Guild Wars 2

As anticipated, the removal of culling did little to nothing in favor of performance. I still get major FPS drops when more than a few players are in the area, even with the new settings in low/low. The only difference is that I get too see many more name plates.

I hope this patch is enough to convince those advocates of the theory proclaiming the game slowed down because of the engine was unable to render so many characters on the screen. What are you going to say now, are those name plates too graphically demanding?

There was no need of monitoring tools to realize that the severe bottleneck has nothing to do with 3d models or rendering, but lies in the way the engine handles data.
Are they passing objects by value, do they have costly constructor/destructor routines? who knows what the severe blunder is. I am positive directx9 isn’t the only guilty though.

In any case, this will be my last post. I have requested a refund for the game and shortly I’ll be unable to post on the forums anymore. As soon as my refund gets processed my forum and game accounts will be void.

I want to take the chance to wish y’all the best of luck, wherever your MMO activities take you. Whether you decide to stay here or move on to greener pastures.

Happy gaming!

Ok firstly, Occlusion Culling has nothing (at all) to do with the network culling they are talking about.

It has everything to do. It is an additional component of the occlusion culling implementation. In fact, we do not know if it is the only part of the culling implementation. We know for certain there is character culling, for evident reasons. But we can’t ascertain if there’s object culling. For all we know character culling could be the only culling system in the game.

Internally, there is no difference between the two, the same system that tells the GPU not to render a room behind a wall, tells it not to render a character that’s behind your back (Umbra). The difference lies in where the data come from. This game being an online game requires character position data to be supplied by a server, as opposed to the position of other objects/NPCs the game is well aware of.

There is nothing specifically wrong with their implementation of network culling

Tell that to the thousands of frustrated WvW players. They seem to disagree with you. Apparently Anet does too, after all they tried to fix it, and ended up discarding the whole thing.

you seem to associate network culling with client side performance

I make no such claim.
I simply state the fact that, after admitting there was a problem with their culling system and working hard to fix it, Anet developers decided to discard it and implement an alternative, player controlled system. Why this? Perhaps they realized fixing the problem would require making major changes into the network structure of the game engine and decided the alternative would be more feasible/cost effective.

My whole point is, if they decided fixing a relatively simpler issue such as culling would prove to costly/difficult, how likely is it that they’ll fix the infinitely more complex and entangled performance issue.

This is their response, subtle but clear as water.

Guild wars 2 is an exceptional game in so many ways, and a great disappointment in many others. For a while now there have been many a thread, discussing why this is. I believe that a consensus can be achieved that of all the possible reasons, the greatest by far is the terrible performance.

If you never do WvW, PVP, or participate in any crowded dynamic events or visit any major city, you might be prone to disagree on this. For the rest of us this issue is a latent, constant let down and is shoved in our faces daily. Turning an otherwise enjoyable experience, into a cesspool of frustration and sourness.

Countless of times I’ve asked and seen others ask Anet; what is going on with the game, why this atrocious performance and what are they doing about it. All questions falling on deaf ears. They haven’t even acknowledged the issue.

Why is this so ? Why haven’t they at least given us some comfort in the fact that they recognize there is a problem, instead of attributing any performance issues to a user-related problem that only affects a small portion of the players instead of being a game-wide issue? IMO the answer is very simple: Because they can’t do anything about it.

Many have achieved the same conclusion. Some attribute this inability of Anet to fix the game to the way the game was programed; only directX9 and poorly threaded.
Solving this would mean going back to the workbench and working on the game engine almost from scratch. Too late to do it now.

Attributing the reasons for this to a console port is just a conjecture. Personally I think they programmed the game that way to target windows XP users, and because dividing some problems into smaller chunks (multithreading) is not always easy…for less skilled developers. Not to mention they probably weren’t concerned about performance until very late in the development process. (The game wasn’t even optimized until after the first BWE) Probably they expected compiler optimizations would be enough.

In any case, with their latest announcement, The End of Culling, we have finally received our much sought answer. The answer to the infamous question “Will the Lousy FPS be fixed” Now we know what the game programmers have been working on for the past few months.

The performance issue is simple: The game clientt slows down to a slide-show when having many players in the area. What is the solution?
Well they can either:

1-) Investigate the bottleneck in their implementation and rewrite their algorithms to either be more efficient or to utilize more efficiently the machines resources.

2-) Create a “workaround” to the problem.

They went with choice number two. Which essentially consisted of releasing culling computations from the CPU and putting it under your control. To let you decide what is or not to be displayed.

This will most definitely NOT improve performance. At least not by much, or to the extent that is needed. But then again that wasn’t their intention. They simply aimed to fix the “mob invisibility” problem, which resulted from a faulty culling implementation.

Occlusion Culling is a technique used in video games to tell the game engine what to display and what not to, in order to save graphics resources at the expense of adding some CPU burden.
Now, not only was Anet’s culling implementation malfunctioning, but their programmers could not find an adequate solution for it, so they decided to eliminate it and create an alternate system whereas users get to decide what they want to be displayed on the screen. (as far as characters is concerned).

Now if they were unable to properly fix a relatively simpler issue such as culling (instead they removed the system altogether) in the five or so months they’ve been working on it (and they acknowledged the problem), how can we expect for them to fix the much harder, bigger issue that is game performance (which they have never admitted to be a problem with in the first place).

Similarly to a building’s foundation that can’t be changed after the building is built, they can’t make the game engine to perform adequately without tearing down the whole (recycled) thing and making it anew.

Will the lousy FPS be fixed ?
The answer is no my friend. Don’t count on it. Accept it or request a refund and move on. Your call.
I certainly know what I’ll be doing.

Dear sir, thank you for taking the time to help others.

My question: between Power, Prowess and Precision, what is the most important and in what order should we spec these.
I’d be inclined to max my power, then go into precision and finally prowess. What’s
your input?

Krendar.2139:

It’s pretty ridiculous that you say ANet hadn’t considered performance when building this game.

is it ?
They might have considered performance but at at a point in development when it was too late to do much about it, without rewriting much of the code base from scratch. Hence, why we get the terrible performance issues we have, with no acknowledgement from them(since they can’t do anything about it) and no hope for improvement.

Obviously it went through the same types of discussions that every one of the other 100 graphical, third person MMOs has gone through.

Did it ? Maybe, but I bet performance wasn’t there on the design table much, other than the limits to polygon count for the artists and the “target hardware” they were aiming for.
Apparently, they made a severe blunder with their technical implementation, as shown by their stated minimum hardware requirements. When top of the line hardware can’t keep playable performance on the lowest graphical settings, how can the much older hardware they claim should be able to run the game.

It clearly shows that what they “Intended” and the ending results were completely opposite things.

You would have to be a fool to think that ESO and {insert developer} will have no issues with their game because they used 200 players in beta tests.

I’m not claiming they will have no issues. I’m assuring you, if they set as one of their goals to have good performance while having up to 200 players on the screen, thats how it will be. Why ? because they are explicitly looking for that level of performance.

By the number of main threads in GW2 (three), you can tell they weren’t really thinking about performance, as a priority.

It shows no technical leadership, expertise or talent

To have video game developers think about CPU performance from the start of a project, shows expertise, leadership and talent to me. Today’s developers only care about performance when there’s an issue with it. They program first and foremost for readability, maintenance and re-usability of code, leaving performance to compiler optimizations. Performance programming is left only to kernel and drivers developers.

This reminds me of a quote from a book I was reading some time ago, called “Efficient C++: Performance Programming Techniques”, and I paraphrase:
“Contrary to C, you can not stumble on good performance on C++ by accident, you have to actively code for it, navigating through costly language abstractions like on a minefield "

Is this your first MMO? Welcome and please leave your tissues at the door

Assuming that bad performance is inherently equal to an MMO, because your experience with all MMOs has been the same, is understandable. However, let me repeat, this has been the case because those developers have not coded for performance or set it as one of their goals from the start.
In computer science every problem can be solved as long as there are three things, technical expertise, time and money.

Jazhara Knightmage.4389:

watch this one and be in awe
7600gt and core2duo e8300@3.0

http://youtu.be/2qVZLBcXtbg

That’s what happens when capable and talented developers actively code for performance. Bravo.

Jazhara Knightmage.4389:

ESO really is the next mmo Im really looking forward to….
http://tamrielfoundry.com/2013/03/interview-dan-dunham/

read that, and then ask yourself why anet cant get 200people on screen and have a playable situation…….

Jazhara Knightmage.4389, that’s what happens when programmers code for performance from the beginning. It clearly wasn’t the case with this game. But ESO seems to be doing it right.

If ESO delivers at least 50% of what they promised, I will switch in heartbeat as well.
This is so far my favorite quote from the article:

We test every build with around 200 players having a massive fight to make sure that even in the thick of it, the frame-rate stays smooth and the experience is great.

Now that’s what I’m talking about. That shows they have great technical leadership, expertise and talent and that they have great common sense.

Remember to claim your refund and state the reason was because the terrible performance.

IMHO gw2 would have sold more and gotten more good press AND more good word of mouth had they launched it dx11

What are you talking about ?
Didn’t you hear GW2 was the best game of 2012 … according to Anet…

the only excuse for these problems not being discussed and fixed is, anet just dont care…..they dont give a kitten because, they are still making money off people who blame their systems for the problems……..

My hero !!!
Can I use this as my signature?

the question was if guardian shield is better than warrior shield…

No, the question of the OP was for a good pve build that included a shield.
I simply stated that there was no such thing, as any warrior pve build that includes a shield is not going to be good.

Now to back to our argument.
The trait you talk about works only for missiles and works not only for shields but for every weapon that blocks. (sword,mace, etc) So its not really a shield specific trait.
I’m sure there are a lot of guardian traits that benefit and maximize the shield skills for guardians and even if there aren’t, guardian’s shield skills are still far superior.

Let’s do a comparison:
Shield Comparison
I think arguably the engineer has the more powerful shield skills, however the guardian in comparison to the warrior wins hands down.
even though the shield skills for the guardian take a little longer to recharge, its effects are strong enough to merit it’s use.

Guardian first shield skill:
1- gives protection to up to 5 allies and yourself for 5 seconds.
2- damages multiple enemies for 302.
Second Skill:
1-Pushes foes out of the area.
2-Deflects projectiles for 4 secs.
3- Heals for a decent amount.
4- Creates a combo field.

First warrior Shield skill:
1- Damages one foe for 302.
2- Stuns one foe for 1 sec.
3- Executes a combo finisher.

Second shield skill:
1- Block all attacks from all direction for 3 secs.

I think guardian wins hands down. In fact, If it wasn’t for the long recharges, the guardian shield skills would be not only good but great.

PS: when I said mace/sword I meant mace with a sword off hand. Though I agree a mace off-hand is much better than a shield, I was referring to the particularly defensive and effective combination of mace main hand and sword off-hand.

Lalangamena.3694:

IvanHoe.3468:

If you want to play a warrior with a shield you are out of luck. I would recommend you to play a guardian if you want to use a shield, effectively.

not only this is an exaggeration, this is also not true.

warrior shield skills are better for solo play and when traited properly (reflect on block) are superior to guardian shield.

And what trait is that ? I’m not aware of any warrior trait that makes reflect on block.
In fact that was my suggestion to improve the skill.

Even if what you say is true, the recharges are still too long to make the shield viable. 16 sec and 24 secs and thats with the recharge reducing trait.

If you wanna go solo and have a very defensive build and still kill things, nothing beats a mace/sword.
You have an 8 sec recharge block/riposte with mace #2 (with recharge trait). You have a 15 sec recharge block/riposte that causes bleeding with sword #5. You have 12 sec recharge interrupt with mace #3 and finally, you have a decently lasting stun with the mace elite.

Any time I wanna be defensive, I carry mace/sword and it beats the hell out of any shield.

I wasn’t aware that the build editor I posted had outdated skills. In any case, for the warrior at least, the gear and traits seem to be fine. I don’t know why you claim traits are outdated, AFAIK Anet hasn’t changed any warrior traits, just fixed bugs…

I will verify to make sure skills and traits aren’t current for other classes. Also the fact that you claim there is no PVE gear in the build editor shows you have not used the site very much…go take a second look (I fixed the link)
Yep, The build editor is very outdated. 2012/10/07 according to the site. Also there is no PVE back-Item slot. A shame.

Hmmm….Kinda think it’s a little bit silly that warriors aren’t effective with the shield as the traditional stereotype of a warrior is sword/shield

I agree completely. Go tell that to Anet and see if they listen. (good luck with that)

IMO to make the current shield skills remotely effective they would need to reduce the recharge time of #4 to 10 sec and #5 to at least 15 sec. Even then, without any other effect, they would be mediocre. To make them more useful they should increase the stun of skill #4 to be adrenaline based and add a damage reflection effect to skill #5.

I’d also like to take this opportunity to show yall a better build editor. It pains me when I see you use that gw2BuildCtaft site….

Here it is better build editor

If you want to play a warrior with a shield you are out of luck. I would recommend you to play a guardian if you want to use a shield, effectively.

I, like you, love the idea of using the classic sword/shield combination, sadly for warriors, the shield skills are lackluster to say the least and take 100 years to recharge. Even with traits. The sword has mediocre damage and bleeding is the weakest condition in the game, even though it is stackable.

loseridoit.2756:

ummm, Anet acknowledged that culling and fps is a major issue a long time ago.

They said they were investigating why SOME high end computers were having performance issues with the game. As if only a few computers had problems when is well known the game performs terribly for everyone.

Pardon me, but that doesn’t sound like they were acknowledging there’s a problem with the game, much less any major anything.

Anet care so much about how poorly the game runs that they keep us up to date on all the gem shop items they add to the game, or are going to add, but dont say a single word about performance for months on end….

Though I agree 100% with you, I fear for your safety. Keep speaking the truth and you might end up banned !!!
Didn’t you get the memo ? …They…are…watching…

I noticed the link I posted is not working. Something wrong on AB’s site. anyways here’s the comment AB made after Bethesda finally released a patch fixing the issues with the game.

As you may already noticed my last activity was associated with Skyrim, so all below is about Bethesda and Skyboost. The last one was started as a continuation of the project called Tesval by Arisu, he was the one who found out about huge bottlenecks in Skyrim.

To be honest I couldn’t even imagine that all that stuff which was the reason of bottlenecks can cause so huge performance drop. All that Bethesda developers should do was to add optimization flags to the compiler, if someone will say that something-in-the-code was incompatible with optimizations I’ll answer that this is a bullkitten because PS3 and X360 versions were built with turned on optimizations.

Anyway let’s assume that only PC code part was causing some troubles with optimization … and what ? coding is always about glitches and troubles, Bethesda could make the fix easily. But they didn’t ‘til got punched with Skyboost, Tesval and tons of reports from users that two guyz are making things which Beth coders couldn’t.
I can’t believe that so skilled programmers as Bethesda has just forgot to turn on so life critical thing as optimization and only for PC builds since 2006 when Skyrim development started.

So the question is why they didn’t ignore X360 and PS3 optimizations and give a kitten to PC version, for me this becomes pretty obvious when you see that profit from console version almost for every new game is significantly larger than profit from PC, thats my theory, I have no other explanation of what happened. Without Tesval and Skyboost this wasn’t even possible, this projects forced Bethesda to do in patch 1.4.15.0 what they must did along time ago. Also Fallout3 and Fallout: New Vegas have the same problem – PC version optimizations were turned off.

Anyway I’m not stopping the development of *boost project.

Rampage.7145:

-GameBryo is date, slow and poor engine overall, console designed engine, not really a good example it is probably one of the worst engines out there.

It might be so, or not. Still is good enough for what it does. Could it be better ? Absolutely, but the game already performs well on adequate hardware. The same can not be said about GW2.

The fact is not many MMOs out there can handdle a lot ppl casting spells at the same time, GW2 runs bad but not as bad as others to be fair

Name any game that performs worse than GW2 on the same high level hardware and I will gladly buy it and install it just to prove you wrong.

MMO gamers are used to sub 20 FPS is not a big deal, arenanet won’t listen to us, cuz they know 90% of the players don’t see any performance issue don’t know any better.

I completely agree.

TBH i don’t think Arenanet will ever fix this problem

Truer words have never been spoken.

in 2 years CPU will be capable of run this game properlly

You are wrong on this one.

Jazhara Knightmage.4389:

I know that this game uses anets own in house engine, thats just an updated gw1 engine, fingers not typing what brain is thinking….

You didn’t type that, lol. It was the other guy.

after all, quagan backpacks are more important then the game running half decently on peoples systems.

I chuckle every time you say that. So true.

Espionage.3685:

Jazhara Knightmage.4389:

… in-comparison to GW2 (Unreal Engine 3).

GW2’s engine is a proprietary in-house modification of GW’s engine, which was developed by Anet. GW2 does not use the Unreal Engine 3 game engine, LOL, we wish buddy !! that would solve a big chunk of our problems.

Regardless, it is pretty clear that ANet can do some optimizations to make GW2 perform better

Can they ? Perhaps. But they choose not to, so whether they can or not is irrelevant.

I myself have no problem with performance

Even if you play at the lowest settings I doubt that’s true. Go play in any event with more than 30 players on your screen and comeback and say that again.

Asking Anet to acknowledge the problem is like asking a baker to admit his pastries are causing people indigestion. It ain’t going to happen.

TBH, the more I play the more I believe there’s really serious issues with the game. Not just the speculated issues with the directX9 implementation. I’m talking about issues in the engine dating as far back as the first GW. They were less evident because of the much smaller scale and features of the project.

The root of the problem lies in today’s developers mentality. The “Only worry about performance when theres a problem with it” mindset, that most modern developers have. Well I have news for you Einsteins; when you realize there IS a problem it might be too late to fix it without going back and writing your code from scratch.

It’s funny, I’ve been playing the game for less than two weeks and I already realized that crafting, the way it currently is, has no use other than to help level up your character a few levels. That’s it.

I’ve been selling all the mats I get, and with that money I buy the best consumables, armor and weapons for my level on the TP. After that, I still get plenty of $$$ left to put on my bank.

of course, if everyone thought like me, there would be no armor/weps on the TP as no one would craft. Luckily, there are players who craft just for the love of it.

The community paying for the game to be fixed is an excellent idea, however, it needs some adjustments.

First: It needs to be an external contractor, not a game developer, the person or persons paid to do the job. Developers due to contractual obligations would be prohibited from doing such activity. Besides I have my own opinions on the efficacy of the game devs which I’ll hold to myself.

I recall a similar case where a skyrim fan and developer named Alexander Blades fixed skyrim’s atrocious performance before the game developers felt so embarrassed they they decided to fix the game themselves.

Second: due to the way kickstarter works, it isn’t convenient, if at all possible, to use it for this purpose. It needs to be a chip-in style site, where people contribute money via paypal or whatever. Once the money starts getting high enough it will certainly attract the attention of free lance developers. The higher the money the higher chance of attracting the right mind for the job.
It would be similar to a Power Ball lottery ticket, when the money goes high enough, people go crazy buying more and more tickets trying to gain the prize. Similarly we would have more and more developers trying to solve the problem.
And this will be no simple task. Without access to the source code and anet’s server side profiling and benchmarks, only a low level wizard would have any chance of stumbling with a fix.

Solving the underlaying performance problem would most likely require a major rewrite of a big portion of the game engine. I think at this point in time it’s safe to say it’s not going to happen.

If enough people complained loud enough though, ArenaNet would be forced to at least acknowledge the issue, publicly. My bet is this isn’t going to happen either.

if you happen to live nearby, I’d say go to their community meet up, and bring out the issue. I would die to see their response.

Q: I think is common Knowledge the game is severely CPU bound. Why is it so and what is the team doing about it. Can we expect a solution soon ?
A: We are always actively working on improving the game to run on a variety of systems and configurations.

Q: you advertise GW2 as a MassiveMO game. Why is that a $1000 state of the line CPU needs to be over-clocked in order to maintain minimum acceptable FPS on a big dynamic event, city, festival or any place with a sizable gathering of players.
A: We are currently investigating why the game is not performing optimally on some high spec systems.

Q: Can you tell me a high spec system where the game does perform optimally ??
A: There are many possible configurations and combinations of hardware, many more that we could cover in this time segment.

https://www.guildwars2.com/en/community-news/guild-wars-2-community-meet-and-greet-in-seattle-on-march-1/

The way I see it, all the problems with crafting can be solved by increasing the importance of crafted items. This can be done by either:

1. Increasing the Armor/weapon stats of crafted items for their level.
2. Decreasing the level limit of crafted items.

This would allow crafted items to be superior in stats to other items of the same level.
So, there would be an incentive for players to craft items if they want to possess the best equipment available for their level.

Since crafted equipment is now “Valuable” the prices of crafting mats would go up. Similarly, the crafted equipment prices would also go up.
These change in the economy would also give an incentive to players of all levels to go out in search of crafting mats to sell on the TP or craft equipment for later resale.

IvanHoe.3468:

It is not an equivalent to crossfire/SLI. The IGP will not collaborate with your discrete graphics to help render the game faster.

I take that back. After investigating a little, it seems that this application IS like sli/crossfire. However it works with any graphics card, it is not restricted to specific brands or models.
Here is a fairly recent review: http://www.hardwaresecrets.com/article/Virtu-Universal-MVP-Review/1548/1
Official product description: http://www.lucidlogix.com/product-virtu-mvp.kittenml

Even though there are certainly positive performance gains, GW2 is cpu bound. If your graphics card is being held back by your struggling CPU, you are not going to see much if any performance gain. Secondly, if you have an OCed CPU, by activating the IGP you will generate more heat and consume more power. You might need to revise your overclocking settings, and possibly adjust the voltage to obtain a stable OC.

In any case I’m definitely trying it out

I think you have mistaken the factual functionality of this application.
It is not an equivalent to crossfire/SLI. The IGP will not collaborate with your discrete graphics to help render the game faster. No, this application will simply allow the use of the IGP instead of the PEG, for regular desktop usage. It will also allow the for hardware accelerated video encoding. It will switch back to PEG , when the application detects a demanding 3D environment, such as a video game.

The only benefit of using this would be the reduced power usage, heat and possibly noise production, as well extended product life as the PEG card is powered off until it’s needed.

Jazhara Knightmage.4389:

I doubt anybodys working seriously on fixing the problem, because anet management probably dont see that the perf is costing them money…

They don’t see it because it’s not.
The majority of users attribute the perf issue to network Lag. The rest think the game is so advanced, that it will necessary run slow on todays hardware and a very few actually come here to complain, yet they still play the game.
From a managerial point of view, ArenaNet is actually doing the right thing. Remain silent and wait for the problem to go away on its own.

ZaxanRazor.6235:

Any game that has to deal with rendering interactive objects and large numbers of players will be CPU bound, in the end.

That won’t change unless there’s a shift in GPU technology.

That is blatantly false. Provided decently optimized shaders and not counting for postprocessing effects, the only thing limiting the rendering speed of a scene is the vertex and polygon count. The CPU interaction here is minimal, little more than an operator transferring calls to the GPU.

Depending on the middleware they used on the game, the CPU may be more taxed. By having to process physics, animation, collision and reflections the CPU can quickly become overburdened when too many calculations are required at once. This can be remedied by using adequate middleware (Nvidia’s physicsX instead of havok for example) that offloads these tasks to the GPU, and/or by using good occlusion techniques.
Of course with today’s quad core CPUs being so common, they can set one or two cores to process these tasks and have the rest free to transfer API calls.

The CPU should never be a limitation, if it is there is a severe design blunder.

Speaking of which, I just bought the game and came here to inquire about the game installer setting my CPU and GPU to max performance power state, causing my temps to go up and the fans to become noisy. This is ridiculous considering the download is going to take at least 4-5 hours.
I any case, I had never heard of any performance problem until now and the Youtube videos I’ve seen appeared very smooth and fluid but If the game has performance issues that is a definite showstopper for me.