When will this game update engine to DX12?

[Stand The Wall.6987]

it would be nice to not have to deal with miles of skill lag, you know to actually be able to play the game.

 

if gw3 comes out, I hope they use software from a video game graphics company.

[Ayrilana.1396]

> @"hunkamania.7561" said:

> > @"Trise.2865" said:

> > Do you have ANY idea what it will take to rewrite the entire game engine?

>

> excuses. That's for losers. How come someone made some dx12 thing for gw2 for FREE? People say making the game run smooth isn't worth it is laughable.

 

The amount of performance boost may not be worth it for the resources spent. Players have many ways to improve the performance on their end.

[Obtena.7952]

> @"hunkamania.7561" said:

> > @"Trise.2865" said:

> > Do you have ANY idea what it will take to rewrite the entire game engine?

>

> excuses. That's for losers. How come someone made some dx12 thing for gw2 for FREE? People say making the game run smooth isn't worth it is laughable.

 

No, it's not excuses ... GW2 is part of a business. They aren't going to do something to the game unless they feel it's worth it. DX12 ... not much value there ... high cost, low returns.

[Turkeyspit.3965]

> @"Ayrilana.1396" said:

> > @"FONGORE.7410" said:

> > > @"Obtena.7952" said:

> > > Hopefully never, because it's a low value proposition to players. I've played a game that upgraded their DX once ... it was fun for about 5 minutes ... then it was the same game all over again.

> >

> > Low value? The d912pxy addon alone that's basically wrapping dx9 to dx12 gives a 30 - 40% frame rate boost. That's not a low value. I call 30%+ frame improvement pretty dang significant bump up in value.

>

> Sure... 30%...

>

> And when I tested it on both my PC's, I saw roughly half that or lower depending on which PC.

>

> Will you see a performance boost? Sure. Will it be as large as people are claiming? No.

>

> Come to think of it, giving a percentage boost when nobody shares the same baseline is misleading. If I was getting 2 FPS during world bosses and now get 10 FPS with it, that's like a 500% performance increase! Would it be accurate to say that everyone would see that? Would that increase really be significant when it comes to performance? Am I going to notice the difference between 2FPS and 10FPS?

 

It's the same thing as when people claim a tax break only favors the wealthy. Logic (and math) would stipulate that the more money you earn, the more taxes you pay, and therefore the more savings you would see from a tax break.

 

The lower your hardware capabilities are, the less impact you would see from switching to the d912pxy addon. My computer is very much due for an upgrade (coincidentally, which is coming soon thanks to a tax refund :sunglasses: ) but at the moment I am seeing about a 30% bump in min FPS. Those with more powerful systems would obviously see less of an impact.

[Turkeyspit.3965]

> @"Obtena.7952" said:

> > @"hunkamania.7561" said:

> > > @"Trise.2865" said:

> > > Do you have ANY idea what it will take to rewrite the entire game engine?

> >

> > excuses. That's for losers. How come someone made some dx12 thing for gw2 for FREE? People say making the game run smooth isn't worth it is laughable.

>

> No, it's not excuses ... GW2 is part of a business. They aren't going to do something to the game unless they feel it's worth it. DX12 ... not much value there ... high cost, low returns.

 

That's your opinion. The only way to know for sure would be to take a measure of how many players have quit GW2 / discontinued following through with a purchase after trying it out as F2P, mainly because of performance issues.

 

It isn't unreasonable for a player to assume a 7 year old game would perform effortlessly on modern hardware, but as you can see in the bugs section of the forum, there are players with $2000 computers struggling to keep above 30FPS. That can't be good for business. Perhaps upgrading to DX12 isn't attractive to veteran players, but I think it is fair to assume it would discourage new players from purchasing the product.

[TinkTinkPOOF.9201]

> @"Trise.2865" said:

> Do you have ANY idea what it will take to rewrite the entire game engine?

 

What does rewriting the game engine and the API it uses have to do with each other?

 

People seem to not understand that DX12 is an API, it is NOT a game engine, nor does changing the API of a game engine mean starting over. Some game engines have many API's. And API is an interface, the game engine is the functionality that can be made up of all sorts of assets including tools that are used by the developer to aid in faster creation of the game, such as authoring tools, model editors, importers etc etc. Most game engines include an API, but the API does not define the game engine and can be changed.

 

The API standardizes how the engine interacts with other software and hardware so it knows what to expect, direct hardware access is also banned by most OS's, such as Windows, Linux and OSX. However, each bit of hardware is different, AMD and NV is not even the same kind of architecture, and that does not include proprietary hardware designs such as RT cores for AI or Ray tracing acceleration. Those vendor specific drivers would never work without a standard API such as DX12.

[Ayrilana.1396]

> @"Turkeyspit.3965" said:

> > @"Ayrilana.1396" said:

> > > @"FONGORE.7410" said:

> > > > @"Obtena.7952" said:

> > > > Hopefully never, because it's a low value proposition to players. I've played a game that upgraded their DX once ... it was fun for about 5 minutes ... then it was the same game all over again.

> > >

> > > Low value? The d912pxy addon alone that's basically wrapping dx9 to dx12 gives a 30 - 40% frame rate boost. That's not a low value. I call 30%+ frame improvement pretty dang significant bump up in value.

> >

> > Sure... 30%...

> >

> > And when I tested it on both my PC's, I saw roughly half that or lower depending on which PC.

> >

> > Will you see a performance boost? Sure. Will it be as large as people are claiming? No.

> >

> > Come to think of it, giving a percentage boost when nobody shares the same baseline is misleading. If I was getting 2 FPS during world bosses and now get 10 FPS with it, that's like a 500% performance increase! Would it be accurate to say that everyone would see that? Would that increase really be significant when it comes to performance? Am I going to notice the difference between 2FPS and 10FPS?

>

> It's the same thing as when people claim a tax break only favors the wealthy. Logic (and math) would stipulate that the more money you earn, the more taxes you pay, and therefore the more savings you would see from a tax break.

>

> The lower your hardware capabilities are, the less impact you would see from switching to the d912pxy addon. My computer is very much due for an upgrade (coincidentally, which is coming soon thanks to a tax refund :sunglasses: ) but at the moment I am seeing about a 30% bump in min FPS. Those with more powerful systems would obviously see less of an impact.

 

Likely why the PC I built last year didn’t show as much of an improvement compared to the one I build in 2012.

 

One of the things I had brought up was whether the performance increase would actually be noticeable. I know that theirs a physiological impact that if you expect something to perform better then it may appear to do so even though it doesn’t.

 

I had originally posted a link to a YouTube video that compared the performance with and without the add-on. It wouldn’t leave it as just the link so I opted to remove it entirely. In that video, they found that their average FPS increased by 23% which was about 5-6 FPS. I also looked at start FPS of 10, 20, 30, and 40 and found that a 30% increase you only be 3, 6, 9, and 12 FPS respectively.

 

Are people, on average, able to tell the difference with FPS that small? I mean, can you look at two videos and tell which one is 3 FPS faster? And when it comes to performance in the game, are those extra FPS enough to really make a difference in the performance of the game? If I’m doing the Svanir WB where I get 10 FPS without the add-on, would the 30% boost of 3 FPS with the add-on really be noticeable or make a difference?

[Turkeyspit.3965]

> @"Ayrilana.1396" said:

> > @"Turkeyspit.3965" said:

> > > @"Ayrilana.1396" said:

> > > > @"FONGORE.7410" said:

> > > > > @"Obtena.7952" said:

> > > > > Hopefully never, because it's a low value proposition to players. I've played a game that upgraded their DX once ... it was fun for about 5 minutes ... then it was the same game all over again.

> > > >

> > > > Low value? The d912pxy addon alone that's basically wrapping dx9 to dx12 gives a 30 - 40% frame rate boost. That's not a low value. I call 30%+ frame improvement pretty dang significant bump up in value.

> > >

> > > Sure... 30%...

> > >

> > > And when I tested it on both my PC's, I saw roughly half that or lower depending on which PC.

> > >

> > > Will you see a performance boost? Sure. Will it be as large as people are claiming? No.

> > >

> > > Come to think of it, giving a percentage boost when nobody shares the same baseline is misleading. If I was getting 2 FPS during world bosses and now get 10 FPS with it, that's like a 500% performance increase! Would it be accurate to say that everyone would see that? Would that increase really be significant when it comes to performance? Am I going to notice the difference between 2FPS and 10FPS?

> >

> > It's the same thing as when people claim a tax break only favors the wealthy. Logic (and math) would stipulate that the more money you earn, the more taxes you pay, and therefore the more savings you would see from a tax break.

> >

> > The lower your hardware capabilities are, the less impact you would see from switching to the d912pxy addon. My computer is very much due for an upgrade (coincidentally, which is coming soon thanks to a tax refund :sunglasses: ) but at the moment I am seeing about a 30% bump in min FPS. Those with more powerful systems would obviously see less of an impact.

>

> Likely why the PC I built last year didn’t show as much of an improvement compared to the one I build in 2012.

>

> One of the things I had brought up was whether the performance increase would actually be noticeable. I know that theirs a physiological impact that if you expect something to perform better then it may appear to do so even though it doesn’t.

>

> I had originally posted a link to a YouTube video that compared the performance with and without the add-on. It wouldn’t leave it as just the link so I opted to remove it entirely. In that video, they found that their average FPS increased by 23% which was about 5-6 FPS. I also looked at start FPS of 10, 20, 30, and 40 and found that a 30% increase you only be 3, 6, 9, and 12 FPS respectively.

>

> Are people, on average, able to tell the difference with FPS that small? I mean, can you look at two videos and tell which one is 3 FPS faster? And when it comes to performance in the game, are those extra FPS enough to really make a difference in the performance of the game? If I’m doing the Svanir WB where I get 10 FPS without the add-on, would the 30% boost of 3 FPS with the add-on really be noticeable or make a difference?

 

Well for me, running an 8-core with a GTX 960, I went from 15-20fps in Mistlock sanctuary to around 30-35ish. The most noticeable difference for me though was before simply panning my camera displayed visual stuttering, while now it's a smooth rotation. Before the mod there were times when during the AB meta I'd drop to 5-7fps, where now when things get bad, it's in the 15-20fps range. And while I still fall below 30fps in world boss fights or WvW zerg battles, the new wrapper has definitely helped keep the min level of FPS to an at least playable level.

 

My wife's computer is an i5 with a 1060, and she never drops below 35fps, so for her it makes no sense to bother installing the mod.

[TheQuickFox.3826]

It should be relatively easy for a skilled programmer to make a DX11/DX12 renderer, but this will not be the ultimate solution fo the performance issues. But it may help a bit, I would love to try. (DX11 here, because of Windows 7.)

 

More performance can be gained with engine improvements, but this will be much harder to achieve.

[Obtena.7952]

> @"Turkeyspit.3965" said:

> > @"Obtena.7952" said:

> > > @"hunkamania.7561" said:

> > > > @"Trise.2865" said:

> > > > Do you have ANY idea what it will take to rewrite the entire game engine?

> > >

> > > excuses. That's for losers. How come someone made some dx12 thing for gw2 for FREE? People say making the game run smooth isn't worth it is laughable.

> >

> > No, it's not excuses ... GW2 is part of a business. They aren't going to do something to the game unless they feel it's worth it. DX12 ... not much value there ... high cost, low returns.

>

> That's your opinion. The only way to know for sure would be to take a measure of how many players have quit GW2 / discontinued following through with a purchase after trying it out as F2P, mainly because of performance issues.

>

> It isn't unreasonable for a player to assume a 7 year old game would perform effortlessly on modern hardware, but as you can see in the bugs section of the forum, there are players with $2000 computers struggling to keep above 30FPS. That can't be good for business. Perhaps upgrading to DX12 isn't attractive to veteran players, but I think it is fair to assume it would discourage new players from purchasing the product.

 

Not really my opinion ... I don't think majority of the people that play this game do it because it performs well on their computer. I am confident that the primary reason people that play value it's content and other selling features above all.

[Ashantara.8731]

> @"Kanok.3027" said:

> > @"Blocki.4931" said:

> > Maybe it's not completely out of question, but highly unlikely as of now. Though wouldn't it also mean that people with older rigs would get screwed over? Wouldn't be the best decision.

>

> Well, DX12 can work on Windows 7 now (since they backported it to Win 7 recently), but since Win 7 is going to be phased out soon, yeah. DX12 would not be ideal for those that will still be running Windows 7.

 

I am a huge fan of Windows 7. But running it after support expires (and thereby security patches will no longer be provided) would be careless, to say the least. Best to update to Windows 8.1 then or look for another alternative (anything _but_ Windows 10).

[Emberstone.2904]

> @"TinkTinkPOOF.9201" said:

> > @"Trise.2865" said:

> > Do you have ANY idea what it will take to rewrite the entire game engine?

>

> What does rewriting the game engine and the API it uses have to do with each other?

>

> People seem to not understand that DX12 is an API, it is NOT a game engine, nor does changing the API of a game engine mean starting over. Some game engines have many API's. And API is an interface, the game engine is the functionality that can be made up of all sorts of assets including tools that are used by the developer to aid in faster creation of the game, such as authoring tools, model editors, importers etc etc. Most game engines include an API, but the API does not define the game engine and can be changed.

>

> The API standardizes how the engine interacts with other software and hardware so it knows what to expect, direct hardware access is also banned by most OS's, such as Windows, Linux and OSX. However, each bit of hardware is different, AMD and NV is not even the same kind of architecture, and that does not include proprietary hardware designs such as RT cores for AI or Ray tracing acceleration. Those vendor specific drivers would never work without a standard API such as DX12.

 

Yes, the API can be changed, but let's spin this a little bit differently.

 

Let's say you have a program written in C++, but you want to port it to Java. Most programming languages offer the exact same set of features at a base level (various kinds of loops, function libraries with a plethora of math/memory/etc. operations figured out for you, etc.). Even if you're only using basic libraries containing functions that can be, more or less, found across many different languages (like square roots), a program with 10,000 lines of code in total will still take a pretty large amount of time to make the like-for-like translation, and then do all the necessary R&D to ensure it works as it did before. This isn't a perfect example, but it should get the point across.

 

Now, I'm pretty sure a dev said somewhere that GW2 has more than 10,000 lines of code, and surely a game of this scale uses a lot of custom-made stuff in its graphics pipeline rather than canned DX9 functions in order to get the game to work the way they want it to. Making that translation on custom code on a scale this large likely won't happen, even if it's just an API change. Sure they wouldn't be changing the core logic of the game, but every graphics call is going to have to be altered and that's still no small feat. DX12 is markedly different than DX9.

 

So, even if you're just translating GW2's code from one form to another, there's still a boatload of work to do to get it all done and ensure it works. And it can't just be a simple translation from DX9 to 12 since there are some major differences that have to be accounted for.

 

Don't get me wrong, I would absolutely love to see this game get a big engine update. Dipping to 20 FPS unless I lower shadows and character models in Dragon's Stand is not fun, especially when I have an overclocked 8700K and GTX 1070. My GPU is barely utilized even at 1440p in these scenarios since the CPU can't keep up. But like I said, because of the time investment a drastic update to the graphics pipeline likely isn't going to happen.

 

But I'm praying that it does with you.

[Ayrilana.1396]

I know that the number of player models on screen had a large impact on a computer’s performance. What about the visual clutter from other players’ skills and abilities?

 

Instead of revamping the entire engine to work with DX12, would a better option be to separate those visuals from other players out from your own into a toggle? Would that be doable? Would it take up less resources? What impact would it have on performance?

 

Aside from performance, I know that this has been requested so you can simply see what is going on rather than stand there staring at the light show. Being able to toggle this may have a bigger benefit to the player base than simply upgrading to DX12. That all assuming that the visual clutter impacts FPS by any noticeable degree and that adding the toggle would be worth the resources spent. .

 

[Kanok.3027]

> @"Ashantara.8731" said:

> I am a huge fan of Windows 7. But running it after support expires (and thereby security patches will no longer be provided) would be careless, to say the least. Best to update to Windows 8.1 then or look for another alternative (anything _but_ Windows 10).

 

Same about Windows 7, but this is why I tested GW2 on Linux and it's great. Same performance as Windows. Install was easy and got right into the game without issue. Will probably be moving there...and keep at least one PC as Windows because I have no choice (digital musical workstation PC where iLok doesn't work in WINE/VM).

[Ben K.6238]

> @"Turkeyspit.3965" said:

> It isn't unreasonable for a player to assume a 7 year old game would perform effortlessly on modern hardware, but as you can see in the bugs section of the forum, there are players with $2000 computers struggling to keep above 30FPS. That can't be good for business. Perhaps upgrading to DX12 isn't attractive to veteran players, but I think it is fair to assume it would discourage new players from purchasing the product.

 

The only way a modern $2000 computer is going to struggle to keep above 30FPS is if the player cheaped out on the CPU and is trying to run it on a 4K screen. I spent half that two years ago, and am averaging 100-120 FPS in this game at 1920x1200. "Skill lag" is a different story, since the only way I could get a reasonable ping is with a server that isn't halfway across the world.

[Emberstone.2904]

> @"Ben K.6238" said:

> > @"Turkeyspit.3965" said:

> > It isn't unreasonable for a player to assume a 7 year old game would perform effortlessly on modern hardware, but as you can see in the bugs section of the forum, there are players with $2000 computers struggling to keep above 30FPS. That can't be good for business. Perhaps upgrading to DX12 isn't attractive to veteran players, but I think it is fair to assume it would discourage new players from purchasing the product.

>

> The only way a modern $2000 computer is going to struggle to keep above 30FPS is if the player cheaped out on the CPU and is trying to run it on a 4K screen. I spent half that two years ago, and am averaging 100-120 FPS in this game at 1920x1200. "Skill lag" is a different story, since the only way I could get a reasonable ping is with a server that isn't halfway across the world.

 

I use an overclocked 8700K and still dip below 30 FPS in events like AB, Dragon's Stand, and Teq if I run high shadows and high character model limit.

 

It's actually really easy to dip below 30 even on powerful hardware.

[Ashantara.8731]

> @"Kanok.3027" said:

> > @"Ashantara.8731" said:

> > I am a huge fan of Windows 7. But running it after support expires (and thereby security patches will no longer be provided) would be careless, to say the least. Best to update to Windows 8.1 then or look for another alternative (anything _but_ Windows 10).

>

> Same about Windows 7, but this is why I tested GW2 on Linux and it's great. Same performance as Windows. Install was easy and got right into the game without issue. Will probably be moving there...

 

Yeah, if you use your PC for only one game, that's probably an easy choice. ;) But people like you and me also work on their computers (music, graphics, office stuff), and that's a different story if you don't want to clogg your desk with two computers.

 

Also, I am ashamed to admit that over 20 years of using Windows has made me very dependend on its user interface, and I have trouble to adapt to Linux or other operating systems (although, in the future, switching to a different OS will be the only alternative to using a horrible system like Windows 10). I will definitely install Classic Shell once I switch to 8.1.

[bennypig.6428]

I've been using d912pxy for a while. It's working like a beast. I've never dropped below 30 fps even in meta event. I can never go back now lol.

All this work is by one person. It wOulD TaKE tOo mUcH ReSoURceS.

[Kanok.3027]

> @"Ashantara.8731" said:

> Yeah, if you use your PC for only one game, that's probably an easy choice. ;) But people like you and me also work on their computers (music, graphics, office stuff), and that's a different story if you don't want to clogg your desk with two computers.

 

Not really. Most, if not all, of my gaming library is playable on Linux. It's come a VERY long way since it's days as a niche alternative to Windows and Mac. For gaming, video editing and image editing, streaming, etc. I could easily cut the cord. The only issue, and I mean the only one I face, is iLok products, of which I only have for my main instrument library. If I shelled out more money, I could replace it and never bother with Windows ever again, but since I'm not rich, I still have to dual. Until iLok can be successfully run under a Linux OS in some way, shape or form, or until they bring a version of it to Linux, I and others that use iLok software/hardware are left with little choice. However, some have reported success, so I might look into that very soon.

 

 

> @"bennypig.6428" said:

> All this work is by one person. It wOulD TaKE tOo mUcH ReSoURceS.

 

A very hacky way of running it, yes, with no official support. I'd rather not have it and not deal with it. Believe me, if it really was all that simple, Anet would have done it by now, but obviously people think that because someone made a hacky workaround that it's really simple. lol Oh well...

[crepuscular.9047]

> @"mercury ranique.2170" said:

> > @"crepuscular.9047" said:

> >

> > the GPU side of the things are fine, the problem is on the CPU side of the things that requires massive amount of computation like model positioning, skills executions, damage counters, etc

> >

> > anet just keep on adding more and more things that the CPU needs to calculate within seconds, it's really bloated

> > unless Anet figures out a way to properly separate calculation processes between CPUs to work in parallel, we are pretty much stuck like this; that said, multi-thread processing is difficult because of the computation is very serial

>

> You asume, this is cause they are incapable. I think that assumption is wrong and it is on purpose.

> GW1 and GW2 has always been friendly on low end computers. I can run the game reasonable well on computers running intel GPU. GW1 could run on almost every system capable of running windows XP. This has the advantage that you are inviting and open for almost all new players to the genre and even to online games. People who are new will not likely want to invest into a brand new computer that cost a 4 figure amount of money. So the filosophy seems to be that someone must be able to install the game and try it out, regardless of their system. That is why it is this way. They know it limits those with high end systems (I also have a high end rig), but it is a well thought through trade off and not a lack of skills

 

oh, i dont assume, i know, my old 4770K was running perferctly smoothly on medium settings for years until PoF came out, was unplayable unless set to lowest setting

Especially when people jumps at mob spawn groups with raptor tail swipe, the machine will go nuts maxing out the CPU and getting freezes for couple of seconds

 

i was basically forced to upgrade to what I'm running now, 8700K thanks for PoF

[Stalkingwolf.6035]

> @"kharmin.7683" said:

> Yeah, it would take a complete re-coding of the entire game. Check the forums search function as this has been discussed quite a bit.

 

if this would be the case, the code is a mess. you have layer/broker for audio, video, database so you can change it without change your whole code.

i can not imagine that they did not develop with this in mind.

World of Warcraft got DX12 support and it was a huge performance boost.

[Ashantara.8731]

> @"Kanok.3027" said:

> > @"Ashantara.8731" said:

> > Yeah, if you use your PC for only one game, that's probably an easy choice. ;) But people like you and me also work on their computers (music, graphics, office stuff), and that's a different story if you don't want to clogg your desk with two computers.

>

> Not really. Most, if not all, of my gaming library is playable on Linux. It's come a VERY long way since it's days as a niche alternative to Windows and Mac. For gaming, video editing and image editing, streaming, etc. I could easily cut the cord.

 

Really? Can you run any MS Office products on it? Or an Adobe Suite with Photoshop etc.? If so, that would be great as Windows 8.1 will certainly be the last Windows I'll use, unless MS stops their silly Windows 10 policy and goes back to their OS not being one that constantly updates to new versions, causing tons of issues for gamers and others. (P.S. AFAIK, there are alternatives to Windows, Linux and Mac altogether, but I would have to thoroughly look into them first.)

 

 

[Thomanson.4765]

> @"Ben K.6238" said:

> > @"Turkeyspit.3965" said:

> > It isn't unreasonable for a player to assume a 7 year old game would perform effortlessly on modern hardware, but as you can see in the bugs section of the forum, there are players with $2000 computers struggling to keep above 30FPS. That can't be good for business. Perhaps upgrading to DX12 isn't attractive to veteran players, but I think it is fair to assume it would discourage new players from purchasing the product.

>

> The only way a modern $2000 computer is going to struggle to keep above 30FPS is if the player cheaped out on the CPU and is trying to run it on a 4K screen. I spent half that two years ago, and am averaging 100-120 FPS in this game at 1920x1200. "Skill lag" is a different story, since the only way I could get a reasonable ping is with a server that isn't halfway across the world.

 

Someone hasn't been in a crowd lately. The game's bottlenecked to hell. Getting away from dx9 would only help.

[TinkTinkPOOF.9201]

> @"Emberstone.2904" said:

> > @"TinkTinkPOOF.9201" said:

> > > @"Trise.2865" said:

> > > Do you have ANY idea what it will take to rewrite the entire game engine?

> >

> > What does rewriting the game engine and the API it uses have to do with each other?

> >

> > People seem to not understand that DX12 is an API, it is NOT a game engine, nor does changing the API of a game engine mean starting over. Some game engines have many API's. And API is an interface, the game engine is the functionality that can be made up of all sorts of assets including tools that are used by the developer to aid in faster creation of the game, such as authoring tools, model editors, importers etc etc. Most game engines include an API, but the API does not define the game engine and can be changed.

> >

> > The API standardizes how the engine interacts with other software and hardware so it knows what to expect, direct hardware access is also banned by most OS's, such as Windows, Linux and OSX. However, each bit of hardware is different, AMD and NV is not even the same kind of architecture, and that does not include proprietary hardware designs such as RT cores for AI or Ray tracing acceleration. Those vendor specific drivers would never work without a standard API such as DX12.

>

> Yes, the API can be changed, but let's spin this a little bit differently.

>

> Let's say you have a program written in C++, but you want to port it to Java. Most programming languages offer the exact same set of features at a base level (various kinds of loops, function libraries with a plethora of math/memory/etc. operations figured out for you, etc.). Even if you're only using basic libraries containing functions that can be, more or less, found across many different languages (like square roots), a program with 10,000 lines of code in total will still take a pretty large amount of time to make the like-for-like translation, and then do all the necessary R&D to ensure it works as it did before. This isn't a perfect example, but it should get the point across.

>

> Now, I'm pretty sure a dev said somewhere that GW2 has more than 10,000 lines of code, and surely a game of this scale uses a lot of custom-made stuff in its graphics pipeline rather than canned DX9 functions in order to get the game to work the way they want it to. Making that translation on custom code on a scale this large likely won't happen, even if it's just an API change. Sure they wouldn't be changing the core logic of the game, but every graphics call is going to have to be altered and that's still no small feat. DX12 is markedly different than DX9.

>

> So, even if you're just translating GW2's code from one form to another, there's still a boatload of work to do to get it all done and ensure it works. And it can't just be a simple translation from DX9 to 12 since there are some major differences that have to be accounted for.

>

> Don't get me wrong, I would absolutely love to see this game get a big engine update. Dipping to 20 FPS unless I lower shadows and character models in Dragon's Stand is not fun, especially when I have an overclocked 8700K and GTX 1070. My GPU is barely utilized even at 1440p in these scenarios since the CPU can't keep up. But like I said, because of the time investment a drastic update to the graphics pipeline likely isn't going to happen.

>

> But I'm praying that it does with you.

 

C++ and Java would be closer to the engine, as both already have their own APIs defined in the SDK. Java and C++ are also very different animals, as Java assumes a full fledged OS, while C++ could be the programing of the OS it self. One assumes an OS and the output being an application of some sort, while the other makes almost no assumptions about anything, as C++ can be used for almost anything and as such the standard library assumes almost nothing about what it will have available, and so the standard library doesn't make any dependencies on those features.

 

In the case of GW2 we are talking porting it to DX12, not rewriting the whole engine like your example above would require. Now, not saying it's as easy as some people assume, however it seems people either assume it's a click of a button, or they assume it requires a ground up rebuild of the engine, when it's really closer to the middle. As has already been stated, a single unpaid person has written a wrapper for GW2 that implements many of DX12 performance benefits. So what you seem to be saying is that anets whole dev team is less capable than this single unpaid programmer?

 

I am also not praying for anything, as I stated in my first post, DX12 is never going to happen for GW2, the next game to come from this line might, that is assuming there will be anything like a GW3. If they plan on working with GW2 for some time, we might see standalone expansions that move to DX12, WoW saw a 25% gain in FPS across the board when they did this and patched in DX12.

[AlexxxDelta.1806]

I would like to see it too, even if it meant a middling improvement of around 10-15 fps. I don't think it's very likely given the current state of things though.

 

A much more feasible solution would be Anet focusing more on proper optimization. There is no reason for that bone wall in Istan to kill fps even on beastly rigs, for example. And after that, they should consider toning down the graphical improvements they 've been doing in new items and environments. The difference in fps from doing a meta event in core compared to PoF is very noticeable even on pure gaming systems. Less graphical fidelity is a weird thing to consider but if your engine can't handle what you are creating, it will stop being a matter of choice at some point.

 

Crowded events in both OW and WvW are such a big part of this game, you want your players to be able to experience that at a good enough framerate. When even high-end gaming PCs drop below 30 in some of your most popular game activities, there is a problem.