64bit and multithreading

the topic does say it. for a game released Q4 2006 it has a lack of present and comon technology. may we get some eta when we can expect delivery of those missing must-have features?!

Simon

This goes doubly for the server files. Most dedicated servers offered on the market now are multi-core.

However there is already a post about this in the Multiplayer forum.

Thanks for answering Suma.

I know that you guys have a lot on your plate and are working around the clock to solve other pressing issues. I just hope this isn’t discounted completely because as technology progresses, more and more systems will have multi-threading. Especially on the server.

I think a good start would be to release a server file which can best utilize it because i am hard pressed to find a premium server that does not have multiple cores.

Again, thanks a bunch and keep up the great work.

I wouldnt expect it for a lonnggg while...

Maybe if they release an expansion like Resistance for OFP they will implement it.. But by the sounds of it its a lot of work and so there are much bigger priorities to sort out first (bugs).

And if not... theres always game 2

Bet ya everyone will be waiting for higher level programming language to take care of the thread issue before any games will start taking advantage of multicore.

Whatever you do, don't go buy a quadcore right away

Quad cores have theire place. we currently have 3 retail servers running and a demo server running on a quadcore dedicated box.

The issue comes when we try to have 60 people at once which more than maxes out one core so we gotta run distributed which is not much better.

There is a completely other topic on it in the Multiplayer forum so I dont want to dwell on that here.

I think this thread focuses mroe on client side.

Yeah the issue is moreover for servers, which are rapidly becoming multi-cored to improve performance... Support for multi core in ArmA (at least for the server side) would increase performance massively on MP. (and also allowing the "100+ players" to become reality )

I would suspect, that until compilers can automatically and intelligently compile for dual-core processors that it won't happen on any meaningful scale for games.

Agreed, and as multi threading is not very widespread yet the actual value of multi threading versus workload is not completely tested yet.

Although i'd like to see it in ArmA (even if purely for server side purposes) it is quite doubtful unless, as you say, the process becomes simpler.

Although should we have high hopes for Game2? It is 'next generation'

Still, even without any specific optimization on our side, thanks to the optimization done by both nVidia and ATI in the drivers, you can see a measurable performance gain in ArmA (my measurements have shown 5-20 % improvement depending on the scene).

Got me drueling

I dont see anyway of running 100+ players on a map and keeping decent FPS without major re-working of the Dedicated server.exe. The server must be multithreaded, or we will be going down the same path as I did in OFP. Trying to get a server to support that many clients will not be sustainable for CTI matches, which is the gameplay type that the majority of the community want.

Quote[/b] ]Multithreading: Yes, ArmA is missing specific optimization for Dual Core systems. The same can be told about many (perhaps most) games released (including Microsoft flagship product Flight Simulator X). While this is unfortunate, there are reasons for this - implementing multithreaded architecture is very complicated and time consuming process. Still, even without any specific optimization on our side, thanks to the optimization done by both nVidia and ATI in the drivers, you can see a measurable performance gain in ArmA (my measurements have shown 5-20 % improvement depending on the scene).

Well, although i think MSFlightX is not a good example, you are right that not every game or simulation has multithreading and 64bit support. But the point is that we dont talk about MS and its Flight Simulator. We are talking about ArmA. Just for the files: Half-Life (yes even HL1) and HL2, Battlefield2, Ghost Recon Advanced Warfighter, Rainbow Six Vegas, Americas Army, Gothic 3, ALL Unreal Engine based Games since Unreal1 are multithreaded, Ofcourse ID Software's Quake IV Engine, ...

we could expand this list a lot more.

multithreadin is complex and complicated but that is the comon business of professional developers. im pretty sure that adding multithreading afterwards is rarely possible due the immense amount of time (and money).

just take it as some kind of critics for the responsible head-of of this project. usually when development has begnu multithreading should have stood on the list. even when ofp was recent, there already was known that intel and amd would produce multicore processors for the consumer market. also classic smp systems (multiprocessor boards) were already pretty popular. so i dont think ppl would understand reactions as "oops, how we should know?!".

Quote[/b] ]Regarding 64b, this technology does not bring any magical speedup, what it brings is an increased address space, which would bring almost nothing to ArmA. Thanks to the streaming technology the ArmA is handling huge datasets (over 4.4 GB of data) with little physical and virtual address space needed. Until more than 2 GB is needed to be present in the memory at the same time for one scene (which will quite likely be necessary for future products) , such technology is not needed and no specific support is therefore planned for ArmA.

well 64bit does not only achieve addressing. offering 64bit optimized code also offers native code. running 32bit code on a 64bit machine means that the processor has to "emulate" the opcodes of the program which is not as fast as running it on a native 32bit processor. also moving large amounts of memory is a lot more faster on a recent 64bit machine. wasting that advantage just because there is no 64bit optimization is a little frustrating.

there are good examples for performance increasing just by adding 64bit support. best known example is hl2 of valve. when they decided to add 64bit native support and did release it, their technology demo did run up to 40% faster just because of the optimized code - on a single core 64bit processor. so its not logical to say that such an improvement is impossible to ArmA.

however, Paul Statham himself said that up to the last second of release features will be added and enhanced. the engine of the game and processor support (64bit) or even mt support is a comon feature and not a rare thing.

this is not meant as some kind of offense. we enjoy the game-concept and simulation capabilities. BUT getting an oldfashioned engine in Q4 2006 with no support for 64bit or multicore machines while most of the present games DO support one of those (sometime both) features is a little bit disappointing, especially if you think about those dramaticly raised expectations created by your public relations department. (just reread press releases or interviews)

just my point of view

Simon

... CTI matches, which is the gameplay type that the majority of the community want.

I highly doubt that...

BTW, Suma said they'll do what they can ("We hope we will be able to provide some more support in future patches, but I cannot make any promises on when (or even if) that will be. "), so please stop asking for multithreading support (FYI: I'm dual core user, too).

Lange good post I am with you, I fail to see why MT wasn't high on the list when Arma was developed. I along with many other server admin's have discussed ways of achiving decent FPS for years. I chuckle when I read about Arma server threads and ppl trying all different ways to increase performance. i.e setting affinity. I'm afraid none of this will achive any significant improvment. I can see us travelling down the same road as we did with OFP. MT is the only real way of the community achiving what we have wanted for so many years. Large scale battles involving 100+ players without the constant worry of "damm the servers down to 8FPS its gunna crash"...LOL

Quote[/b] ] highly doubt that...

Thats a shame when a moderator doesn't know what the majority of the community would like to play.

Quote[/b] ] highly doubt that...

Thats a shame when a moderator doesn't know what the majority of the community would like to play.

Please edit your post next time instead of double posting. Thanks

I rather guess that you're overestimating the CTI community. Sure, it's a part of the community, but definitely not the majority. If you feel the need to discuss that further, PM me or open up a new thread, it's too OT here.

Unfortunately CTI and multicore topics are linked. Normally I would agree with you, but if you are advertising 100+ people battles on your website, and it is not even possible, then your base is going to get upset and it needs to be resolved. In order to do that, multithreading (at least on the server), must be made possible.

Unfortunately CTI and multicore topics are linked. Normally I would agree with you, but if you are advertising 100+ people battles on your website, and it is not even possible, then your base is going to get upset and it needs to be resolved. In order to do that, multithreading (at least on the server), must be made possible.

Sure, but there's still no need to discuss how many % of the community want CTI and if its more than 50% here. Also the MT version of ArmA won't be done earlier (if there ever will be one) when you keep complaining here

No, have never advertised anywhere that we offer 100+ player matches..lol My thoughts are mearly the advertising hype made it seem that we could run 100+ plyr huge battles easily, I think also when the English version is released we may see that many ppl on a server, I think our Australian server may well have to cope with the influx of players.

Just read a good interview from john carmack at CES 2007 talks about multithread games and DirectX 10 on pages 2 and 3 which explains why Multithread games are costly to make and might give a better understanding as to why BIS isnt supporting it yet.GameInformer

No, have never advertised anywhere that we offer 100+ player matches..lol My thoughts are mearly the advertising hype made it seem that we could run 100+ plyr huge battles easily, I think also when the English version is released we may see that many ppl on a server, I think our Australian server may well have to cope with the influx of players.

ummm... http://www.armedassault.com/multiplayer.html

No, have never advertised anywhere that we offer 100+ player matches..lol My thoughts are mearly the advertising hype made it seem that we could run 100+ plyr huge battles easily, I think also when the English version is released we may see that many ppl on a server, I think our Australian server may well have to cope with the influx of players.

ummm... http://www.armedassault.com/multiplayer.html

He's talking about his page, as he didn't get that you were talking about BI's page...

Sure, but there's still no need to discuss how many % of the community want CTI and if its more than 50% here. Also the MT version of ArmA won't be done earlier (if there ever will be one) when you keep complaining here

I agree with you, not everyone is using the game for CTI . All I am saying is that it is something that BIS should look at at some point. it is by far not the most important thing that needs to be addressed.

I don't mean to come off as complaining, I just wanted to make sure that it is addressed and not thrown to the wind. When you start throwing around 100+ battles, and when the US and UK wave hits, people are going to get very angry.

He's talking about his page, as he didn't get that you were talking about BI's page...

Oh, looks like there is a miscommunication there :-)