.
They had a lot of ideas they wanted to put into the next version, but kept pushing it back because there wasn't a whole lot of interest. Eventually someone finally said "we're doing it," and work began. Not sure when anymore.
Damn I feel old. LOL.
Fixed it for you.
I think you are correct. The "AMD forced Microsoft to release DX12" was something the news stories purported at the time. A more accurate statement, if there was any impact of one event on the other, would be "caused Microsoft to move up their timeline to release DX12".
To:
Try again? Show me 1 innovation AMD has created since 2007 (15 years), that wasn't a copy of something nVidia did first.