AMD's GPUOpen site is up

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Rvenger

Elite Member <br> Super Moderator <br> Video Cards
Apr 6, 2004
6,283
5
81
This is not AMD vs NVidia., this is regarding development tools. If you don't get back on track I am locking this one and calling it a day.

-Rvenger
 

Despoiler

Golden Member
Nov 10, 2007
1,966
770
136
@sontin
Not sure why you are pretending that NV is promoting its tech as open source. If you refer to the GitHub, it was last updated in 2014, with only a few older effects for D3D, and not the CURRENT GameWorks libraries which remain closed source, obfuscated for most developers. Devs who want a look at the raw source code have to sign up and pay a fee to be able to optimize it.

The OpenGL stuff is primarily for Android development, their Shield device is struggling to get exclusives.. and so few big PC games are OpenGL and when they are, they don't use GW.

Don't forget that the biggest incentive that AMD gives is that they are using the MIT license. That license allows zero restrictions on a dev's customized code including SALE of the customized code. That is huge. Sourced to a slide in below article.

http://www.anandtech.com/show/9853/amd-gpuopen-linux-open-source
 

Azix

Golden Member
Apr 18, 2014
1,438
67
91

Despoiler

Golden Member
Nov 10, 2007
1,966
770
136
The sale bothers me. It would mean that the modified code would see less adoption. Kind of hurts the benefits of GPUOpen.

I don't think it will end up like that. I see it more likely to be shared than to be put up behind a paywall. That is just the nature of where AMD has positioned itself. The allowance for charging is just for the coders that take the source and do serious amounts of work on it to the point that it might not look very much like the code they started from. In those cases I think it's reasonable to charge for your work. The fact that AMD chose the MIT license and not GPL or something else means they agree with that type of scenario. Also, the MIT license transfers with the modified code although I do not know if it only applies to the base source or the fully modified code. I would hope the latter so we don't run into situations where lawsuits are filed for people coming up with the same or similar solutions.
 

Essence_of_War

Platinum Member
Feb 21, 2013
2,650
4
81
I don't think it will end up like that. I see it more likely to be shared than to be put up behind a paywall. That is just the nature of where AMD has positioned itself. The allowance for charging is just for the coders that take the source and do serious amounts of work on it to the point that it might not look very much like the code they started from. In those cases I think it's reasonable to charge for your work. The fact that AMD chose the MIT license and not GPL or something else means they agree with that type of scenario.

You know you're allowed to sell GPL software, right?

https://www.gnu.org/philosophy/selling.html.en

What you can't do is distribute your software for money and then not provide source code for it.

Also, the MIT license transfers with the modified code although I do not know if it only applies to the base source or the fully modified code. I would hope the latter so we don't run into situations where lawsuits are filed for people coming up with the same or similar solutions.
The MIT license is extremely permissive. If you have a piece of MIT licensed code, you can freely take that code, incorporate it into a non-free-software product, distribute only binaries, and you'd be under no obligation to distribute any license or any type of source code whatesoever.

Here's a recent example:

The Sony PS4 actually runs a modified FreeBSD, itself distributed under the BSD license, which is very similar to the MIT license (lax, permissive, non-copyleft...). Sony certainly does not release source code for it.

Edit:

The choice of license is certainly strategic. MIT/BSD/Apache licenses are lax, permissive, non-copyleft licenses and are a great choice when what you want to do is get developers on-board to try things ASAP minimum hassle. They're especially good choices when proprietary software exists that fills a similar niche, and your free software will be competing with it. Some developers or companies might choose the proprietary software over the free software simply because the free software with, for example, a copyleft license puts requirements on them to redistribute source code. Stallman outlines some of these issues in a blog post about LGPL vs. GPL for libraries. Given that proprietary graphics libraries exist (Gameworks, etc), the MIT license seems like a good choice here.
 
Last edited:

Headfoot

Diamond Member
Feb 28, 2008
4,444
641
126
In my experience, the vast majority of companies won't even touch a library with any flavor of GPL, especially not v3, and not even LGPL many times. And customers won't buy your product if you have LGPL in it even if you've managed it correctly and you're in the clear. People just dont want the liability of any flavor of GPL. It's simply not worth the effort of managing it and the risk of screwing it up for many.

So if AMD wants people to actually build, sell and buy things on it, they made a good call using MIT.
 

Headfoot

Diamond Member
Feb 28, 2008
4,444
641
126
So, most of their "Open" techniques require their proprietary toolset like the DepthBoundsTest11:
http://gpuopen.com/gaming-product/depthboundstest11-directx-11-sdk-sample/

Wow really dude? Could you spread misinformation any harder than you already are?

"Proprietary toolset" like AMD GPU Services (AGS) library. Well no shoot Sherlock, you can't access GCN specific/AMD specific hardware and driver information through DirectX11. Thus they offer an add on library to expose those things.

Here's the description of this "proprietary tool" you're spreading FUD about:
AMD said:
This library provides software developers with the ability to query AMD GPU software and hardware state information that is not normally available through standard operating system or graphic APIs. The library includes support for querying graphics driver version info, Crossfire (AMD&#8217;s multi-GPU rendering technology) configuration info, as well as Eyefinity (AMD&#8217;s multi-display rendering technology) configuration info. AGS is available in dynamic library form for 32 and 64 bit versions of Microsoft® Windows® 7, Windows 8.1 and Windows 10. Latest version of AGS now exposes driver extensions.

How evil and proprietary. I definitely need source code access to this...
 
Reactions: kawi6rr

Essence_of_War

Platinum Member
Feb 21, 2013
2,650
4
81
So if AMD wants people to actually build, sell and buy things on it, they made a good call using MIT.

Yup. There are already existing proprietary software options, and companies have 0 trouble using them already, so you can't create any problems for the free software libraries if you actually want anyone to use them. MIT was the right choice here.
 

xthetenth

Golden Member
Oct 14, 2014
1,800
529
106
In my experience, the vast majority of companies won't even touch a library with any flavor of GPL, especially not v3, and not even LGPL many times. And customers won't buy your product if you have LGPL in it even if you've managed it correctly and you're in the clear. People just dont want the liability of any flavor of GPL. It's simply not worth the effort of managing it and the risk of screwing it up for many.

So if AMD wants people to actually build, sell and buy things on it, they made a good call using MIT.

Yep, it's the right choice, Ice-9 licensed code gets rejected out of hand in a lot of places.
 

Dribble

Platinum Member
Aug 9, 2005
2,076
611
136
http://motherboard.vice.com/read/glimpse-of-the-purehair-hair-rendering-engine-at-gdc

I note that "Pure Hair" in Tomb Raider 2 is modified TressFX (i.e. they took TressFX as a base and then made their own better version). I strongly suspect none of those changes will go back to any open source libraries but will be kept as Square Enix's intellectual property - it now belongs to them. AMD doesn't even get mentioned in relation to it in general TR2 advertising, and no other dev will benefit from that work. All we read about is the other nvidia additions which because they are directly using nvidia libraries they have to credit nvidia. In addition Tomb Raider 2 including that Pure Hair now runs better on Nvidia hardware (it probably won't eventually but Nvidia released a driver for it in time for all the reviews and AMD didn't so all the review benchmarks now show Nvidia looking stronger).

This is the reality of what AMD's open source becomes. Sure it's a nice building block for someone, but it's not going to get updated by dev's as they will keep the changes and it's not going to help AMD.
 
Last edited:

Essence_of_War

Platinum Member
Feb 21, 2013
2,650
4
81
Yes, this is the reality of lax, permissive, free and open source licenses. No one has to "give back" their contributions.

The real value in licensing their code under a permissive license is to try to get developers to actually use it, and prefer to develop for their open platform rather than, for example, Nvidia's closed one. As I mentioned above:

They're especially good choices when proprietary software exists that fills a similar niche, and your free software will be competing with it. Some developers or companies might choose the proprietary software over the free software simply because the free software with, for example, a copyleft license puts requirements on them to redistribute source code.

Edit:

And just to be very clear, the fact that no one is compelled to "give back" doesn't mean that no one will. There are certainly anti-patterns to fall into here, the classic one would be eschewing any sort of leadership and hope/guess/praying that either no one will notice, no one will care, or that someone will step up and lead for them, but if AMD's developers treat this like an ongoing project that they are the Benevolent Dictators of, this could end up being really cool.
 
Last edited:

zlatan

Senior member
Mar 15, 2011
580
291
136
I note that "Pure Hair" in Tomb Raider 2 is modified TressFX (i.e. they took TressFX as a base and then made their own better version). I strongly suspect none of those changes will go back to any open source libraries but will be kept as Square Enix's intellectual property - it now belongs to them.
It doesn't needed to port back the specific changes. One of the goal of GPUOpen is to make these effect fully accessible. Even if TressFX 3.0 is open source it is still a middleware level thing. It can be properly optimized for the actual rendered, but it can't be integrated deeply in the engine. This means AMD and their partners can use the R&D to make the effect better, but a simple integration may lead to high performance penalty. PureHair is the deeply integrated version of TressFX 3.0, and integration is the main goal to run this effect really fast.

AMD doesn't even get mentioned in relation to it in general TR2 advertising, and no other dev will benefit from that work.
Any dev can do the same thing what Square Enix does. They can download TressFX 3.0, and integrate it deeply to the engine.

This is the reality of what AMD's open source becomes. Sure it's a nice building block for someone, but it's not going to get updated by dev's as they will keep the changes and it's not going to help AMD.
PureHair is a very specific modification of TressFX 3.0. Even if Square Enix opens up the source it will be unusable for the other devs without the Crytal or Dawn Engine. If the other devs want TressFX 3.0 quality with PureHair speed, they need to this job by their own.
 
Last edited:

lilltesaito

Member
Aug 3, 2010
110
0
0
http://motherboard.vice.com/read/glimpse-of-the-purehair-hair-rendering-engine-at-gdc

I note that "Pure Hair" in Tomb Raider 2 is modified TressFX (i.e. they took TressFX as a base and then made their own better version). I strongly suspect none of those changes will go back to any open source libraries but will be kept as Square Enix's intellectual property - it now belongs to them. AMD doesn't even get mentioned in relation to it in general TR2 advertising, and no other dev will benefit from that work. All we read about is the other nvidia additions which because they are directly using nvidia libraries they have to credit nvidia. In addition Tomb Raider 2 including that Pure Hair now runs better on Nvidia hardware (it probably won't eventually but Nvidia released a driver for it in time for all the reviews and AMD didn't so all the review benchmarks now show Nvidia looking stronger).

This is the reality of what AMD's open source becomes. Sure it's a nice building block for someone, but it's not going to get updated by dev's as they will keep the changes and it's not going to help AMD.

I would say this is more of a good start for AMD, I do not think they care to much about being praised currently. I think they want to stop Gameworks and make their own Hardware look better. AMD has a long road ahead to catch up to Nvidia and this is just a stepping stone for them.

Now that Square has shown other companies what can come out of this, might start seeing more from other games.
 

Azix

Golden Member
Apr 18, 2014
1,438
67
91
It doesn't needed to port back the specific changes. One of the goal of GPUOpen is to make these effect fully accessible. Even if TressFX 3.0 is open source it is still a middleware level thing. It can be properly optimized for the actual rendered, but it can't be integrated deeply in the engine. This means AMD and their partners can use the R&D to make the effect better, but a simple integration may lead to high performance penalty. PureHair is the deeply integrated version of TressFX 3.0, and integration is the main goal to run this effect really fast.


Any dev can do the same thing what Square Enix does. They can download TressFX 3.0, and integrate it deeply to the engine.


PureHair is a very specific modification of TressFX 3.0. Even if Square Enix opens up the source it will be unusable for the other devs without the Crytal or Dawn Engine. If the other devs want TressFX 3.0 quality with PureHair speed, they need to this job by their own.

Makes sense. I think this is also the benefit of GPUOPen and things like Havok. Deep integration into the engine. I realized recently that a lot of the issue with gameworks would be that it cannot be made part of the engine like other tech. its going to almost always be an addon at least with the more demanding parts. Being able to make these things actual integral parts of the game engine is key imo.

They are doing that somewhat with unreal engine. But GPU physx, hairworks etc aren't likely to ever be that deeply integrated.

Could be worse if devs have to make sub-optimal changes to their code to get gameworks features to run right rather than being able to optimize in either direction.
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
So, people are surprised devs aren't giving their updated AMD code back for all to use? Devs are scum. They steal from each other and wouldn't credit their mothers for their birth unless there was a buck in it.

At least we have hair effects that work, and don't cause relatively huge performance hits between the IHV's. Isn't that what we want?

While I can understand the devs not wanting to credit someone else, I think it's sad we have gamers not wanting to credit AMD with doing something good for gaming. This is exactly what they wanted to happen when they released the library. They've also asked that nVidia and Intel do the same. Looking at the example of TressFX, who here doesn't want this to happen? Well, if you do, support it. If you don't support it no other company will see any benefit and will feel completely justified continuing down the closed source back stabbing course they are on now.
 

sontin

Diamond Member
Sep 12, 2011
3,273
149
106
At least we have hair effects that work, and don't cause relatively huge performance hits between the IHV's. Isn't that what we want?

Hairworks works. And at least there are more options then "looking bad" and "looking worse than TR2013".
I want quality. Otherwise i wouldnt care about "simulating" hair in games. BTW the alpha blend hair of the NPCs in Rise looks much better than PureHair. A shame that Lara's standard hair is worse than the one in Tomb Raider Legend... :\
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
Hairworks works. And at least there are more options then "looking bad" and "looking worse than TR2013".
I want quality. Otherwise i wouldnt care about "simulating" hair in games. BTW the alpha blend hair of the NPCs in Rise looks much better than PureHair. A shame that Lara's standard hair is worse than the one in Tomb Raider Legend... :\

Well, if you prefer wolves that look like a Pomeranian that's just had a shampoo and blow dry, and kills performance, that's up to you. Nothing wrong with the way her hair looks keeping in mind it's a video game.
 
Reactions: kawi6rr
Mar 10, 2006
11,715
2,012
126
So, people are surprised devs aren't giving their updated AMD code back for all to use? Devs are scum. They steal from each other and wouldn't credit their mothers for their birth unless there was a buck in it.

At least we have hair effects that work, and don't cause relatively huge performance hits between the IHV's. Isn't that what we want?

While I can understand the devs not wanting to credit someone else, I think it's sad we have gamers not wanting to credit AMD with doing something good for gaming. This is exactly what they wanted to happen when they released the library. They've also asked that nVidia and Intel do the same. Looking at the example of TressFX, who here doesn't want this to happen? Well, if you do, support it. If you don't support it no other company will see any benefit and will feel completely justified continuing down the closed source back stabbing course they are on now.

Wow, developers are scum, huh? Without their hard work, you wouldn't have anything to play.

The ADF is going overboard at this point, IMO.
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
Wow, developers are scum, huh? Without their hard work, you wouldn't have anything to play.

The ADF is going overboard at this point, IMO.

They get paid for "their hard work". And, I'm not defending anyone. Nor do I belong to ADF or any other AMD force. If you can't respond without making it personal maybe you shouldn't.
 

Elixer

Lifer
May 7, 2002
10,376
762
126
Today, AMD did a huge release of CodeXL 2.0, and made it open source as well.
They have a new Frame Analysis Mode, which will be interesting for those that love to benchmark.

https://github.com/GPUOpen-Tools/CodeXL

CodeXL is now part of the GPUOpen initiative. You can find all of CodeXL’s source code (barring a few parts that are IP-confidential) on the CodeXL GitHub project. Version 2.0 is also available in pre-built binary form, like previous versions.

We believe that by adopting the open-source model and sharing the CodeXL source base with the world we can help developers make better use of CodeXL and make CodeXL a better tool.

To encourage 3rd party contribution and adoption, CodeXL is no longer branded as an AMD product. AMD will still continue development of this tool and upload new CodeXL versions and features to GPUOpen.
 

Elixer

Lifer
May 7, 2002
10,376
762
126
AMD is pumping out these things, while nvidia only talks about open sourcing.

Today, AMD releases TressFX3.1
We are releasing TressFX 3.1. Our biggest update in this release is a new order-independent transparency (OIT) option we call “ShortCut”. We’ve also addressed some of the issues brought up by the community.
and FireRays 2.0,
A New Milestone

After the success of the first version, FireRays is moving to another major milestone. We are open sourcing the entire library which allows complete integration and contribution from the entire developer community to bring improvements in the rendering world. In fact FireRays 2.0 brings support for Windows, OSX, Linux, AMD, NV, Intel GPUs and CPUs with many back-ends.
and GeometryFX 1.2 – Cluster Culling
Today’s update for GeometryFX introduces cluster culling. Previously, GeometryFX worked on a per-triangle level only. With cluster culling, GeometryFX is able to reject large chunks of the geometry – with corresponding performance increases. Cluster culling is not a new idea – last year at SIGGRAPH, Ubisoft presented a GPU based rendering pipeline which incorporated cluster culling as well.
 

Elixer

Lifer
May 7, 2002
10,376
762
126
Here is another thing that AMD is opening up...
https://radeon-prorender.github.io/

It really is too bad that nvidia is all talk, they did say they were going to open things up, but, since they made that announcement, we see how "open" they are, which is to say, not much has changed with them.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |