Upcoming games to support Mantle

[Rusty0611]

What makes you think my post was aimed at you. I was just stating a fact from a year or two ago. I suggest you grow up yourself.

I don't see anybody else making the point that the graphs aren't the be all and end all (in your eyes because they're favouring AMD) in this thread and uksoldierboy was talking to me in the quote you used. Therefore it is a sound assumption to make.

What does what I have said result in a "grow up" remark? What I have said isn't an immature comment. Thank you for furthering my point.

It's funny though as you base your results on a friends machine when i myself have access to much more powerful machines myself.

Good for you but that has no relevance to any point I have made either historically or in this thread.

I wasn't measuring our performance with an FPS counter - merely making sure his PC was running as it should and as part of that we fired up BF4 side by side with my own.

I'll leave you to it as I have better things to go round in circles with you while you confuse yourself with what you're actually saying .

Completely agree bud, but they play a big part, no worries though, just would have just been nice to see what sort of frametimes your getting.

I'm sure another member will help.

Sorry mate - normally I would be up for a bit of user benching but I just really can't be bothered lol. I'm still burned out from 7950 CF OC vs 780 OC at 1080 and 1440. That killed my enthusiasm.

 

[Uncle Petey]

The VRAM thing and the fact it crashes after 10 seconds unless I alt-tab to the desktop and back. Doesn't matter if you're in a menu or in the game, once the game has loaded it will happen, doesn't matter what server.

Obviously the VRAM thing means my fps drops to near single digits (if not single digits, I seem to recall seeing a 3) and the game stutters and becomes unplayable. I don't think I've recorded any frametimes while it's happening, but I'll try to remember to so someone can make a graph.

Weird. I only got the vram thing (after a round)... Which is still unacceptable by itself of course.

I don't think you need worry about recording the frametimes. The spiking and stuttering pretty much say it all when dat wall is breached.

 

[shankly1985]

I'd suggest you ask Thracks (wouldn't call him a liar to his face though).
Does every Mantle tweak get mentioned in the patch notes then?

As LtMatt would say, given the choice between you and an AMD employee I'll believe the AMD employee in this case regarding AMD's Mantle.

I don't after ask him anything, does he work for dice? No does he develop Mantle? No
So until it's listed as fixed in the patch notes it will remain broken. Simples

 

[GoogalyMoogaly]

I don't after ask him anything, does he work for dice? No does he develop Mantle? No
So until it's listed as fixed in the patch notes it will remain broken. Simples

Oh, you're talking Frostbite issues? I'm talking Mantle issues.
Maybe this should be mentioned in all the Mantle threads that talk about Frostbite based games then? A little disclaimer "Does not work on GCN 1 cards".
Mantle has a small enough audience as it is and now the Frostbite engine is limiting it massively further? And they've got other companies to use this engine?

I'll ask Johan on Twitter, see if he agrees that Frostbite only supports GCN 1.1.

Also "after" = "have to"?

 

[GoogalyMoogaly]

Touchy, learn scouse and you understand. And where did I say frostbite? We talking about mantle go read the bloody release notes baby.

** edited **

Well, Thracks says Mantle has parity between GCN 1.0 and GCN 1.1 and he works for AMD, so unless you actually work on Mantle, I'll trust Thracks here. LtMatt seems to have a lot of communication with him, maybe he can clear up if Thracks is worth believing?

I've also asked Johan and he said Frostbite isn't broken on GCN 1.0.

 

[Gregster]

+1

I remember when the shoe was on the other foot. Frame times and then FCAT were what decided how smooth your gaming experience was. Now it does not not matter and it's i see no difference. The graph's that were so in favour of Nvidia now haunt them.

You do realise that FCAT is a tool made by nVidia to see what happens with frames and AMD were dropping frames (runt frames) and hence the smoothness that needed to be fixed?

 

[adolf hamster]

Civ wants mantle because of the CPU overhead gains.
And a game that's graphically demanding probably wants Mantle less.

Mantle doesn't need to justify its existence in performance gains.

i'm not saying its a bad thing, its just graphical fidelity in civ compared with other games doesn't make it seem like its intensive/demanding. although after watching that video i can kind of see where they're coming from, as there is a definite correlation between how far into a game you are and the framerate [and inter-turn lag]

if it offers any improvement at all i'm down for that. my experiences in theif and bf4 have thus far been pretty damn good, having this in civ would be a godsend [i have a nasty habit of always playing marathon games and man do they get laggy by the 1800's or so]

 

[CAT-THE-FIFTH]

You do realise that FCAT is a tool made by nVidia to see what happens with frames and AMD were dropping frames (runt frames) and hence the smoothness that needed to be fixed?

You might need to be careful about that.

The AMD big issues started primarily with the "Never Settle" drivers IIRC.

If you look at the GTX500 and HD6000 series,the former could have much worse frametimes in a number of games,going by the TR reviews of the time. It was not such a clear cut thing.

Fast forward to the GTX660,and it was not so clear cut even when compared to the competition at the time:

http://techreport.com/review/23527/review-nvidia-geforce-gtx-660-graphics-card/4

 

[Gregster]

You might need to be careful about that.

The AMD big issues started primarily with the "Never Settle" drivers IIRC.

If you look at the GTX500 and HD6000 series,the former could have much worse frametimes in a number of games,going by the TR reviews of the time. It was not such a clear cut thing.

I am just saying what FCAT picked up. It was an nVidia tool and used by nVidia for seeing frame smoothness/stuttering. PCPer got hold of it (I believe) and tested AMD cards and found the runt frames that were not making it to the display and this is what caused all that fuss way back then. Previously we had only fraps and that clearly showed a massive difference.

 

[CAT-THE-FIFTH]

I am just saying what FCAT picked up. It was an nVidia tool and used by nVidia for seeing frame smoothness/stuttering. PCPer got hold of it (I believe) and tested AMD cards and found the runt frames that were not making it to the display and this is what caused all that fuss way back then. Previously we had only fraps and that clearly showed a massive difference.

Well I am talking about single card setups,not Crossfire which at least in the past was a bit hit and miss. However more recent PCPER reviews of HD7970 and R9 290 series Crossfire have been more positive AFAIK.

TR did side by side testing of FCAT and their FRAPs results seemed to follow similar trends.

The whole frametimes measurement and stuttering determination was actually down to TR pushing it first(actually HardOCP was doing qualitative measurements for years before),and if you look at a lot of their tests before the Never Settle drivers,there was not much in it for single cards. AMD in their quest to get more performance out of the drivers,broke something - IIRC it was some memory management issue,although they did fix it eventually.

The FCAT story is a bit more complex.

PCPER already debuted their own version of frame-pacing measurements themselves before FCAT debuted for a few months. Some more info was shed by the HardOCP editor,who said Nvidia had been reaching out to sites regarding development of the methodology for a while. This was strongly hinted at by the senior people at TR and Anandtech too. So their first Geforce Titan reviews probably were a form of FCAT but an early version.

TBH,I was a bit disappointed in PCPER not being more transparent about their dealings at the time.

 

[Gregster]

Just checked and it was TheTechReport that did the FCAT article and not PCPer. AMD certainly have improved frame smoothness though, so credit due to them for that.

The bit that is being bandied about is nVidia are having issues and I can honestly say from my own gaming, this isn't the case. I will do some recording tomorrow but as youtube doesn't show what is shown on my screen, that could well get lost in youtubes 30fps/compression.

 

[Martini1991]

Techreport was the first site I saw to make a deal of the AMD stuff. I've always heard from LT that it was someone else?

 

[humbug]

There are what 4 Mantle games? 3 of which use the same engine and thus Mantle implementation?
So saying Thief is the best implementation so far isn't really saying that much is it?

Your right, there could be even better implementations to come yet, no doubt they will with the 20+ games coming that we know of

 

[Final8y]

But we dont see in Frames? Do you have Camera eyes? you a robot? Its not about seeing its about feel.
Also added effect of less input lagg

So what does having 120hz over 60hz give if you can just have the frames 180fps on a 60hz?

Less input lag
Less motion blur
Faster response time
Less screen tear

Beside the smoother and less frame rate dips when the going got tough, i didn't expect less input lag as i use Vsync and i had to check that Vsync was on when i was running Mantle.

 

[LtMatt]

Beside the smoother and less frame rate dips when the going got tough, i didn't expect less input lag as i use Vsync and i had to check that Vsync was on when i was running Mantle.

Yes, I've also noticed this. I used to think the input lag was completely removed, it's not but it is less than it is under DX. If you use an fps cap of 59, it's all but eliminated for me in BF4 using Mantle.

 

[Final8y]

Yes, I've also noticed this. I used to think the input lag was completely removed, it's not but it is less than it is under DX. If you use an fps cap of 59, it's all but eliminated for me in BF4 using Mantle.

I use the fps cap of 59

 

[LtMatt]

I use the fps cap of 59

 

[Rusty0611]

Time to get a 120 Hz monitor Matt so you can go pro.

Some of the newer panels have decent colours. Decent enough that your average user isn't going to notice a massive difference between that and an IPS.

 

[LtMatt]

Time to get a 120 Hz monitor Matt so you can go pro.

Some of the newer panels have decent colours. Decent enough that your average user isn't going to notice a massive difference between that and an IPS.

Nah, my plan is to upgrade the monitor when a 60-120hz freesync 4k screen comes out, hopefully IPS but may have to settle on TN. Not willing to settle for anything less. I'd still rather have a 60hz IPS than a 120hz TN. I don't need a high referesh rate to have a decent ish aim lol.

 

[Orangey]

I think we will have to wait for OLED for 120hz without pleb colours.