MANTLE VS GSYNC !!!!!!!!!!!!!

[Sammael]

The terrible question becomes the terrible TOPIC !!!!!!!!!!!!!!!!!!!!!!

 

But seriously, what do you all think has more value?

 

We know what camp linus/ryan are in, they are solidly in the gsync camp.  I have a couple of questions and comments though. 

 

 

 

What would be harder, for Nvidia to implement some sort of mantle alternative or general support, or for AMD to implement some sort of gsync alternative or support?

 

 

From a technical standpoint, doesn't it seem like an AMD version of gsync is a MUCH easier technical lift to implement later on?  Would they need specialized hardware on the gpus, or could they just do a software update coupled with some amd equipped monitors later on to get the same effect?

 

 

Also, while the guys were talking about the relative value of the Kaveri APU, they seemed to miss a key point as to why it might be worth the 40 dollar premium.  Kaveri has a GCN gpu, that comes with mantle support and true audio.  One of the pieces of info that came out of the APU event was that mantle allows for asynchronous GPU scaling, such that you could use the gpu in the kaveri apu to work in tandem to MUCH better effect than before with a discreet gpu.

 

http://www.youtube.com/watch?v=tDPgJB2x7dQ#t=33m16s

 

 

With mantle, they are far less cpu constrained, so the lower ipc of Kaveri would not be as much of a negative for gaming (not sure about this, core count might still matter if a game uses 8 cores vs the 4 cores [2 modules] that kaveri brings).  But a Kaveri apu would give a "FREE" gpu boost to anyone who intended to couple it with a discreet gpu with mantle supported games.

 

Mantle + Kaveri + discreet gpu can radically bring down the cost of a performance gaming system.  You could save maybe 50 dollars going with kaveri vs an intel, perhaps another 100 dollars by going with a lower card than you would have gone with before with the knowledge that the gpu in kaveri would make up the difference in mantle games.

 

Imagine notebooks, kaveri mobile chips with modest discreet gpus, imagine the relative cost savings and performance for the price that could be achieved.

 

This seems like a much bigger deal, as nice as gsync is, it seems like lower hanging fruit that amd could easily implement quickly later on, and perhaps on better monitors.

 

imagine real time strategy games that slow to a crawl when too many units come on screen, these types of games would sing with greater draw call support, what would cripple an nvidia gpu might be much more playable on an AMD gpu.

 

It just seems obvious to me MANTLE is a bigger deal in terms of potential.

[Altecice]

You cant compare Gsync to Mantle. Two different things

 

ones a hardware thing to match refresh rates the other is a software graphics API

 

Also linus/Ryan are ultra Nvidia  fan boys so that's obv.

 

I am nether but i would say Mantle is the better achievement. (bypassing the 5k limit etc)

[delet my acc pls]

We don`t know since mantle is not out yet.

 

BTW you are comparing apples and oranges mate.

[Scottonpc]

These are completely different things. Why are we comparing them?

[cjeffcoatjr]

I think G-Sync will be much better. imo mantle will just be another api in the list with DX and Open CL, and it wont be any better. I think that the '%50' percent improvement numbers that AMD is giving is on small stuff like Athlon's's with huge GPU's like 7990's. I dont think I will achieve 20 more frames on my 7870. 

[Guest GRRigger]

You cant compare Gsync to Mantle. Two different things

 

ones a hardware thing to match refresh rates the other is a software graphics API

 

Also linis/Ryan are ultra Nvidia  fanboys so thats obv.

 

I am nether but i would sat Mantle is the better achievement. (bypassing the 5k limit etc)

Yea, but you only can have one Nvidia G-Sync or AMD Mantle. You have to choose one you want to use.

It all comes down to G-Sync price and Mantle performance battle. 

My opinion is that G-Sync will win on the long run, but mantle will be a thing when it comes out IF it works. 

And adoption is the biggest thing of all in the beggining. On both technologies.

[skullbringer]

From my experience, even with dual 780s, games are still not as fluid or enjoyable as they could be with more exact image delivery by the monitor / gsync. I dont care about mantle ant its 10% more performance at all, gsync is the much more senseable graphics card feature IMHO.

[MzCatieB]

yay... lets compare something thats unrelated...

Boobs and minecraft... :3 

 

Also:

Mantle = AMD

CUDA = Nvidia

 

Would of been more logical  

[Guest GRRigger]

yay... lets compare something thats unrelated...

Boobs and minecraft... :3 

 

Also:

Mantle = AMD

CUDA = Nvidia

 

Would of been more logical  

Boobs or Minecraft. Why not both? Same with g-sync and mantle We want both, on all platforms But i guess that is not possible.

[Sammael]

From my experience, even with dual 780s, games are still not as fluid or enjoyable as they could be with more exact image delivery by the monitor / gsync. I dont care about mantle ant its 10% more performance at all, gsync is the much more senseable graphics card feature IMHO.

 

 

And therein lies the rub, it sounds like mantle is not just about something as trivial as a 10% performance improvement.  It could allow something like well over a thousand units on the screen vs a couple hundred on nvidia hardware with similar performance.  That's not a trivial fps achievement.

 

 

To those saying they can't be compared, it kind of misses the point of the topic.  To put it another way, what is a bigger deal?  It may be that with current games designed with the current draw call limitations, gsync provides better benefits, but all amd would need to do is flip a switch in their gpus with some open source alternatives for monitors, and they would be at parity there, and rather quickly.  But the technical advantages of mantle seem to greatly outstrip what nvidia currently has or allows, we'll see how it all shakes out.

[Sky Daddy]

Pretty sure nvidia cards will still benefit from mantle, just not as much. Mantle will definitely be superior in cpu optimization. 

 

Only thing with mantle that makes AMD special is its console-like optimization for their GCN architecture.

[TheMissxu]

Mantle with 144Hz monitor trumps G-Sync with 144Hz monitor.

 

Because the G-Sync would cost ~$200 more (~$100 more for the GPU and ~$100 more for the monitor), it's out of the question. Now if G-Sync was free... That would be a different story; none-the-less, my ties are with AMD.

[Altecice]

Mantle with 144Hz monitor trumps G-Sync with 144Hz monitor.

 

Because the G-Sync would cost ~$200 more (~$100 more for the GPU and ~$100 more for the monitor), it's out of the question. Now if G-Sync was free... That would be a different story; none-the-less, my ties are with AMD.

 

G-sync will also only be TN panel only. the tech doesn't work with IPS.

 

IF mantle can provide its 20-50% boost like they say (and lot of big names are using it now) then it will win this "race".... id much rather have a PC that can produce 60fps + than a PC that pushes 40 with Gsync.

[airdeano]

Mantle with 144Hz monitor trumps G-Sync with 144Hz monitor.

 

Because the G-Sync would cost ~$200 more (~$100 more for the GPU and ~$100 more for the monitor), it's out of the question. Now if G-Sync was free... That would be a different story; none-the-less, my ties are with AMD.

 

you say mantle isn't going to cost anything? doubt that. game devs have to

learn/manipulate and implement the new API isn't going to be cheap. so on

the purchase it will be tacked onto the app/game. so developmental costs are

going to passed into the title not the hardware. so you really can't pass judgment

on amount of cost until the API has been implemented. sure hardware will always

have a cost involved price, but so do new platforms.. just an outlook.

[skullbringer]

And therein lies the rub, it sounds like mantle is not just about something as trivial as a 10% performance improvement.  It could allow something like well over a thousand units on the screen vs a couple hundred on nvidia hardware with similar performance.  That's not a trivial fps achievement.

 

 

To those saying they can't be compared, it kind of misses the point of the topic.  To put it another way, what is a bigger deal?  It may be that with current games designed with the current draw call limitations, gsync provides better benefits, but all amd would need to do is flip a switch in their gpus with some open source alternatives for monitors, and they would be at parity there, and rather quickly.  But the technical advantages of mantle seem to greatly outstrip what nvidia currently has or allows, we'll see how it all shakes out.

 

Also the count of units or poligons, generallay graphical elements, isnt as important, as the most accurate frame timing possible.

How much is a beautiful picture worth when its always been torn into jaggy shreds? (exaggerating a bit, but I hope you get my point)

[TheMissxu]

you say mantle isn't going to cost anything? doubt that. game devs have to

learn/manipulate and implement the new API isn't going to be cheap. so on

the purchase it will be tacked onto the app/game. so developmental costs are

going to passed into the title not the hardware. so you really can't pass judgment

on amount of cost until the API has been implemented. sure hardware will always

have a cost involved price, but so do new platforms.. just an outlook.

Let's take Battlefield 4 for example:

 

AMD GPU $400 + 144Hz monitor $266 + Mantle enabled BF4 $20 = Total $686

NVidia GPU $500 + G-Sync enabled 144Hz monitor $400 + BF4 $20 = Total $920

 

BF4 doesn't cost any more when you used Mantle; I don't see how "[the cost] will be tacked onto the app/game." (Also compare BF4's price to CoD:Ghosts' price... Clearly this learning, manipulating, and implementing Mantle costs more money.)

 

Game developers do R&D all the time. Take CryTek for example:

 

Crytek's R&D technical director, Michael Glueck, said 'yes, that [a low level API like Mantle] would appeal to us. ... This will not only improve performance, but it will also allow better use of the available hardware features.''

Don't forget the current list of developers willing to implement Mantle.

 

And another point brought up during an AMD AMA:

 

Q. What was the influence to do something like Mantle when we have seen other projects like "GLIDE" fail? What do you think will make your solution better or more enduring?

A. Game developers requested Mantle themselves. That's the key difference. The industry told us they want it.

Developers are done with DirectX and are willing to move forward, and games wont magically cost $10-20 more because of Mantle support (see BF4 for anecdotal evidence).

[LinusTech]

G-sync will also only be TN panel only. the tech doesn't work with IPS.

 

IF mantle can provide its 20-50% boost like they say (and lot of big names are using it now) then it will win this "race".... id much rather have a PC that can produce 60fps + than a PC that pushes 40 with Gsync.

 

Where is the perception that G-Sync only works with TN panels coming from? Who is reporting this?

 

It's false.

[Askew]

Where is the perception that G-Sync only works with TN panels coming from? Who is reporting this?

 

It's false.

 

I think people may be assuming that because it's marketed as a gaming feature that's initially upcoming on gaming monitors which are very likely to be TN panels.

 

I've not actually seen it reported anywhere.

[TheGhostVII]

They're completely different and do completely different things.

It's like comparing apples to oranges.

Actually, not even apples to oranges.

You've comparing potatoes to pineapples.

[bob345]

not another one of these. Mantle and g-sync are totally different things. There no way you compare them and see "which one is better" because they each serve completely different functions.

[TheGhostVII]

not another one of these. Mantle and g-sync are totally different things. There no way you compare them and see "which one is better" because they each serve completely different functions.

What's better?

This case fan or this thermal paste?

[SlaughterSmurf]

2 completely different concepts..... G-Sync makes shit FPS look good, Mantle gives good FPS, depends what you want.......

[Keff87]

So I've had an idea on this whole back and forth concept, and I've narrowed it down to, "What kind of games do you play?" I see games that are very fast paced like a generic FPS or a racing game would benefit most from G-Sync. Where as games like Elder Scrolls Skyrim would benefit more from Mantle.

 

Since G-Sync, simplified, is the GPU and monitor talking to eachother on when to show frames, missing frames in a FPS game, specially in a PRO setting, could cause a very horrible noob-like death of some crazy proportion.

 

Mantle simplifies allows more graphical 'clutter.' Massively increasing the draw calls cap, means more objects in scenes, and a much farther draw distance from the player. In turn more imersion to the world.

 

So.. if you play a lot of fast paced games, or plan to go PRO... get G-Sync. If you rarely play any fast paced games get mantle.

[rentaspoon]

To those that say you can't compare, he's asking the value, weither you see it as the value to the companies or community or the future of gaming tech that's up to you

[TheGhostVII]

To those that say you can't compare, he's asking the value, weither you see it as the value to the companies or community or the future of gaming tech that's up to you

Again, what they do is too different to compare.