Discussing GCN Architecture, Performance & The Future

[LtMatt]

Very interesting article here, well worth a read.

Article discusses GCN, AMD CPU/APU performance as well as features like TrueAudio/Mantle/TressFX.

I've highlighted a couple of points made about Mantle, but this is not the theme of the article, hence not in the Mantle thread. Just making that clear for those in doubt.

What were the factors prompting AMD to create the Mantle API? Was it simply a case of helping developers solve the major weaknesses of PC gaming architecture (the overhead) in preparation for the next generation of consoles and game engines?

Game developers did the rounds in the industry, asking all of the hardware vendors with a stake in graphics for a solution to make PCs more “console-like” with respect to hardware utilization efficiency and programming simplicity. They recognized that PC gaming could learn a lot from its siblings in the living room. Only AMD took these requests from the negotiation stage to the manpower and money phase, and Mantle was born!

Regarding Mantle’s performance, it was mentioned at CES that the performance increase with Battlefield 4 was up to 45 percent. Will this sort of performance boost be the normal for titles under Mantle?

Every game is different, of course, but we’ve said all along that we didn’t undertake the Mantle project to chase two or three percent performance improvements. With each game, we’ll be publishing a full set of data and we’re confident people will be pleased with the uplift.”
Even if the performance of the Mantle API isn’t up to 45 percent per game, and only works out to be half as much, that’s still going to be a significant difference. This can turn a title barely running in 30FPS into the comfortable to play territory, or titles which are just shy of the target of 60 shooting way past that number. Obviously it’s still too early to make judgement calls quite yet, but for titles that have AMD’s Mantle API available, Nvidia could be in for a fight.
I’m hoping there’ll be news of Nvidia embracing the technology, or some alternative to start to resolve this.

Audio technology on PC’s has been fairly stagnant for a number of years, and requires high CPU overhead for processing. Does AMD’s TrueAudio processor take care of all of the processing work with audio, or is there some processing left for the CPU to perform on the audio?

AMD TrueAudio fully offloads the CPU if the developer maximally utilizes its capabilities. It’s a programmable audio pipeline, so it will only shoulder as much of the burden as it’s told to. Regardless, AMD TrueAudio sends the voice(s) it’s processing and rendering to the user’s existing audio hardware as a passthrough to the endpoint device (e.g. headphones).

Full Article
Part1:http://www.redgamingtech.com/exclus...n-architecture-performance-the-future-part-1/

Part2:http://www.redgamingtech.com/exclus...tecture-performance-the-future-part-1-page-2/

Part3:http://www.redgamingtech.com/exclus...ntle-performance-rop-gpu-steam-machines-more/


So its up to Nvidia to embrace Mantle it seems.

 

[~Mike~]

You won't get a ' discussion ' on this forum.

Good read though.

 

[Gregster]

This has been covered in a couple of threads including G-Sync threads and Mantle threads.

 

[LtMatt]

You won't get a ' discussion ' on this forum.

Good read though.

That was the title of the article to be fair, but one can hope.

 

[Silent_Scone]

To your point, compute is more important now than ever. Consider the following effects used by game devs: high definition ambient occlusion (HDAO), global illumination, FXAA, MLAA, TressFX Hair, diffusion depth of field, and contact hardening shadows. What do they have in common? They’re all GPU compute-driven effects. That’s just a small selection of the effects you can accelerate with compute resources rather than the traditional graphics pipeline. Where possible, using GPU computer is a rather efficient way of rendering!

I take it he’s obviously talking about Mantle making use of Compute? Seeing as there’s no benefit from using it with DX (currently)/

higher pixel rates for UHD content, off-chip buffering improvements to enhance tessellation performance, more robust data storage in the geometry processors to improve geometry shader performance, a smaller and more efficient memory controller, the addition of our “XDMA” technology, support for up to four primitives per clock, and of course we were able to scale Graphics Core Next out to 2816 total shader units. Overall, though, this is the basic GCN Architecture we know and love, but with an extra ounce of love to make it a meaner and more capable engine for multi-purpose work. (note, GCN 1.1 compute is similar to PS4

So with any luck the R2 series will match / exceed Kepler with tessellation in Mantle . I’m interested to see what they can do with the extra compute performance compared with the PS4.

Let’s assume I’m a games developer and I’m in the middle of creating a game for the PS4. There are reports that Mantle is similar to the Playstation 4′s API. Since the PS4 uses the GCN technology, Is it relatively easy for me to port my game to the PC from PS4 using the Mantle technology, particularly if I were using an engine like Frostbite 3 or say Unreal?

Misinformed question, PS4’s OPGL based API already has low level access to GCN arch and has nothing to do with Mantle. The GCN arch is where the link ends between the two. I'd like to see where he got his report.

I, along with many others in the industry feel it was a great move by AMD to allow both Nvidia and Intel owners the chance to use TressFX. Although a very different beast, could you clear up how Nvidia would be able implement Mantle if they wished to support it?


Robert Hallock: It wouldn’t be prudent of me to comment on behalf of other companies.

Lol . With difficulty.

Could you offer any comments regarding FreeSync technology and how it functions?
Robert Hallock: I’ll have to get back with you soon. We only just made the announcement at CES, and I actually have to learn a little more about the technology myself.”
Because the interview took place pretty much a day after the CES event, much of the technology behind Free-Sync is still being worked on. AMD are currently in working on ways for it to work with Desktop monitors, and are making progress. See news here. They’re not quite ready to speak about this yet.

Awh.

From my own perspective as a PC gamer, I believe the belief that PC gaming is more hassle is one of the core problems that must be tackled. With both AMD (using thei Gaming Evolved app) and Nvidia with their Geforce experience App helping to pre-configure games, it’s becoming easier. The problems with beginners is that they often don’t understand how to best configure the graphics options with games, and that puts them off.
The other issue is hardware – inexperienced gamers just don’t know what to buy. Fortunately, there are less ‘lemon’ products out there than there used to be. Back several years ago, it was fairly common practice to sell GPU’s with high amounts of slower RAM. These cards would seem like bargains, often with double the RAM, but the customer didn’t know the memory was often extremely slow, crippling the cards performance due to bandwidth constraints. Fortunately this is less of a problem now, but pre-configured PC are still somewhat hit and miss when it comes to the value to the customer. Hopefully with standards set by Valve’s Steam Machines, we’ll see some of this getting resolved.

Some good points, most interestingly the fact they’ve removed pacing issues with XDMA via the PCI-E 3.0 bus. Why they didn’t do that years ago I have no idea, but it’s good that it’s finally been resolved. A lot of vague remarks but still interesting, and hopefully Mantle will speak words for itself eventually. I’m still waiting!

So its up to Nvidia to embrace Mantle it seems.

Mhm, no that's not the aim at all. They cannot, and won't. The fact it's suggested they can only implies that Mantle isn't as 'down to the core' as they're implying. Which I'm not doubting it is, but I think AMD know full well Nvidia cannot use this without making some drastic changes to their GPUs. And with the greatest of respect Matt, you need to stop the double standard opinion of how Mantle really is lol. I say that as just-as enthusiastic about Mantle as you are believe me, but Mantle definitely does not have an open door policy, it is a smoke screen.

 

[pgi947]

The way I look at nvidias stance on mantle is gsync, one company is trying to get lower frames up and the other is smoothing out the lower frames. One is offering a free solution with at present a limited amount of game support, the other offering a priced premium and an almost entire back catalogue of support.

I don't see nvidia embracing mantle either.

 

[LtMatt]

Mhm, no that's not the aim at all. They cannot, and won't. The fact it's suggested they can only implies that Mantle isn't as 'down to the core' as they're implying. Which I'm not doubting it is, but I think AMD know full well Nvidia cannot use this without making some drastic changes to their GPUs. And with the greatest of respect Matt, you need to stop the double standard opinion of how Mantle really is lol. I say that as just-as enthusiastic about Mantle as you are believe me, but Mantle definitely does not have an open door policy, it is a smoke screen.

Well you obviously know more than Johan Anderson who worked alongside AMD to create it. He's on record as saying the aim is for Mantle to become open. Who to believe, forum user frosty or technical director who helped create Mantle? Hmmm...

Johan Andersson: I agree, yeah, and, also the vision for Mantle - well it can play out in many ways. You can't really say for sure but the vision for Mantle is that it can become an open API also and that's something we're completely open for. It's something AMD is open for also, it's just that that's not the right way to go at it initially...

Source
http://techreport.com/news/25651/ma...ite-games-dice-calls-for-multi-vendor-support

But the "pink elephant in the room," as he called it, is multi-vendor support. Andersson made it clear that, while it only supports GCN-based GPUs right now, Mantle provides enough abstraction to support other hardware—i.e. future AMD GPUs and competing offerings. In fact, Andersson said that most Mantle functionality can work on most modern GPUs out today.

Source
http://www.tomshardware.co.uk/johan-andersson-battlefield-4-interview,review-32839-8.html

 

[andybird123]

could and will are two entirely different things, since this is covered hundreds of times over in the main mantle thread I really fail to see why we needed a new thread, it could have been contained to the wretched hive of scum and villainy that is the mantle threadnought

 

[LtMatt]

could and will are two entirely different things, since this is covered hundreds of times over in the main mantle thread I really fail to see why we needed a new thread, it could have been contained to the wretched hive of scum and villainy that is the mantle threadnought

The article is not just about Mantle though if you'd bothered to read it before passing judgement, so yes it does deserve its own thread. Those were just two questions i picked out but there are lots more about GCN tech/features as a whole.

Nice to see another Nvidia user looking forward to Mantle though.

 

[Silent_Scone]

Well that is half the battle isn't it. With you at least...

You seem to believe every single thing you read. Is Johan Anderson qualified to comment? Barely any more than I am. Would he have the foggiest how to programme drivers? NO! He wouldn't. And trust me, he wouldn't...

You're looking at it as if I am saying to you "Mantle is closed". No it's not closed, nor is it an open standard yet. I've no doubt it will be in some fashion eventually for others to use at their own discretion. But what you're not realising is Johan Anderson is looking in the long term. This is something he's been wanting at an industrial level for quite-some-time. So yes, I'm sure he would very much like to see Mantle operating on other vendor hardware. But unfortunately that doesn't make him qualified to say it will be.

It sounds so easy to say out loud, just programme some instruction sets for your own hardware and you're in. But unfortunately in the real world that's not how it works (well it is, but it's not that simple). And speaking morbidly Mantle has far more chance of being dead in the water, than it has Nvidia adopting it anywhere near as primordially as AMD.

 

[Rroff]

Well you obviously know more than Johan Anderson who worked alongside AMD to create it. He's on record as saying the aim is for Mantle to become open. Who to believe, forum user frosty or technical director who helped create Mantle? Hmmm...

You mean the same guy who is behind one of the most broken high profile game releases in recent history? not sure I'd take him very seriously.

 

[LtMatt]

Well that is half the battle isn't it. With you at least...

You seem to believe every single thing you read. Is Johan Anderson qualified to comment? Barely any more than I am. Would he have the foggiest how to programme drivers? NO! He wouldn't. And trust me, he wouldn't...

You're looking at it as if I am saying to you "Mantle is closed". No it's not closed, nor is it an open standard yet. I've no doubt it will be in some fashion eventually for others to use at their own discretion. But what you're not realising is Johan Anderson is looking in the long term. This is something he's been wanting at an industrial level for quite-some-time. So yes, I'm sure he would very much like to see Mantle operating on other vendor hardware. But unfortunately that doesn't make him qualified to say it will be.

It sounds so easy to say out loud, just programme some instruction sets for your own hardware and you're in. But unfortunately in the real world that's not how it works (well it is, but it's not that simple). And speaking morbidly Mantle has far more chance of being dead in the water, than it has Nvidia adopting it anywhere near as primordially as AMD.

Well ill stick with the people who are most knowledgeable about this and im sorry but their opinion and thoughts hold far more sway with me than yours. Time will tell i guess. Considering he worked closely with AMD creating it, id say he has a pretty good idea about what it can do and has gone on record several times say it can support other and competing gpu's. Just that initially it will only work with GCN tech as Mantle is currently Beta.

You mean the same guy who is behind one of the most broken high profile game releases in recent history? not sure I'd take him very seriously.

It was EA's decision to rush the game out though in an unfinished state and that was to damage cod sales.

 

[ubersonic]

I think it has a very bright future as the current development has obviously been planned by AMD for a very long time, since before they even considered buying ATi, and with their red steamroller* building momentum they look set to knock Nvida/Intel/Microsoft for six unless they band together which is unlikely.

My predictions are that Nvidia will try and counter Mantle with something and fail then fall behind, have to embrace GCN or something but end up diminished, much like 3DFX in it's later years. Intel will not realise the trouble they are in until £500 Haswell-E CPU's are being matched in games by sub £200 AMD APU's and by the time they mount a counterattack (possible partnership with Nvidia?) it will be AMD's biggest rout since the Athlon XP. Microsoft being Microsoft will try and deal with mantle with FUD and Windows issues like they did with OpenGL, but gamers/devs today won't believe the FUD and the Windows issues will do more damage to Windows as a gaming platform than Mantle.

*A WW2 metaphor not a CPU related one.

 

[Silent_Scone]

I know I'm likely right, I'll happily admit when I'm not.

I'll just have to wait till the tabloids post it so you can quote it on here for your own benefit

My predictions are that Nvidia will try and counter Mantle with something and fail then fall behind, have to embrace GCN or something but end up diminished, much like 3DFX in it's later years.

If you honestly believe that, that's hilarious.

 

[Rusty0611]

The way I look at nvidias stance on mantle is gsync, one company is trying to get lower frames up and the other is smoothing out the lower frames. One is offering a free solution with at present a limited amount of game support, the other offering a priced premium and an almost entire back catalogue of support.

I don't see nvidia embracing mantle either.

Agree 100%. They're different approaches to essentially (at a high level) the same end point: a better gaming experience. Once it all becomes mainstream they can use the low level access on Mantle to push new boundaries on graphical settings similarly with G-Sync you can push new boundaries and have low FPS "feel" like a much higher frame rate.

Lads: I have to be honest - finding this brand level baiting / defending extremely tiresome to read as it's spiralling into multiple threads now. Can you all just agree to disagree and stop the remarks which are bound to incite multi-page rants backing up or flaming a GPU brand?

 

[Silent_Scone]

I dunno, there's a difference between ranting and trying to point out the bleeding obvious.

 

[ubersonic]

If you honestly believe that, that's hilarious.

It's a solid prediction, the day AMD bought ATi, Nvidia became the number two fish in the GPU market, however as they were still "winning" they never seemed to acknowledge it in their business practice. Now AMDs advantage is shifting up a gear and Nvidia will most likely stay true to form and keep their head in the sand while everything kicks off around them (like 3DFX when ATi/Nvidia were overtaking them), then scramble to do something about it but be running uphill at that point.

I am not saying that it will happen, just that it's looking likely due to what AMD is doing and the way the other parties will most likely react (based on previous actions).

 

[uksoldierboy]

You mean the same guy who is behind one of the most broken high profile game releases in recent history? not sure I'd take him very seriously.

To be fair i can't remember the last time it crashed for me, got to be over a month at least.

 

[Silent_Scone]

It's a solid prediction, the day AMD bought ATi, Nvidia became the number two fish in the GPU market, however as they were still "winning" they never seemed to acknowledge it in their business practice. Now AMDs advantage is shifting up a gear and Nvidia will most likely stay true to form and keep their head in the sand while everything kicks off around them (like 3DFX when ATi/Nvidia were overtaking them), then scramble to do something about it but be running uphill at that point.

I am not saying that it will happen, just that it's looking likely due to what AMD is doing and the way the other parties will most likely react (based on previous actions).

You're vastly overestimating Mantles success. Given the current climate, that whole analogy is properly flawed. How you can possibly come to that conclusion from a list, (albeit promising in content) drip fed over the next two years? You're talking 2 to 3 years before Mantle best case has half the platform covered, then possibly half that again before NV would contemplate adopting it as a base platform. More likely as we're seeing already, OpenGL will have more coverage given SteamOS. Also being that it's controlled and maintained by Intel, AMD, and Nvidia etc via the Khronos Group, so makes far more sense for Nvidia at least, to use this truly open library. Instead of one which is currently controlled by AMD only. And I imagine this factor will not change. How that means they're going to suffer I don't know...

I'm not even having a dig at Mantle as a technology, which is why I find the blow back from some of you hilarious. It doesn't reflect badly on the technology at all. But Matt stated that basically NV have no choice but to adopt, which is pure nonsense.

I'm pro Mantle, just not pro fan talk - if it doesn't ring true lol.

For the record my remark about OpenGL is not suggesting it will overtake Mantle as an API - just a more likely route for NV, again I think personally that is obvious. Also Uber, you're talking about something that happened almost 2 decades ago. I don't know how else to respond lol.

 

[LtMatt]

But Matt stated that basically NV have no choice but to adopt, which is pure nonsense.

No i didn't. I said its up to Nvidia to embrace it and i got that from interview in the OP. If Nvidia don't embrace it, then it won't work on their gpu's end of. The only person that will suffer in this situation is Nvidia gpu users. I never said they had to embrace it or they will go bust or whatever, far from it. I don't think they will embrace it. The point is and was that the long term plan for Mantle is not to lock Nvidia out from using it.The stumbling block in all this will likely be Nvidia themselves.

Spoiler

Every game is different, of course, but we’ve said all along that we didn’t undertake the Mantle project to chase two or three percent performance improvements. With each game, we’ll be publishing a full set of data and we’re confident people will be pleased with the uplift.”
Even if the performance of the Mantle API isn’t up to 45 percent per game, and only works out to be half as much, that’s still going to be a significant difference. This can turn a title barely running in 30FPS into the comfortable to play territory, or titles which are just shy of the target of 60 shooting way past that number. Obviously it’s still too early to make judgement calls quite yet, but for titles that have AMD’s Mantle API available, Nvidia could be in for a fight.
I’m hoping there’ll be news of Nvidia embracing the technology, or some alternative to start to resolve this.

Johan Andersson: I agree, yeah, and, also the vision for Mantle - well it can play out in many ways. You can't really say for sure but the vision for Mantle is that it can become an open API also and that's something we're completely open for. It's something AMD is open for also, it's just that that's not the right way to go at it initially...

Source
http://techreport.com/news/25651/ma...ite-games-dice-calls-for-multi-vendor-support

But the "pink elephant in the room," as he called it, is multi-vendor support. Andersson made it clear that, while it only supports GCN-based GPUs right now, Mantle provides enough abstraction to support other hardware—i.e. future AMD GPUs and competing offerings. In fact, Andersson said that most Mantle functionality can work on most modern GPUs out today.

Source
http://www.tomshardware.co.uk/johan-andersson-battlefield-4-interview,review-32839-8.html