Gameworks, Mantle and a pot calling a kettle black

Martini1991 · 12 Jun 2014 at 10:19

And developers lie, like every time they're bringing out a game

"IT'S GOING TO BLOW YOUR MIND"

LtMatt · 12 Jun 2014 at 10:19

andybird123 said:
SNIP

I was responding to your deferred contexts claims.

Martini1991 said:
You may well have responded to it, but it was nothing relevant or meaningful.

Actually he responded to me first after i asked a question about them.

Rusty0611 said:
It's not as simple as that. They have knowledge - don't think anybody has disputed they know better - but they're also (currently) intrinsically linked to AMD. If you're taking what they say at face value then you are being gullible. Not to say everything they say has been spun or anything like that but you can't just follow what they say and regurgitate it as absolute like a sheep.

Well we'll see won't we. No doubt the next civ game will have DX11 DC and we can compare it to Mantle and see if they were telling the truth or not. I know what my money is on.

Rusty0611 · 12 Jun 2014 at 10:20

There is that but Matt doesn't like it if you call his mates liars

.

DICE: "We will never charge for DLC"

Awkward.

LtMatt · 12 Jun 2014 at 10:22

Rusty0611 said:
There is that but Matt doesn't like it if you call his mates liars .

DICE: "We will never charge for DLC"

Awkward.

What relevance does that have to this?

Its just funny when you see resident experts in action. Can't help but chuckle.

Rusty0611 · 12 Jun 2014 at 10:23

LtMatt said:
Well we'll see won't we. No doubt the next civ game will have DX11 DC and we can compare it to Mantle and see if they were telling the truth or not. I know what my money is on.

What is it with you lot taking examples of a more general point as the point itself

.

To be honest, I couldn't care less about the specific example of Civ 5 - I don't think Andy does either. He was using that as an example of a more general point and to my eyes the example fits the point he was making. Focusing on the example itself is bad debating skills

.

LtMatt said:
What relevance does that have to this?
Its just funny when you see resident experts in action. Can't help but chuckle.

Martini1991 said:
And developers lie, like every time they're bringing out a game "IT'S GOING TO BLOW YOUR MIND"

Martini1991 · 12 Jun 2014 at 10:23

LtMatt said:
Well we'll see won't we. No doubt the next civ game will have DX11 DC and we can compare it to Mantle and see if they were telling the truth or not. I know what my money is on.

Unless I'm being very silly, is anyone actually refuting that Mantle will give the greatest performance in Civ? Because I'd be pegging on my AMD GPU set up over 780, but that doesn't make the developers inherently correct.

LtMatt · 12 Jun 2014 at 10:26

Rusty0611 said:
What is it with you lot taking examples of a more general point as the point itself .

To be honest, I couldn't care less about the specific example of Civ 5 - I don't think Andy does either. He was using that as an example of a more general point and to my eyes the example fits the point he was making. Focusing on the example itself is bad debating skills .

I asked him a question about DC as ive heard a lot about them and responded to his reply to me. Not interested in the other point he was making which is why i didn't respond to it. It's not the first time ive seen andy harping on about deferred contexts and he'll know what im talking about.

Martini1991 said:
Unless I'm being very silly, is anyone actually refuting that Mantle will give the greatest performance in Civ? Because I'd be pegging on my AMD GPU set up over 780, but that doesn't make the developers inherently correct.

Deferred contexts improve performance for the first five minutes of gameplay, then the accumulated contexts eventually slow down the framerates below an engine without them. But how long do benchmarks take?

Shrewd, no? How long is StarSwarm? 5 minutes.

Martini1991 · 12 Jun 2014 at 10:29

Well, bird said he hasn't seen a tail off in performance, so either he's a liar, or it's not the case.

And that again, doesn't actually directly link to your comment.

triss · 12 Jun 2014 at 10:34

LtMatt said:
I was responding to your deferred contexts claims.

Actually he responded to me first after i asked a question about them.

Well we'll see won't we. No doubt the next civ game will have DX11 DC and we can compare it to Mantle and see if they were telling the truth or not. I know what my money is on.

How exactly is the Next Civ game having mantle relevant to If command lists work for 5 mins or not?

And how is it relevant to what happened on a old game that didnt have mantle?

And lets say the next gen Civ has mantle and for DX has command lists how does that prove if they were telling the truth back then since they wernt comparing or even mentioning mantle?

Mantle without a doubt is better for most AMD users than DX so far But its not a tool to prove what was said about something in the past

Gregster · 12 Jun 2014 at 10:52

Getting back on topic and not calling people names, if GameWorks is so bad, how come it took a 780Ti at over 1400Mhz to beat a 290X at 1250Mhz in Batman Arkham Origins Matt? I don't care what happens when the settings are turned down, you have close relations with Joel and he never replied to that question either when I asked him. If GameWorks was gimping performance on AMD or even not allowing AMD to optimise, how is GameWorks so bad for the industry?

I can see why the nVidia guy was annoyed at all the volleys from extremetech and AMD.

LtMatt · 12 Jun 2014 at 10:54

Gregster said:
Getting back on topic and not calling people names, if GameWorks is so bad, how come it took a 780Ti at over 1400Mhz to beat a 290X at 1250Mhz in Batman Arkham Origins Matt? I don't care what happens when the settings are turned down, you have close relations with Joel and he never replied to that question either when I asked him. If GameWorks was gimping performance on AMD or even not allowing AMD to optimise, how is GameWorks so bad for the industry?

I'm not getting drawn back into the GameWorks debate again Greg, I've said my piece on that for now. I can put you in touch with Joel via trust if you want though, you can ask him yourself then.

CAT-THE-FIFTH · 12 Jun 2014 at 11:12

andybird123 said:
BF3 also used Command Lists and that game also showed a 10-20% vendor bias.

BF3 also used NVAPI.

Details of NVAPI:

https://developer.nvidia.com/nvapi

Frame Rendering
Ability to control Video and DX rendering not available in DX runtime. Requires NDA Edition for full control of this feature.

Gregster · 12 Jun 2014 at 11:23

LtMatt said:
I'm not getting drawn back into the GameWorks debate again Greg, I've said my piece on that for now. I can put you in touch with Joel via trust if you want though, you can ask him yourself then.

Lol this thread is about GameWorks and Mantle and it is a simple question surely? If GameWorks locks AMD out of optimizations, how come the 290X is faster than the 780Ti in Batman AO when all the settings are turned on?

From what you and others keep spouting, surely nVidia would stomp on the 290X with ease.

CAT-THE-FIFTH · 12 Jun 2014 at 11:34

Gregster said:
Lol this thread is about GameWorks and Mantle and it is a simple question surely? If GameWorks locks AMD out of optimizations, how come the 290X is faster than the 780Ti in Batman AO when all the settings are turned on?

From what you and others keep spouting, surely nVidia would stomp on the 290X with ease.

Is this stock to stock or overclocked??

andybird123 · 12 Jun 2014 at 11:36

both, stock for stock and the 780ti needs a much much larger OC to equal / best the 290X at max settings

CAT-THE-FIFTH · 12 Jun 2014 at 11:44

andybird123 said:
both, stock for stock and the 780ti needs a much much larger OC to equal / best the 290X at max settings

Do the GTX780TI cards boost much over 1GHZ?? Not been following them or the R9s that much after launch TBH,and do you mean this thread:

http://forums.overclockers.co.uk/showthread.php?t=18585432

That is at least 4 months after the game launch - I thought most of the complaints were at launch or within two months of launch??

Could be wrong OFC.

andybird123 · 12 Jun 2014 at 11:51

Yeah, most Titan's top out at bout 1200mhz without voltage mods (e.g. custom bios) and 780's not much better, but 780ti's will hit 1300+ pretty easily
VSG's score in that thread was at 1463mhz

where as Kaap's 290X score is at 1250mhz

hence why Gregster is saying, what is the problem with Arkham City / GameWorks, if you need an abnormally high 780TI OC to beat a 290X, surely if the game was rigged then the 780ti's would be stomping all comers with no 290X sitting at the number 2 spot

the extremetech article that started all this was written well after release, and the article writer ONLY tested a 290 an a 770 and specifically at lower settings, when you start running the game at different settings and on a range of cards, all of a sudden the "vendor bias" is not so apparent

ICDP · 12 Jun 2014 at 11:58

AMD = good guys
Nvidia = bad guys

How hard dis it to understand. Why can't anyone accept that AMD are only out to help us the poor consumer. They would never take advantage of use or lie to us to get a sale.

Disclaimer: The above statement is pure sarcasm

CAT-THE-FIFTH · 12 Jun 2014 at 12:11

andybird123 said:
Yeah, most Titan's top out at bout 1200mhz without voltage mods (e.g. custom bios) and 780's not much better, but 780ti's will hit 1300+ pretty easily
VSG's score in that thread was at 1463mhz

where as Kaap's 290X score is at 1250mhz

hence why Gregster is saying, what is the problem with Arkham City / GameWorks, if you need an abnormally high 780TI OC to beat a 290X, surely if the game was rigged then the 780ti's would be stomping all comers with no 290X sitting at the number 2 spot

the extremetech article that started all this was written well after release, and the article writer ONLY tested a 290 an a 770 and specifically at lower settings, when you start running the game at different settings and on a range of cards, all of a sudden the "vendor bias" is not so apparent

Like I said the whole issue was at launch and the time period afterwards,so it will be interesting to see at what point the performance difference started degrading.

BTW, high GTX780TI overclocks could be explained in part due to the Turbo mechanisms.

If the GTX780TI does boost quite highly with a decent cooler and cooling at stock above 1GHZ(this is why I asked),then that means a overclock on a R9 290X would lead to a higher percentage increase at the same clockspeeds. IIRC the R9 290 and R9 290X have a max Turbo limit of 1GHZ,with a throttle limit of around 700MHZ to 750MHZ. At least that is what I gathered.

_Alatar_ · 12 Jun 2014 at 12:31

LtMatt said:
Deferred contexts improve performance for the first five minutes of gameplay, then the accumulated contexts eventually slow down the framerates below an engine without them. But how long do benchmarks take? Shrewd, no? How long is StarSwarm? 5 minutes.

Bullcrap.

I'd go as far as saying that Nvidia's DX11 implementation does better over time than Mantle in star swarm.

Here's a 15min run of DX11 (underclocked titan) and 15min run of mantle (900MHz 7950)

That btw is the Titan averaging 39fps while the 7950 is doing 23. Underclocked titan being 71% faster than an overclocked 7950.

And to top it off Star Swarm's animation / events are tied to frames, not time. Meaning that the 7950 had a much easier time because it didn't have to render the huge amount of ships for as long as the Titan did. As you can see the Titan rendered a much higher amount of frames.

We see Star Swarm fps drop over time because the demo throws more and more crap at you over time. But this happens on both mantle as well as DX11.