Operation: Make Faster Game

Well, this might mean the end of fun for the phoenix, I suppose. We'll see.

I look forward to the results. Nice post.

Great to see a developer finally break through the corporate PR.

Hopefully we'll see the benefits of these focused optimisation efforts before the end of October.

I fear any longer and majority of the player base will have left for cleaner pastures like battlefield 4.

Still keep the servers running, they'll come back to check in a years time.

sorry for the bold!

but Ryan said that they are making thread / parallelization optimizations more apparent on AMD processors.

Does this mean that a multi-core AMD processor will be at an advantage to an Intel core?

That's pretty interesting with the PS4/XB1 launch and both using AMD CORES.

Can you please answer this, as it will impact the next system I build.

If it means that a system with NVidia GFX (for GPU Physics benefit) with AMD core will be best, then so be it.

Can you shed light on that?

Thank you! Right now, I'm usually running 20fps, I can't participate in any battle with more than 40 people, since my fps drops to 1, and that's on the lowest possible settings despite my comp supposedly meeting specs. This game is amazing, and I look forward to finally joining the larger battles. Hooah!

If you search up some benchmarks for well threaded games (Crysis 3 and Metro Last Light especially) the fx-8350 (an 8 core processor) beats the intel i5 3570k (and probably the i5 4670k) by a small margin. If you overclock the i5 it will beat the fx 8350, but were talking differences of 10 frames at the most. Both processors will not significantly bottleneck a powerful graphics card. The i5 will do better in cpu bound, but not well multithreaded games such as Skyrim, but either will give you 60+ frames a second in most games given a powerful enough graphics card.

Right I understand, but my question was more about specific thread optimizations done for AMD, since AMD is on the other gaming platforms. That's pretty straightforward to achieve, especially since they are pushing more to CPU in these optimizations to cater to a larger base of players with average gfx cards.

A good direction to start with. The game's performance suffered badly from tight deadlines and never-ending demands from design, user retention, purchase optimization. If you didn't start fixing it now, it would be too late.


I always thought a complete rework of the game's engine would be a good idea, but from this post I can see that there are lots of "peripherals" need fix. It could be the case that once theses are fixed, the engine core wouldn't need major refactor until too much more content is added.

As always SOE, too little too late. There's hardly anyone left playing the game to benefit from these optimizations.

The lattice and alert system are just ways to concentrate the remaining players into specific areas, to disguise the fact that hardly anyone is playing.

Just some random questions...

Does this mean things like lighting will be overhauled so we can get darker nights/useful lights again?
Is Hossin on indefinite hold? Resource revamp?
Will INTEL hyperthreaded processors benefit just as much as AMD compute modules?

so, can we haz hossin now?

The basis of this (occlusion) is "don't draw things that the camera can't see". The camera is everything you see on your screen, whether that's your player view or the TV cam you get when guiding the missile. The game will still keep track of that tank behind a rock that you're hitting. It just won't draw the model until the camera is in a position where it could see the model.

Hossin has tons of rocks, trees, plants, water, as well as new building assets. I think this OMFG update is required before large-scale battles can be fought there. I'd expect Hossin to follow shortly if these optimizations go well.

Good work SOE. You got your priorities straight!

Hey codeforge, while you're answering all kinds of technical bollocks. I've always wondered what it is that gives some games, Planetside 2 included, that texture fuzz issue and if there's any possibility of it being remedied in the future.

I assume it's some texture filtering method that gives you good bang for your buck because I always see it in games that are trying to render a huge battles, other offenders I've seen recently being War Thunder and Wargame Airland Battle. It's hard to describe but I'm always acutely aware that it's there. It's a kinda muddy, furry, fuzzy feeling to the textures and ambient occlusion that makes everything look slightly out of focus.

I'm not complaining, I understand you have to make sacrifices somewhere, it's just out of curiosity. Wondering what it is that does it and if there's driver settings or anything that might alleviate it.

I'd say its much more difficult than that.

What these games are doing, and why PS2 runs slow atm compared to them in 'normal play' is for a plethora of different reasons.

I'd suggest people wait and see.

I'm hoping to get a solid performance from my Fx-8350 Oc'ed to 4.51 though. If that happens, I could really give a damn what someone who paid more for an i5 is getting as I don't just have this for gamzoring.

Yep, this is absolutely vital to the future of PlanetSide 2--- if only because if it's not running smoothly, ready for Sony's PS4 launch, I imagine the whole game and senior staff would be binned.

This good news, well delivered, imo. Thanks Ryan.

In the meantime, anyone daring enough can gain about 10 fps simply by disabling the HUD in Settings. Hardcore mode.

A lot of what you mentioned here in the OP excites me extremely. As a programmer I appreciate the complexities of refactoring systems for optimisation and am amazed at the amount of changes you've found might be possible to make.

Looking like a good list! Hopefully you guys keep your eyes on the target and get these powered through. Especially the AMD and multi-core supporting ones.

And remember, neat and well commented code supplies an easy to maintain system!

Why Operation: Make Faster Game?
Because someone really wanted the acronym to spell OMFG.

Personally I'm curious as to what constitutes low-priority physics simulations. I'm guessing things like distant debris and smoke behaviour?

My biggest concern in all of this is the fact that every member of the team is looking for ways to make the game faster. While this sounds good, reading between the lines I'm seeing a little bit of 'I always wanted to make that little tweak to that level...' which could sneak through without the usual checks and balances and most importantly data mining because it falls under the umbrella of optimisation.

Case in point: Quartz Ridge. It's probably a framerate nightmare right now due to its layout, but it is also one of the most fun bases to fight at/through and is, IMO, an example of a defensive base done right. Throwing up extra walls will do wonders for occlusion, but will also ruin the defenders' fire zones and turn many base fights into purely knife-fights that favor neither side.

And speaking as someone who has just done an optimisation refactor for the game we're working on... good luck and make sure to send greeting cards to your immediate family!

Oh my god


THANK YOU!