AMD Mantle API [updating]

this does not change anything in my speech, not the stuff of fanboys here, if today were nvidia in partnership with DICE (or anyone else) to launch its own API, I'd do the same considerations. And speaking of fanboys, I can only see how AMD's ones suddenly become advocates and lovers of industry standards, in spite of the open ones.

I am not exactly sure which forum the ”I heard you liked the fanboy word so I put more fanboy in my post so that you can read all about fanboys as you read all about fanboys” attitude is coming from, but it needs to stay there, and not join us on B3D. Keep it away, no matter how strongly you feel about the people that like the IHV other than your favourite. I will also avoid the rather obvious ”pot calling the kettle black” observation.
 
Exactly. The existence of Mantle doesn't imply a worse experience for non-AMD users. However devs do have limited resources and Mantle is yet another thing they have to deal with...

One would expect that, until (if) it reaches critical mass, dev resources needed for this will not be exuberant, as AMD devrel will do a fair bit of legwork, and AMD will also sweeten the financial proposition of adopting Mantle. Not much different from things like PhysX or NVIDIA engineers porting stuff to CUDA themselves to persuade partners to adopt it, and rather reasonable if one wants to get people to adopt a non-portable, proprietary thingamebob.
 

The BF4/Mantle slides mention unused GPU functionality as a current problem, and that Mantle allows full access to the graphics hardware capabilities.

http://www.frostbite.com/connect/#battlefield-4-frostbite-mantle

Slides 13 and 18.

I hope all this bickering about Mantle wont result in like 2-3% gain compared to DX path.

Two different AMD people have gone on record to state that AMD wouldn't have wasted time on sub 5% gains.

I see a lot of discussions without real numbers. With those everything can be much clearer, that is telling us if Mantle is worth or not.

Repi, I know they were claiming 9x performances in drawing calls, but how this gets translated in a practical scenario? ie high end pc with frostbite3 in bf4
 
Last edited by a moderator:
I see a lot of discussions without real numbers. With those everything can be much clearer, that is telling us if Mantle is worth or not.

Repi, I know they were claiming 9x performances in drawing calls, but how this gets translated in a practical scenario? ie high end pc with frostbite3 in bf4


december 2013.
before that I dont expect any numbers out not even at the november conference.
 
Repi, I know they were claiming 9x performances in drawing calls, but how this gets translated in a practical scenario? ie high end pc with frostbite3 in bf4

Just to be clear, I'm not repi :p

I think it may not boost avg FPS in a very meaningful way (assuming a fast CPU), since they have to keep the number of draw calls low for the majority of PC gamers with non-GCN parts, but it should prevent dips in CPU limited scenes where there's a ton of stuff going on. That would be very cool. :cool:
 
december 2013.
before that I dont expect any numbers out not even at the november conference.
Incorrect

http://forum.beyond3d.com/showpost.php?p=1795291&postcount=252
sI329jq.jpg
iB9zn2Q.jpg
 
Last edited by a moderator:
To respond directly to your example, yeah I think console APIs provide a con for pc consumers. Forcing developers to divert resources away from their platform directly impacts the experience. I happen to think that trade off is more than worth it for developers, but there is a con (to some).
"Divert resources" implies that resources would have been available for PC development in the first place. How about "add resources"?
 
"Divert resources" implies that resources would have been available for PC development in the first place. How about "add resources"?

Where is this fantasy wonderland where companies actually add resources to do anything? :)
 
Just saw the nvidia hosted interview with Carmack, Sweeney and Andersson yesterday and was amused when guest asked Andersson about mantle. Now I'm hoping the same guy asks about G-sync on AMD's thing, haha.
 
You're approaching this too much as a developer. You don't think AMD fans were negatively impacted about msaa being locked to nvidia on batman?

Correct. It were Batman fans which have been "screwed over" by Rocksteady, who have been "screwed over" by their middle-ware provider, for not providing native & generic MSAA support.

If you think there was some Nvidia conspiracy going on, you might ask yourself who is really the sinner in this case, Nvidia or Rocksteady. Who defines and implements the engine's features?

What if MSAA would be a Nvidia only feature?

Proprietary APIs hurt consumers. That's not to say all proprietary APIs are bad or that Mantle is bad, but you're naive in saying they have no negative impact.

Proprietary APIs are completely okay, if a common baseline API exist. Having two different network protocols for different areas or devices as the only option is inconvenient - but having 500 million different network protocols each with it's own strengths and awesomenesses as switchable alternative to the baseline say TCP protocol is AAA-okay.

People are paying for distinctive services all the time. And if you would indeed be following market sense, then Mantle versions should cost $1 more. That Mantle versions will come out without added cost is actually a free distinctive service, which is only to rejoice about.

And btw, last time I checked customers are paying for your product; they can dictate whatever they want.

No, they can't. If a consumer has no contract with a game producer, then he can't. He can't dictate CryTek to make RPGs, he can't dictate Crystal Dynamics to make a side-scroller, he can't dictate DICE to make iPhone games. He can't dictate them to make the games accessible to the blind, which would actually pretty cool, I would love to hear high quality aural games.

Particular developers don't even go to the market and ask people what kind of game they would like and then do that. They just make the game they are either familiar with making, or they want to make for "personal" reasons.
In fact, lots of games come from individuals thinking about it for some time, and manage to find a sponsor to produce it, when the sponsor thinks it's marketable.
Monster companies like EA might look to the market and think it's time for another RPG, and then put one of their studio to do it. Which in effect might end up boring because of it.

Its relative and comparative, compared to amd users, nvidia users have less performance.

No it's not.
You don't agree that when you buy a 8bit monitor, connect it to a 14bit graphics card, and then getting told that the monitor producer who offers a 14bit patch for the monitor firmware is supposedly harming consumers (of 8bit monitors of other brands, or even the entire industry)!

Please. If you get only 50% of X because of Y, then of course you say Y is subpar or inefficient, and don't criticize the fact that someone is able to fix X by using Z instead!

Exactly. The existence of Mantle doesn't imply a worse experience for non-AMD users. However devs do have limited resources and Mantle is yet another thing they have to deal with...

No, they don't. They can be completely ignorant of the fact that Mantle exists and hell is not going to break loose.
 
I think it may not boost avg FPS in a very meaningful way (assuming a fast CPU), since they have to keep the number of draw calls low for the majority of PC gamers with non-GCN parts, but it should prevent dips in CPU limited scenes where there's a ton of stuff going on. That would be very cool. :cool:


They don't have to. Not for the Mantle version of BF4, at least.
They could just implement an "Epic Battle" mode that shows an enormous amount of NPCs/vehicles/debris and plays smoothly using Mantle + dirt-cheap APU or requires a $1000 overclocked CPU in a $300 motherboard to get the same results in the non-Mantle system.
 
So all other IHV's should pursue their own IHV-specific graphics API because "targeting another renderer is not that much of an overhead" and is a "selling feature"??? I'm sorry, but an IHV-specific graphics API such as Mantle is hardly a good long-term solution, even if it being opportunistically promoted by some in the short-term.

Targeting Mantle is not that much of an overhead because it:

1. Can reuse console code 'in whole or part (AMD)"
2. Already has a large GPU user base that can benefit from Mantle - the 7xxx cards.
3. Addresses both the GPU and CPU - all bulldozer based CPUs and APUs will benefit from Mantle.
4. GCN is a long term architecture. Mantle optimizations done today will work on AMD processors far into the future.
5. Mantle is designed from the start to be multi platform - Linux and Android can be easily accommodated.

With their x86 PC Mantle experience, DICE should be able to efficiently incorporate a Linux render path into Frostbite. Same should be true for any targeted platform. Or any engine developer incorporating Mantle.

EA made a point of saying they insisted Mantle be cross platform. AMD said Mantle was developed in consultation with several other developers (Nvidia might have rounded up the only two developers/engines that aren't currently on board with Mantle, Carmack and Sweeney. If there had been a third it is likely they would have been there instead of Johan Andersson).

With most of the major developers (and engines) on board and cross platform future proofing Mantle is looking very strong.
 
no i'm italian, it's the translation via google translator of "trovata commerciale", i don't speak perfect english so please suggest me the correct word in english, thank you.
gimmick =
A device employed to cheat, deceive, or trick
or
something designed to attract extra attention, interest, or publicity
the first one is the most common definition

3. Addresses both the GPU and CPU - all bulldozer based CPUs
since intel cpu's workings are well documented (i believe) amd could make it so it benefits intel cpu's
 
gimmick =
A device employed to cheat, deceive, or trick
or
something designed to attract extra attention, interest, or publicity
the first one is the most common definition


since intel cpu's workings are well documented (i believe) amd could make it so it benefits intel cpu's

Ok then what I meant, more in the second sense. Thank you very much, very kind ;)
 
Back
Top