• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Games before mantle on dx

setter

setter

setter

Caporegime
Joined
14 Dec 2005
Posts
28,067
Location
armoy, n. ireland
Just a curious question guys, now that mantle has been out for a bit, albeit in only a few games. It's an interesting development that i applaud. Mainly my question is, on the games that you play that utilise mantle, how was your experience under dx before mantle came out. Was dx smooth enough in said games as it's something weve had for years. My use of a mantle setup is quite limited tbh, a brief play of bf4 on a relatives system using a 3570k and a 290 gpu.

All to often i see people who use amd gpu's bigging up mantle and rubbishing dx, but as weve all used it for so long, surely it wasnt that bad? Serious question guys and i look forward to feedback from users and a good clean discussion with no fanboyism.:)
 
I think it's more to do with the fact that everything feels smoother. Games would work perfectly before hand but when you spend hours maybe overclocking your system trying to get the most performance out of it. When something comes along like Mantle that gives in essence, a free performance boost, everyone raves about it and rightly so.

I have a temp, i3 4130 CPU on my 7990 card system. I never believed it would run Battlefield 4 smoothly and in DX it doesn't. With everything low the game is still sluggish and stutters pretty much constantly.

When I switch to mantle the difference is night and day even though mantle hasn't been properly optimised for 7*** cards yet, I really noticed a difference. Yes it still looks naff but at least I can play it and I often hit 100 fps at times which is crazy when you consider it's basically a dual core CPU with HT.
 
setter said:
Just a curious question guys, now that mantle has been out for a bit, albeit in only a few games. It's an interesting development that i applaud. Mainly my question is, on the games that you play that utilise mantle, how was your experience under dx before mantle came out. Was dx smooth enough in said games as it's something weve had for years. My use of a mantle setup is quite limited tbh, a brief play of bf4 on a relatives system using a 3570k and a 290 gpu.

All to often i see people who use amd gpu's bigging up mantle and rubbishing dx, but as weve all used it for so long, surely it wasnt that bad? Serious question guys and i look forward to feedback from users and a good clean discussion with no fanboyism.:)
Click to expand...

I can really only answer this for BF4, because ive yet to play Thief (i will in time but not yet) and starsawrm ive run maybe three times and thats it. Really need a bigger sample of Mantle games before i can say its the be all and end all. So i base all of my opinion so far on BF4's implementation.

I found DX11 in BF4 to be a bit flaky from the start. Its always been a bit spiky. Its nothing to do with the frame variance of either sides gpu's. I just don't quite think the API is capable enough of providing a perfect gaming experience in multi player. In single player where its mostly gpu limited its less of an issue. In multi player however there is too much demand on the cpu, too many calculations, too many draw calls, too much variance in frame times. This becomes even more applicable on multi gpu. On single gpu, with the cpu under less load it is less noticeable. Its still noticeable, but to a much lesser degree. Now i never tried BF4 on a 6 core cpu, i imagine for these guys its less of an issue, certainly they're in for the best experience DX11 can offer as they have more threads and BF4 loves threads. However from what ive heard they don't have it that much better than the rest of us with mere four core cpu's. They can sort of brute force their way to a smoother experience though assuming they're playing at high enough res to overcome some of the DX limitations. People ignorant to Mantle (they've never tried it) will scoff and assume they're already getting the best/perfect experience. Well they ain't and its not until you try near on perfection that you realise im afraid. You have to look past the fps, the minimums, sometimes it just comes down to how smooth and consistent the experience of Mantle is vs DX.

James post above sums it up pretty well. The best way to describe Mantle in BF4 vs DX is...

uNHnpLY.jpg.png
 
Last edited:
My experience with Thief in Mantle vs DX is a head scratcher.

in DX the average is higher and the maximums are much higher but the minimums in Mantle are far superior. So while the number suggest DX should be better the variation was much narrower and as such a far smoother experience.
 
ICDP said:
My experience with Thief in Mantle vs DX is a head scratcher.

in DX the average is higher and the maximums are much higher but the minimums in Mantle are far superior. So while the number suggest DX should be better the variation was much narrower and as such a far smoother experience.
Click to expand...

I'd take that any day over 200fps+ ;)
 
Directx was superb on bf4 for example, it's when I started using mantle did I start to see the weakness of dx11.
How I see it is, it's like using 120hz for the first time before that 60hz felt so good you know any better, but then you start using 120hz and 60hz shows all its weakness.
 
Last edited:
I'm a bit like LtMatt in that I've not played Thief (just run the benchmark), ran StarSwarm a couple of times and so have just used BF4 to really experience Mantle.
The initial thing there that has annoyed me is that I have to lower settings in Mantle to stop me getting near to the VRAM limit on my 3GB 7950CF setup compared to me DirectX settings. Having to lower settings to get a playable experience would not be on my list of "must haves" for a new API.
As for Mantle being smoother than DirectX, I can't say I really noticed this either. I should probably point out that after about 5 minutes of watching a Blu-Ray I stop being aware that I'm watching HD, I'm just watching the story. So it may just be that in the middle of everything I'm not really concentrating on how smooth it is when I nervously spin around trying to see where the gunfire is coming from.
I don't put too much stock in min/max framerates, because they could be 1-offs that you barely notice with the framerate usually bouncing around 10fps +/- the average. I'd be more interested in seeing groupings of fps (i.e. how many times it was 30-35 fps, 35-40 fps... 100-105 fps, etc.). Maybe the tool to analyse the BF4 frametime files shows this, I've not looked into it. Anyway, I have definitely noticed a slight increase in framerate just by randomly looking that the fps counter that the in-game overlay provides.
So based on this I'd say Mantle is nice to have as it's free, but I wouldn't have paid money for it.

Based on the StarSwarm demo, Mantle does make a massive difference. Any game that is based on this StarSwarm demo wouldn't interest me, but the concept of being able to draw all the extra units with higher detail does seem like it could be useful for things like RTS games, which I have a passing interest in. Not sure how useful it would be for RPGs or Action/Adventure/Puzzle type games, but having the ability to do this is encouraging.

Thief is a tricky one at the minute, partly as I've only run the demo which I understand is more demanding than the actual game. That said, Mantle Crossfire doesn't work in Thief from what I've been told. This is apparently down to the developers and not AMD. I thought the selling point was that the developers knew how to do this and could do it better. In Thief they've not done it at all. I find it a little worrying/annoying that this could become the case with new games as they are rushed out the door to start making money as soon as possible. So currently in the Thief benchmark my average (and max) fps are much better in DirectX (Crossfire) than Mantle (no Crossfire), so I wouldn't use Mantle if I did play Thief, I bought two cards and I'm gonna bloody well use them!

So from that the tech demo makes it look good, BF4 makes it look 'meh' and Thief currently makes it look bad (for me). Based on this I can take it or leave it. If AMD suddenly found something about Mantle was killing cards prematurely and decided to pull it until they could solve the problem I'd have no problem going back to DirectX.
If Microsoft pulled DirectX and I had to game using Mantle or OpenGL where available and anything that didn't support either of them wouldn't run, I'd be far more upset!
 
GM your problems with Mantle and BF4 is because 7000 series dont support Xfire and Mantle. AMD have said this countless times already. Single 7000 GPUs are getting better with Mantle after each patch and driver.

So I think its unfair to give a view on something that isn't even fully supported. Now if you would use your 290's for a test and base around that it would be much better. Surely you can ignore the fans for couple rounds BF4 to work up a review?

Driver 14.3 issue - nothing has been said since.
Mantle performance for the AMD Radeon™ HD 7000/HD 8000 Series GPUs and AMD Radeon™ R9 280X and R9 270X GPUs will be optimized for BattleField 4™ in future AMD Catalyst™ releases. These products will see limited gains in BattleField 4™ and AMD is currently investigating optimizations for them.
 
Last edited:
I am using my 7950CF it's true. Although my 290s aren't as loud as I remembered them being.
I realised the 7000 series cards weren't optimised, didn't realise there was no CF support in BF4 for them. Seem to be getting ridiculous framerates for one card and Afterburner would suggest something is using my second card quite heavily while I'm playing BF4 using Mantle.
Also, considering the 270X/280/280X are basically 7000 series cards I felt there were more likely to be more people using these cards than 290/290X/295x2 cards and so it probably is more representative. We could all base our framerates off Kaap's findings with 4 Titans (or 4 290Xs) but not everyone has this setup.

I feel they should either lock 7000 series cards in crossfire from using Mantle (if it's not supported) or be prepared to be judged on it. Mantle's not even out yet but we're giving feedback on it, why should that be any different with cards that aren't fully supported yet?
 
GoogalyMoogaly said:
I am using my 7950CF it's true. Although my 290s aren't as loud as I remembered them being.
I realised the 7000 series cards weren't optimised, didn't realise there was no CF support in BF4 for them. Seem to be getting ridiculous framerates for one card and Afterburner would suggest something is using my second card quite heavily while I'm playing BF4 using Mantle.
Also, considering the 270X/280/280X are basically 7000 series cards I felt there were more likely to be more people using these cards than 290/290X/295x2 cards and so it probably is more representative. We could all base our framerates off Kaap's findings with 4 Titans (or 4 290Xs) but not everyone has this setup.

I feel they should either lock 7000 series cards in crossfire from using Mantle (if it's not supported) or be prepared to be judged on it. Mantle's not even out yet but we're giving feedback on it, why should that be any different with cards that aren't fully supported yet?
Click to expand...

Crossfire will work, in Mantle just like Single 7000 series it just wont run very well because its not been optimized for it.
Like I said in another thread to you, if you going to keep using 7950's you better of using DirectX, if you want the extra benefit from Mantle then use the 290's
 
CF is working in BF4 with Mantle, its just not as optimized as what it could be is what they are saying.

Saying that CF is not working on the 7xxx with Mantle is just flat out wrong, the only issue i have experienced is the Vram..
 
Last edited:
No need for me to add anything to smoothness ecte that's already been said.

I can give you another angle.

I have a weaker CPU than most, although the CPU actually is not at all weak, its simply that DirectX does not make use of the power it does have, where as Mantle does.
So while DirectX bottlenecks the crap out of my R9 290, with Mantle i actually have CPU power in reserve.

So from my perspective DirectX is junk and Mantle is brilliant.

The performance difference for me between DirectX and Mantle is very significant.

What thats enabled me to do (due to cost) is buy an R9 290 instead of having to pull my rig apart, sell most of it and then put it back together with a lesser GPU and a used i5 or whatever.



 
Last edited:
It's just a shame all other graphics cards except amd are locked out, but that's the graphic card industry I suppose.

I agree with shankly it's like 60hz vs 120hz, I would like a 120hz monitor but can't really see what I am missing as my experience on 60hz feels as smooth as it can be, I am sure my opinion would change if I got a 120hz monitor.
 
8150 at 4.3, stock 290x= smooth and playable bf4 with everything ultra bar aa [just 2x post], running 5040*1200

framerate isn't everything, sure solid 60 is nice, but i've played a couple of games that were locked to 30 and they were as smooth as anything.

haven't tried dx, but i doubt i'd see anywhere near those results. i'm happy to see my system sweating, at least i'm not wasting horsepower that way.
 
If you're asking if there was a downgrade in performance after Mantle was released, then the answer is that the opposite happened. I think BF4's DX performance improved a bit with drivers and patches, for me at least.

Nothing like Mantle, though.
 
You must log in or register to reply here.
Back
Top Bottom