-
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.
AMD Mantle Vs. Direct X Battlefield 4 Multiplayer Benchmarks
- Thread starter LtMatt
- Start date
That was the reason i rolled back to 13.12 WHQL. Bugs and lacking crossfire. I think i instructed that person to roll back because of bugs as well.
No im saying that Mantle and a 290 will offer a smoother experience in rendering times than a 780 on DX.
LOL.
Take look at this Ultra footage I recorded with Mantle. Note how smooth the GPU and CPU line is. Now you go and do the same recording on same map with DirectX and you watch your line will be all over the place. With higher Frame times.
http://www.youtube.com/watch?v=4Q4fR36ix04
FreeSync will be a while yet, and as someone here who dont touch sync I have zero need for it
A fair comparison will need to be someone using an i7 4 core 8 threaded cpu. A 6 core 12 threaded cpu will remove some, but not all, of the DX API bottleneck due to its extra cores/threads.
Last edited:
Soldato
Caporegime
Why not just use DDU?
I did driver fusion, wiped it all clean, and it was still fubar with 13.12 putting back on.
It was fine before 14.1.
It's not the first time I've had AMD Driver problems, once I was stuck on an older driver as no matter what I upgraded to FM13 would cease to work.
Soldato
OK - have you compared? Or got a link?
Not looking at frametimes because they're already low enough (and evidently consistent enough) to not really be noticeably improved upon?
Caporegime
Screen cap. its in Windowed mode as i have to use BandiCam to manually overlay a capture window, no other way, or other app that i know of will record / cap Mantle.
The CPU is a P-II x6 @ stock (3.2Ghz) i'm still experimenting before i overclock, but even at stock it does not hinder performance at all.
The CPU is a P-II x6 @ stock (3.2Ghz) i'm still experimenting before i overclock, but even at stock it does not hinder performance at all.
Last edited:
Soldato
Take look at this Ultra footage I recorded with Mantle. Note how smooth the GPU and CPU line is. Now you go and do the same recording on same map with DirectX and you watch your line will be all over the place. With higher Frame times.
http://www.youtube.com/watch?v=4Q4fR36ix04
FreeSync will be a while yet, and as someone here who dont touch sync I have zero need for it
Again, fella that's AMD DX vs AMD Mantle - I don't know why the AMD DX version is not stellar if it's not frametimes
Soldato
Nothing wrong with other games here. User error dont mean everyone will have same experience
But I can say Benchmark's like Heaven 4.0 and 3Dmark run much worst.
Caporegime
It is a very interesting debate with both sides giving valid reasoning to their points. What you put there is just plain rude!
I'm comparing it to the results from various tech sites that say before Mantle AMD have better frame times and less variance than Nvidia in this title on single gpu, like for like. Couple that with the fact that when i tested a single card, i could notice the difference in the smoothness on moving from DX to Mantle. So i can notice it with lower frame times (cpu+gpu) than you, you should be able to feel a smoothness increase with slightly higher frame times (cpu+gpu) if your 780 was able to move from DX to Mantle.
Caporegime
1280x720 Humbug?
Caporegime
Nothing wrong with other games here. User error dont mean everyone will have same experience
But I can say Benchmark's like Heaven 4.0 and 3Dmark run much worst.
How can it be user error if I'm simply using AMD's installer?
There's no variation of user input (Unless I'm going in custom, and even then, I can't do much/anything to break DirectX on my end)
Soldato
- Joined
- 14 Aug 2009
- Posts
- 3,085
I've never said anything to the contrary
Then what is this?
Proof was given to you, but you chose to ignore it.
Let's take it one step at a time:
1. You have tests in which AMD is better on frame times (the big nvidia is better argument a while back) under DX and it only gets better under Mantle;
2. You have tests in which Mantle, frame rate wise, can go from 0% improvement to double digits numbers, even on high end setups. The empirical data is all there and it clearly shows under which situations Mantle kicks in and under which is not better. 'cause yeah, no one says it's better all the time. Even AMD said this API isn't required for all type of games. A lot are fine with how DX works.
On a ending note, showing a graph that puts 0 distance between the APIs and judging only by that is wrong, just as it's wrong to judge those API by another graph that shows Mantle way ahead. The subject has been discussed so much already, all the info is there. Don't get stuck on the details (and I'm referring here in general, not to one user) , look at the whole picture: (for the users) better performance in CPU bound situations and better frame times, (for developers) easier to work and estimate results compared to DX - at least from what we've heard from the devs that are working with it.
Caporegime
For me, it was broke as opposed buggy
Soldato
Long way of saying no then?
I said exluding frame times unless we're splitting hairs over 3-5ms difference
. Could you tell the difference between frame time x and frame time y with 3-5ms difference? I want your eyes if so .
Caporegime
Screen cap. its in Windowed mode as i have to use BandiCam to manually overlay a capture window, no other way, or other app that i know of will record / cap Mantle.
The CPU is a P-II x6 @ stock (3.2Ghz) i'm still experimenting before i overclock, but even at stock it does not hinder performance at all.
Long way of saying no then?
I said exluding frame times unless we're splitting hairs over 3-5ms difference
You can feel the difference Rusty, so yes. If i can feel the difference with lower frame times then you will with higher frame times. You may not want to believe it, but its true. You just don't realise, but then you've only ever known DX so i guess thats why.
Soldato
Then what is this?
Proof was given to you, but you chose to ignore it.
Let's take it one step at a time:
1. You have tests in which AMD is better on frame times (the big nvidia is better argument a while back) under DX and it only gets better under Mantle;
2. You have tests in which Mantle, frame rate wise, can go from 0% improvement to double digits numbers, even on high end setups. The empirical data is all there and it clearly shows under which situations Mantle kicks in and under which is not better. 'cause yeah, no one says it's better all the time. Even AMD said this API isn't required for all type of games. A lot are fine with how DX works.
1) Frame times are already low on both set up's - what difference is 3-5ms going to make?
2) All the testing I've seen shows 5-10% gain on a proper overclocked CPU set up. Not what I'd call amazing but yeah, something is better than nothing.
I don't know why you keep talking about CPU limited scenarios because I'm not CPU limited nor would the PC have been in the OP. Nor have I denied that even higher gains will be present in CPU limited scenarios. I thought that much was obvious given the huge hoards of data of the gains on lower end AMD processors.
Between 2-5 ms difference? I doubt it. Not when the frametimes are already low.
"I may not want to believe it" - why would I not? You're not doing that putting people into camps again are you? I'm just sceptical is all.
I've compared DX nVidia and AMD Mantle side by side. For me there wasn't much difference in general smoothness but hey ho, I wasn't really expecting anything massive either so...
Last edited:
Soldato
My comment is in jest which seems to be the get out clause flavor of the month at the moment
