The AMD Mantle Thread

[bystander36]

Its one of the wonders of competition. It nearly always works and will absolutely work in this situation. If mantle sucks it will fade away if not it will not. No need to put a lot of feeling into that. Lol.

I think GaiaHunter had it correct. Mantle could be great, but that does not mean it will be financially sound. Mantle shows a lot of promise, and as someone who likes high FPS over high resolutions/settings, it targets an area that interests me, but due to limited hardware support, it may not be financially sound.

 

[krumme]

...somewhere along the lines money changed hands somewhere.

Spooky. Was it in the darkness where the guards never come?

Money is only a symbol. What is happening is a mutual business interest. Its a win win and there even dont have to be any exchange of money for that matter. With the console win this is not the biggest surprice ever. But the agressiveness and speed is. It nothing we are used to from amd.

 

[JimmyH]

I don't think a Sandy or beyond i5 bottlenecks a 7950 at 1080p or above...

A common setup for people that actually play BF4 like myself is 144hz monitor at 1080p low preset and I am certainly 100% bottlenecked by the I5. Even w/ lowest settings + 2x cheapest AA still bottlekneck on I5. Entire clan plays on 120hz or 144hz monitors at 1080 low preset all the I5 (sandy, ivy, haswell) members are bottlenecked regardless of how crazy the overclock is. 1080p low preset is a very common FPS setup for those that want to see campers and play at a decent level.

 

[cmdrdredd]

And so for that you'd burn AMD at the stake for a free market working how a free market works in response to a pseudo currency demand completely outside of their control?

I don't care about AMD. I don't know why you think people have to feel sorry for them. Mantle helps more the higher end card you have. Paying $700 for a 290x for better performance in one game doesn't interest me in the least.

 

[smackababy]

I don't care about AMD. I don't know why you think people have to feel sorry for them. Mantle helps more the higher end card you have. Paying $700 for a 290x for better performance in one game doesn't interest me in the least.

You have to feel sorry for AMD. It is required. I mean, after all these years of operating in the red and they still don't qualify as a nonprofit organization.

Thread crapping helps no one.

-Elfear

 

[krumme]

I think GaiaHunter had it correct. Mantle could be great, but that does not mean it will be financially sound. Mantle shows a lot of promise, and as someone who likes high FPS over high resolutions/settings, it targets an area that interests me, but due to limited hardware support, it may not be financially sound.

Mantle cant be great if its not financially sound. Makes no sense. We build our entire economies of those economic principles. That is except North Korea and the likes.

And no. Mantle is suddenly not the only exception to that rule.

 

[ASM-coder]

OT, since apparently the topic at hand is whether Mantle makes business sense...

In regards to Test Case 3 put forth by Johan. I read this pseudo explanation
from some guy(on a different board) about the impact of Mantle in this example, and he argues...

"No, the Crossfire gains aren’t because of CPU bottlenecking. There’s an entirely different explanation for that, related to how DX shares information across multiple GPUs."

Anyone know more detail of what this might be and is SLI affected the same way?

 

[Skurge]

I don't care about AMD. I don't know why you think people have to feel sorry for them. Mantle helps more the higher end card you have. Paying $700 for a 290x for better performance in one game doesn't interest me in the least.

It clearly interests enough people to keep prices to remain where they are. For the states anyway. As a poster above mentioned. When I play bf4 I play it at lower settings on my 120hz monitor. So mantle will help me immensely. You can parading how it for a small percentage, but that's better than nothing and well, when a gtx780ti and a 290x cost the same, but the 290x is a lot faster in the game I play by far the most, why wouldn't I get the 290x. You can't apply how you would buy something and then make a blanket statement that the 290x isn't worth the money. Also. Its only that expensive in the states, everywhere else its much much cheaper. As much as some would like to believe, the world doesn't revolve around the states.

 

[krumme]

You have to feel sorry for AMD. It is required. I mean, after all these years of operating in the red and they still don't qualify as a nonprofit organization.

Dont cry. Its difficult in one post to say amd pays for it all and in another realizing they have been in red basicly forewer. I know its difficult.

 

[badb0y]

Congrats father. Enjoy your low min fps then, when i frag on with consistently good performance. S m ooth nes s in dx.

Ok?

You run a Gigabyte GTX 780 (Stock) w/ XSPC Full Cover Water block.

Would anything have impressed you?
I don't think you are part of the target audience.

I think I was the target audience and I would have been impressed if the performance boost was higher than 2.4%, that's like within the margin of error. My last 4 cards have been GTX 460, HD 6950, HD 7970, and a GTX 780. I am far more likely to jump ship than other people if I see value and honestly I don't see value in Mantle so far.

 

[krumme]

OT, since apparently the topic at hand is whether Mantle makes business sense...

In regards to Test Case 3 put forth by Johan. I read this pseudo explanation
from some guy(on a different board) about the impact of Mantle in this example, and he argues...

"No, the Crossfire gains aren’t because of CPU bottlenecking. There’s an entirely different explanation for that, related to how DX shares information across multiple GPUs."

Anyone know more detail of what this might be and is SLI affected the same way?

If i remember and understand corectly the answer is in the transparency of the execution model. With mantle and more gpu you basickly gets several layers besides/on top of each other and the numbers of gpu is in theory perfectly scalable. And you can adress them like you want. Its explained in of the more technical presentations from amd at apu13.

 

[tonyfreak215]

AMD is not Apple...nice try. Who is really gonna pay $700 for a 290x that only has an advantage in one game? Only miners seem to be willing to pay the costs.

I understand they aren't Apple. It was an example of how the markets can change. Back in the late 90's they were a minor player, look at them now. AMD will not always be "5% market share". (plus they currently have more than 5%)

Are they? Interesting you use Apple as an example to help show why having increased marketshare in the PC market equates to more games being available on your platform.

I wasn't using Apple as an example of that. Was for market flexibility.

 

[VulgarDisplay]

I look at mantle like a new console generation. They will keep upping the performance in games using it as time passes.

I'm very interested in love quality settings at 1080p benchmarks because that is what applies to me. I think mantle is going to be a huge deal for smoother rendering and minimum fps. The two things that actually matter far more than max fps.

I know bf4 feels like hot garbage below 90 fps @120hz. If mantle keeps my minimums up over that threshold then I'm a believer.

 

[ginfest]

Just read the Extreme Tech article and wow! We have 239 pages of non-facts and marketing drivel here, only to find out the truth that Mantle isn't what we were promised. Hopefully this thread can now be put to rest.

 

[el etro]

So send the driver first to sites so they can test in singleplayer and with only overclocked i7 processors to show pitful gains. And end users that play multiplayer and uses cheaper processors like i5 or FX that have chance to see decent improvments have to wait another day.

They can test with other configurations too. 24 hours will be not sufficient to test Mantle capabilities.

Anand, by example, is a AMD partner, so AMD must ask them to test diferent combinations of CPU+GPU to test Mantle powers.

 

[bystander36]

Mantle cant be great if its not financially sound. Makes no sense. We build our entire economies of those economic principles. That is except North Korea and the likes.

And no. Mantle is suddenly not the only exception to that rule.

Being great does not mean it is financially sound. There are many great technologies out there. There are a few display techniques that are far superior to what we use, but they are too costly to use, so they are not used.

The product may be great, but it also has to be financially sound to be used. Being a good product is only on part of the equation.

 

[Cloudfire777]

If developers relocate resources from the DX version to the Mantle version, the DX version suffers under almost all circumstances, otherwise, they'd have just let those resources go. They already don't allocate OpenGL resources in 99% of game development because the additional market isn't large enough to support the investment costs. What makes Mantle any different? The entire market encompassed by Mantle is included in the DX market.

If, at some point in the future, AMD has Mantle support on a vast majority of their marketshare, it might become a strategy to develop against Mantle and DX. And even then, that is a dubious claim.

I understand developers wanting it. I get it. I am a developer. I love new technologies. I wish I could use all kinds of new stuff. However, it makes little financial sense from a business aspect.

Bingo.

2 years from now if Mantle survives it will be AMD`s PhysX. Support the few who bothers to implement it in the games, the rest of the developers moves on and continue to use the mainstream methods (OpenGL or DX).
There is one good thing with Mantle and this is that it hopefully will get Microsoft`s behind in gear and they will finally update the 11.1 (11.2?) version to a newer one with similar improvements. When that happens, it will be the final nail in the coffin. Whats even the point with another API that gives the same benefit then?

Mantle clearly works by doing some magic with the CPU mostly. Which means it shouldnt be GPU architecture specific. Which means a similar system (or update) will easily offer the same across all systems.

 

[Midwayman]

Just read the Extreme Tech article and wow! We have 239 pages of non-facts and marketing drivel here, only to find out the truth that Mantle isn't what we were promised. Hopefully this thread can now be put to rest.

Right. An article written based on marketing slides without any testing of the actual product is conclusive. We won't have any real useful info for a couple of days.

 

[bystander36]

Just read the Extreme Tech article and wow! We have 239 pages of non-facts and marketing drivel here, only to find out the truth that Mantle isn't what we were promised. Hopefully this thread can now be put to rest.

It may not be what some posters thought it would be, but up to 45% is accurate, it just isn't the norm. That is why it said "up to".

However, those who were saying a single 7970 would beat Titan SLI were quite wrong.

 

[OCGuy]

A common setup for people that actually play BF4 like myself is 144hz monitor at 1080p low preset and I am certainly 100% bottlenecked by the I5. Even w/ lowest settings + 2x cheapest AA still bottlekneck on I5. Entire clan plays on 120hz or 144hz monitors at 1080 low preset all the I5 (sandy, ivy, haswell) members are bottlenecked regardless of how crazy the overclock is. 1080p low preset is a very common FPS setup for those that want to see campers and play at a decent level.

This isn't remotely true. I just ran a 64/64 player map on all low settings @ 1080p (your refresh rate is irrelevant) with the 4670 @ a boring 4ghz and a 7970. Not only was my chip never pegged at 100%, but it ended at a 101FPS average..so even if someday it becomes a chokepoint at an artificially strangling setting, I would never notice, since it is 100FPS.

 

[Venomous]

AMD bit off more then they can chew by screwing up this mantle and driver debacle. They look like real idiots right now...

Besides all of the mantle hype leading up today , turns out they are giving press the driver before actual customers. They lied about their drivers being ready and pointed the fingers at dice for the December delay.

AMD is just straight out lying at this point about mantle, where it's actually at and gcn hardware not being created equal.

the driver not being available at all during the month of January shows it was never ready to begin with. GCN 1.0 lost performance, yet with a supposed mantle driver ready, shouldn't this have already been caught?

I think AMD was better off working with dice silently and setting a goal to release mantle this summer where every supported card works out the gate.


Bottom line is, as a customer, I don't appreciate being jerked around nor lied to.

 

[blackened23]

It is interesting how the narrative from Mantle went from outrageous performance gains from 1000000 draw calls to now "think of what you can do with a low end CPU". Interesting how that suddenly changed. Low end CPU's is definitely what I had in mind, and everyone had in mind when AMD revealed Mantle at GPU14 for their 290/290X cards. Yeah uh, I don't think so.

And now, yet again, AMD missed another promise to their customers with their supposed December then January release date for Mantle. I mean, where do users draw the line? I've been there as a former AMD customer with promises about HD3D, frame pacing, etc etc etc and at some point you have to say you're fed up with not delivering on promises.

I really think AMD needs a shift in terms of trying to pre-hype and pre-promise on things they end up not delivering on. IF they just SHUT UP and quietly do work nobody will get mad. If they poop on customers after promising something and failing to deliver, people will take note of that. I've been there. That's why I don't prefer AMD these days (and I used to like them quite a bit), after a while stuff like this, failing to deliver on what you say you'll do, it gets old. I won't even name the stuff that happened with Tahiti in terms of failed promises, it's just a very extensive list. And not relevant to Mantle.

I guess my main point here is the pre-hype train that AMD relies on so heavily. That needs to stop IMO. Hyping up something to death which ends up not delivering on expectations, and then hyping a driver or feature that doesn't release on time, come on. If AMD had just mentioned Mantle without the 45% and without the two concrete missed time frames, nobody would care and there would be no controversy. If AMD just shuts up and quietly does their thing without pre-hype, that's the way they should do things. When something is READY and AWESOME, then hype it up. But they have to hype it up beyond belief with stupid marketing before it's ready and here we are. 0-5% gain in GPU limited situations and a delayed driver. And i'm sure AMD will get defended non stop about this, but from a consumer standpoint I find it surprising that this stuff continually happens and some users just don't care. Personally, yeah, I prefer nvidia because after years of getting pooped on by AMD with their missed time frames and missed promises,...... and AMD seems to continue that trend with Mantle....it doesn't do their brand any favors. They just need to stop the pre-hype, excessive social media marketing and exaggerated marketing powerpoint nonsense. Especially if they don't meet said expectations or release dates after hyping it to death. Christ.

 

[Venomous]

It is interesting how the narrative from Mantle went from outrageous performance gains from 1000000 draw calls to now "think of what you can do with a low end CPU". Interesting how that suddenly changed. Low end CPU's is definitely what I had in mind, and everyone had in mind when AMD revealed Mantle at GPU14 for their 290/290X cards. Yeah uh, I don't think so.

And now, yet again, AMD missed another promise to their customers with their supposed December then January release date for Mantle. I mean, where do users draw the line? I've been there as a former AMD customer with promises about HD3D, frame pacing, etc etc etc and at some point you have to say you're fed up with not delivering on promises.

I really think AMD needs a shift here in terms of trying to pre-hype and pre-promise on things they end up not delivering on. IF they just SHUT UP and quietly do work nobody will get mad. If they poop on customers after promising something and failing to deliver, people will take note of that. I've been there. That's why I don't prefer AMD these days (and I used to like them quite a bit), after a while stuff like this, failing to deliver on what you say you'll do, it gets old. I won't even name the stuff that happened with Tahiti in terms of failed promises, it's just a very extensive list. And not relevant to Mantle.

I guess my main point here is the pre-hype train that AMD relies on so heavily. That needs to stop IMO. Hyping up something to death which ends up not delivering on expectations, and then hyping a driver or feature that doesn't release on time, come on. If AMD had just mentioned Mantle without the 45% and without the two concrete missed time frames, nobody would care and there would be no controversy. If AMD just shuts up and quietly does their thing without pre-hype, that's the way they should do things. When something is READY and AWESOME, then hype it up. But they have to hype it up beyond belief with stupid marketing and here we are. 0-5% gain in GPU limited situations and a delayed driver. And i'm sure AMD will get defended non stop about this, but from a consumer standpoint I find it surprising that this stuff continually happens and some users just don't care.

Couldn't agree more. I will no longer have to keep waiting for an affordable 290 to be available to replace these pair of 280x's I have. I'm going to get a couple 780s in SLI and dump these AMD cards on ebay for profit. I'm done with AMDs marketing bs. Even when these 14.1 driver drop, is gcn 1.0 will have lost performances, not gained anything.

 

[Cloudfire777]

Question:

What happens say in next year when we get more powerful CPUs from Intel and AMD? Will the Mantle improvement shrink?

 

[bystander36]

It is interesting how the narrative from Mantle went from outrageous performance gains from 1000000 draw calls to now "think of what you can do with a low end CPU". Interesting how that suddenly changed.

Low end CPU's is definitely what I had in mind, and everyone had in mind when AMD revealed Mantle at GPU14 for their 290/290X cards. Yeah uh, I don't think so.

I'm not a fan of multiple API's, so am not really a big Mantle supporter, but I do like what it can do to those who like higher FPS. My biggest hope is that Microsoft will work hard to limit the CPU overhead, and make multithreading easier. This helps even those with high end CPU's in many games.