Mantle Feedback/Bugs

g3po · 3 Feb 2014 at 22:16

Haven't had any crashes and when I first installed I noticed fps gains in bf4, but started gaming today and im only getting 50fps on average with dips of 35. Whereas before i had it capped at 60fps and it rarely went below that, any ideas what im doing wrong ???

MOOGLEYS · 3 Feb 2014 at 22:16

God damit now getting black screens in windows.

Had no issues at all before installing the mantle drivers. Hopefully it's just a driver issues and will be fixed with a re-load.

GoogalyMoogaly · 3 Feb 2014 at 22:20

I think the trouble with BF4 is that most people will spend most time in MP, so benching SP is only a little better than basing gaming scores off a Heaven run. It's representative of something, but not really what 'matters'.

Hopefully Thief will have a built in benchmark to stop the arguing about what is a valid benchmark (do we really need more to argue about!?), although I'm not aware if it has a multiplayer.

Sacrifice · 3 Feb 2014 at 22:22

Guys, I have tried screenshotting in dx and mantle, but for some reason won't work, so my perfgraph looks like a constant green thick line, ranging about 10 up and down. Any ideas? This is running crossfire only btw.

BuzzLloydyear · 3 Feb 2014 at 22:24

Whilst both GPU's on my 7990 are working in Mantle, I am not convinced GPU2 is doing anything overly constructive. I tried a run of Matt's planned benchmark in "borderless" mode to ensure I am only using 1 gpu and my frames per second went up a bit from when I was in full screen with both gpu's working.

With Crossfire, I got Min FPS: 11.09, Max FPS: 78.62, Avg FPS: 58.34

Without Crossfire, I got Min FPS: 19.61, Max FPS: 109.29, Avg FPS: 61.83

I will say, full screen one seemed to feel smoother. Numbers wise, boderless took the crown.

I am not quite sure what I am seeing, it's pretty hard to get my head around. I will have to do more testing I suppose. I know crossfire is dodgy at the moment, but surely if both cards are running quite heavily (95-99%), then I should be seeing improvements. No wonder my DX11 results are faster than my Mantle ones at the moment.

Azic · 3 Feb 2014 at 22:30

Well I have just played in mantle and its pretty sweet I got to say.
My average FPS was a solid 80FPS @ 1080p all ULTRA with 2X MSAA, i72600k @ 4.6GHZ 8GB RAM ASUS BF4 290X

Its now gone up to an average of 133 same settings! Cant ****ing beleive it.
I have not noticed any fogging issues at all. The colour looks much deeper and richer.
The only thing I noticed is an odd stutter when a squad mate spawns on me.

I have an ASUS DG5 sound card the £30.00 one or it was when I bought it, and TBH i think that is a bit **** really and causes the issue as it seems sound related.

Does BF4 use True Audio?

All in all I am super pleased with this and to think I was ready to ditch the card for a 780TI! glad I did not now if this is the shape of things to come.

EDIT****

Tested on the same 24/7 Shanghai Server for 2 25min runs mixture of infantry & tank as a passenger. once in heli on DX 11 run

Rusty0611 · 3 Feb 2014 at 22:32

LtMatt said:
Its not about being rude, its just common sense. I could run a bench of Shanghai though and likely score higher than both of you with my gpu at stock clocks, but as ive no idea what their benching procedure was, its also worthless. Multiplayer is inconsistent so either result is worthless as we don't know the procedure or playstyle of the people who did the bench. They might have spent the round in a tank. Pgi or yourself could have spent the round on foot. That alone will account for a massive difference in fps.

Re-read what I've said about a 15 min representative sample.

LtMatt · 3 Feb 2014 at 22:41

Rusty0611 said:
Re-read what I've said about a 15 min representative sample.

Re-read what i said. Depends how you play the benchmark the scene/play the game.

Gregster · 3 Feb 2014 at 22:42

LtMatt said:
Your data is almost worthless though because of the inconsistencies of multiplayer benchmarking. Unless the same person is running the bench following the same practices, its too inconsistent.

Matt, don't be so rude fella. PGI is one of the good guys who is unbiased and constantly switches between AMD and nVidia. It is good to compare Mantle to DX and you can attain the averages by playing the same map.

setter · 3 Feb 2014 at 22:45

I find that if your going to bench bf4 in mp, it's best to try and use a good mix of vehicle and on foot stuff in your run. Ie no camping on a rooftop/edge of a map, best to get stuck in on the parts were all the action is.

LtMatt · 3 Feb 2014 at 22:46

Gregster said:
Matt, don't be so rude fella. PGI is one of the good guys who is unbiased and constantly switches between AMD and nVidia. It is good to compare Mantle to DX and you can attain the averages by playing the same map.

Its not about being rude as i said to Rusty. Its simply a case of not comparing like for like testing runs. We all know that results can vary greatly in multiplayer depending on a lot of different factors. This is pretty basic stuff and was not meaning to be rude to pgi at all. I just disagreed with his assumption that their testing must be off because he scored similar fps when we have no idea how they tested that benchmark. If one of you knows exactly what they did then i stand corrected and apologise. I think we all know thats not the case though.

Rusty0611 · 3 Feb 2014 at 22:48

LtMatt said:
Re-read what i said. Depends how you play the benchmark the scene/play the game.

Representative.

LtMatt · 3 Feb 2014 at 22:49

Rusty0611 said:
Representative.

That's not how it works on multi player when you have two different people on two different machines benchmarking in different ways Rusty. Single player sure, multi player is not as consistent. Its pretty well known benchmarking practice. Its why so few of the review sites bench multi player.

setter said:
I find that if your going to bench bf4 in mp, it's best to try and use a good mix of vehicle and on foot stuff in your run. Ie no camping on a rooftop/edge of a map, best to get stuck in on the parts were all the action is.

Exactly. It depends how the game was played and scene was benched. Going round in a chopper or sniping on a roof = highest fps. Spending more time in a tank, lowest fps. Running round on foot, somewhere in the middle but generally higher fps. That's why its not reliable comparing multi player benches. Its only representative if both players do the same thing. Also depends on how long each bench is run for.

Gregster · 3 Feb 2014 at 22:52

LtMatt said:
Its not about being rude as i said to Rusty. Its simply a case of not comparing like for like testing runs. We all know that results can vary greatly in multiplayer depending on a lot of different factors. This is pretty basic stuff and was not meaning to be rude to pgi at all. I just disagreed with his assumption that their testing must be off because he scored similar fps when we have no idea how they tested that benchmark. If one of you knows exactly what they did then i stand corrected and apologise. I think we all know thats not the case though.

It came across as rude and all you would have to do is join PGI in the same game/squad and stick together. You will both end up with the same average, so it doesn't make his benching worthless. Like setter said, jump in tanks and get in the action and they will be pretty accurate.

pgi947 · 3 Feb 2014 at 22:55

The point I was making was about cpu usage, if you'd care to read back your own comments, at the bottom of my post I said I would get around to a stock run, my whole point was based around the benches you posted showing an absolute terrible average on both dx and mantle with a 290x on a stock 4770k.

Anyhow, can do without trying to put the effort in to get something blindly dismissed.

**** this I'm out, totally, these forums have become a hell hole over the last few months.

Later fellas.

Rusty0611 · 3 Feb 2014 at 22:55

LtMatt said:
That's not how it works on multi player when you have two different people on two different machines benchmarking in different ways Rusty. Single player sure, multi player is not as consistent. Its pretty well known benchmarking practice. Its why so few of the review sites bench multi player.

It is how it works when it's... representative. If somebody goes out of their way to be representative over a 15 minute sample then that is long enough to smooth it out.

LtMatt · 3 Feb 2014 at 22:56

Gregster said:
It came across as rude and all you would have to do is join PGI in the same game/squad and stick together. You will both end up with the same average, so it doesn't make his benching worthless. Like setter said, jump in tanks and get in the action and they will be pretty accurate.

I apologise to pgi if he took offence. Merely explaining why comparing his benchmark result of multiplayer to another multiplayer benchmark, in which he had no idea what their benchmarking procedure was, was not a very accurate comparison. Sorry pgi if that was rude.

shankly1985 · 3 Feb 2014 at 22:59

Gregster said:
Is it ok in DX then? I have read lots of people have gone back to 13.12 or earlier to get BF4 running well.

No problems here runs fine, only I must lower from Ultra to mix ultra/high to keep 120fps. But yeah runs very good.

Rusty0611 · 3 Feb 2014 at 23:01

LtMatt said:
I apologise to pgi if he took offence. Merely explaining why comparing his benchmark result of multiplayer to another multiplayer benchmark, in which he had no idea what their benchmarking procedure was, was not a very accurate comparison. Sorry pgi if that was rude.

Which is the whole point that if it is representative and long enough then it doesn't matter that it isn't identical.

shankly1985 · 3 Feb 2014 at 23:02

MOOGLEYS said:
God damit now getting black screens in windows. Had no issues at all before installing the mantle drivers. Hopefully it's just a driver issues and will be fixed with a re-load.

When I had black screen in the past it was my psu.