No, the Oxide engine was not built for Mantle. Have you not seen any of the source material? Mantle was added to the engine. That was a fairly big deal in the presentations, how easy it was and how little work it was to add it in.
Just to correct you, i'm not sure if you're referring to the 290 series or the older cards. But, with the 79xx cards and prior, frame pacing is certainly not "fixed". It is still broken in crossfire DX9. It is also broken in eyefinity with crossfire. AMD managed to "Fix" it with the 290 series, but didn't fix it yet with the older cards. They state that they will fix it this month. Yet by the same token, keep in mind that this bug has existed for nearly 2.5 years. That certainly doesn't give you confidence in AMD's software team, does it? Especially since their competitor had this same issue solved 4-5 years ago with the elimination of nearly all microstutter.
I also am under the impression that crossfire will not work simultaneously with Mantle. I do remember reading that somewhere. That COULD be a good thing in AMD's case given their poor software support, but if anyone could clarify whether CF works with Mantle, that would be great.
I'm quite sure that I read that Mantle will be single GPU only. I could be wrong though.
I think he may had seen a slide that mentioned that under Mantle, AFR wasn't the only way to utilize the additional cards. They had the ability to have the cards work on a single frame in unique ways.I thought the problem always existed. nVidia fixed it with Kepler and then released the equipment they had to show it. At that point AMD started working on a fix.
do you recall where you heard that Mantle and Crossfire weren't compatible? I was actually under exactly the opposite impression, that Mantle allowed for Crossfire to be implemented in various ways. Possibly even with shared memory between the GPU's instead of having to mirror it on all of the cards.
No, the Oxide engine was not built for Mantle. Have you not seen any of the source material? Mantle was added to the engine. That was a fairly big deal in the presentations, how easy it was and how little work it was to add it in.
It goes even further back...remeber AMD's "conroe-killer"?
http://www.xbitlabs.com/articles/cpu/display/amd-k8l.html
AMD's PR has always been..."hyperbole"...
My guess is that they spiked to a 45% improvement momentarily in one area of one map for a brief moment. This legally permits them to say UP TO 45% and they would not be lying. Deceitful? Yes. Lying? No. My money is on that scenario.
It goes even further back...remeber AMD's "conroe-killer"?
http://www.xbitlabs.com/articles/cpu/display/amd-k8l.html
AMD's PR has always been..."hyperbole"...
Funny, how this easy and simple thing to add...still is MIA :thumbsup:
I thought the problem always existed. nVidia fixed it with Kepler and then released the equipment they had to show it. At that point AMD started working on a fix.
do you recall where you heard that Mantle and Crossfire weren't compatible? I was actually under exactly the opposite impression, that Mantle allowed for Crossfire to be implemented in various ways. Possibly even with shared memory between the GPU's instead of having to mirror it on all of the cards.
Microstutter will always exist from time to time. However, Nvidia has been using frame pacing tech, in software at least, since before Fermi. It has improved over time, however.Correct. Kepler introduced a hardware fix for micro stutter. Micro stutter still exist on pre-kepler GPU's.
This has to do with Mantle how?
I would however ask all of you to wait until we see something concrete (or at least asphalt) from AMD about mantle.
Yes, we've read ALL the marketing fluff, thanks. How easy and effortless Mantle is to add to any game engine. Devs will abandon DX or OpenGL, etc. etc. But I don't think I'll bite into these shenanigans. I would however ask all of you to wait until we see something concrete (or at least asphalt) from AMD about mantle.
You ask others to wait, but you give your opinions. Those who have talked about DX and OGL being abandoned are the naysayers using that as a fear tactic against Mantle. The Devs surely haven't said they are going to abandon DX and AMD said they had been in communication with M$ all along before releasing Mantle. They had no issues. That doesn't sound like even AMD is presenting it as abandoning other API's.
I'm not the one throwing around performance numbers. I'm not saying 45% or 10X. Nor am I saying 5%-10%, 45% in one specific location on one map, 2 to 3 fps like a new driver release, etc... I just want to talk about what's been presented.
Look at the excitement in this thread http://forums.anandtech.com/showthread.php?t=2362141 This is just all marketing and slides as well. It's not like hype is beneath nVidia. Until we have some concrete info to debunk what's been presented we have hope that it will come out as they say it will. We all realize we need to tune our BS meters with marketing presentations, but it doesn't mean we shouldn't look forward to improvement and increased performance.
Mantle is AMD PR...I'm looking at the history of the company make PR claims.
I know you don't like us bringing those facts to the table...but untill we get more than PR...I will jugde the Mantle PR like any other PR from AMD and conclude:
Their PR never meets reality.
Start a thread about misleading PR and advertising. That's what your post is about. It's not about Mantle.
Show mere where I participate in the hype in that thread (and yes I have posted in that thread)...sorry to burst your bubble, but that argument fails terribad.
Start a thread about misleading PR and advertising. That's what your post is about. It's not about Mantle.
Did I say, Look at Lonbjerg's excitement in that thread? Why should I have to point anything out when I never claimed it?
What do you base that on? While that's surely possible, under the heading that "anything is possible", but why would you assume that's what happened? And if it was, it's also "possible" that it ran 44% faster all of the time and spiked to 45% in one area of one map for a brief moment, by your way of thinking. It's also possible that it hits 45% in many many situations, but just doesn't flat line at +45%. While I don't assume it's best case, it's just as silly to assume worse case as well.
This has to do with Mantle how?
It takes years to build games and game engines. We've had every developer who's worked with it say it's "simple" (that's in relative terms, of course) you are obviously no better informed than any one else around here. That in and of itself should let everyone here know how informed your opinions are. All we can go by is what we've been told and shown.
Start a thread about misleading PR and advertising. That's what your post is about. It's not about Mantle.
Unless you have a download link to anything running via Mantle....this is a PR thread about Mantle...not a "All praise Mantle - no critque allowed - stop anyone not pro AMD/MANTLE from posting"-thread.
So I will keep posting about the terribad PR, since this "easy and simple to implement API" is still MIA and nothing BUT PR still, thank you very much
So I am excluded from "we all"?
FYI, most of this thread is about misleading PR and advertising. So no real need for another thread about the same thing. IMHO.
Between Mantle's powerpoints and AMD stealth marketers perpetuating it...it's just ... ugh. I'll be the first to give props to Mantle when real BF4 benchmarks exist. If it's 45%. Not till then. I'd assume that anyone with any intelligence would feel the same way.
Random side note: I wonder if it's possible to integrate Mantle into the new CG_miner. That would be pretty hilarious.