danilop

Yeah, true but the new GPU's are around the corner. GK110 is here to stay. Not in Titan necessarily, but in the future versions. Cheaper (and faster) version of the Titan are certain. IMO, it's smart strategy to code with NVIDIA CUBLAS/CUFFT library now, especially when it's much more probable to see dual Titan (or whatever future GPU based on GK110 Nvidia may release) than full blown dual or quad Xeon setup in the gaming computer. In the long run, support for this aspect of the newest GPU's would pay off.

What about huge computation increase (FFT and SGEMM/DGEMM) in GTX Titan? In double precision FFT, GTX Titan is more than 3x faster than 680!!! It would be great if DCS could tap in to this almost supercomputer like processing power. We are CPU limited ATM, so newest GPUs are the only "cheap" option to dramatically increase computation power of gaming systems intended for calculation intensive simulations.

Большое спасибо! :thumbup:

Without exact parameters of the rocket engine you will definitely have hard time simulating thrust fluctuations over time of missile flight. So kudos to ED's dev team! I'm now starting to really appreciate the effort involved. It's relatively "easy" to model basic rocket engine behavior (there are mathematical models in public domain) if you have crucial data. If you don't, it's a trial and error until the model start behaving according to RL observations - it's slow process so we will probably have several versions before missile model settle down.

:thumbup: Couple of questions though. Does Thrust Profile (rocket parameters) mean extrapolation (SFM) based on empirical and reported values, or this data makes basis for the new dynamic modelling (AFM)? And speaking of all important Isp ... Is there any exact and valid data on Specific Impulse values for R77 & AIM120 (or any other current military missile in production)? IIRC, Isp is particularly sensitive to propellant mixture recipe (and they are tweaking it from time to time), so getting current data is a bit of a challenge, to put it optimistically. Cooling arrangement and heat performance of the rocket enclosure is also huge part in rocket efficiency. Any data on this? Is this modelled in DCSW?

Well, I was obviously kidding :), but i'll try to guess: First, it was a long time ago when I was studying rockets, have no idea of ATK's engine propellant mix (and subsequent specific impulse and gazillion other parameters depending on the mix) and I don't know what cooling arrangement is in AMRAAM rocket motor (ablative, curtain, film ...). The other huge unknown is the nature of the AMRAAM motor "failure" which is obviously classified. Soooooo ... I guess that they wanted to sack ATK and find another supplier (they've just did that) ... now, that was scientific and educated guess, wasn't it? ;) :lol:

Maybe they've tried to simulate this: Rocket motor mystery continues to halt AMRAAM It's cold in Caucasus region, you know :D

Ahhh ok, thanks GG. Russian language is not that big an obstacle to me ... I'm native Cyrillic reader/writer so I guess visiting lower part of the forum is the way to gain better understanding of what the Hell is going on with ARHs ... :D

Who is directly responsible for missile development (decisions regarding data on which the missile model will be based) within ED? It would be great to hear straight from the horse's mouth!

Thank you Eddie for taking time to respond. I'm certainly grateful for this wonderful sim and all the effort ED is putting in this title. I've bought every single ED product / title since early Flanker times. I'm very well aware of the fact that we have to live with semi-correct data when talking modern military hardware. We have basically two camps here when we talk missiles: whiners and know-it-all crowd. No matter Red or Blue. My position on the matter is that both whiners and know-it-all are basically wrong when we're talking missiles. Not enough independent data! Having too much thrust is less realistic (basically impossible physically) than having too little. That was the basis of my conclusion (about laws of physics) which stunned you. I'm aware of ongoing tuning of missile performance and I never claimed to be expert on the matter. However, I have enough formal education in physics and math to clearly recognize physically impossible behavior which was evident in 1.2.2. At least, all missiles now seems to behave in (physically) correct and plausible way. I don't claim that they behave to exact and "realistic", current military standards though! Looking forward to 1.2.4 :)

No GG, don't have any problem with that ... it's just the notion that you usually declassify old technology which is not important and valid anymore. ;) It all boils down to what you want to model. AIM120C is easier in that regard than any of the Russian design simply because more (not all! ;)) stuff is publicly available. I just don't understand why did you attack me personally :( ? I think that ARH and AI is behaving more "realistically" (so there you go - all this ongoing debate about missiles is in essence talking about different "realities" - simply there are no hard data to define "realism" outside classified circles) in this patch than in 1.2.2 and if you carefully read my original post, it stresses only that. You don't think that 1.2.3 is improvement over 1.2.2?

I would like to meet person from this forum who has a clue about missile performance apart from internet /declasiffied manuals / university / etc ... Real missile parameters, that is ... P.S. I admit - don't have a clue, beside sources already mentioned! :)

:doh:

After a brief testing it seems that missiles (ARH especially) are finally behaving in accordance with laws of physics. No one on this forum knows the exact parameters of R77 & AIM120c, especially in software and sensor department, so we can only judge the obvious - missile behavior within strict laws of physics / aerodynamics. 1.2.3 is definitive improvement in this regard. Also, AI is much more clever now in BVR and WVR missile engagement scenarios.

+1

Yeah, the updater will probably come later ... It's better to wait for a bit to upgrade anyway - ED is under huge pressure to release something so mistakes are possible - don't want to be guinea pig really :D

I'm no expert, but wonder if it's possible to tap into this vast processing power of Titan? Maybe DCSW (and all other calculation intensive simulations) could benefit greatly from offloading CPU to the newest breed of GPUs? All reviewers are stressing that the main advantage of Titan is in it's enormous calculation performance which is a real breakthrough at this price point (Tesla card is like three times more).

ARH missiles are behaving rather strange ATM to make a definitive recommendation regarding Mig29 strategy - there is a ton of threads discussing this already. ED said that ARH missiles (Red and Blue) will be "significantly" improved in the imminent patch. We'll see ...

I was in your dilemma recently, Friedric. In the end I've got the X65F + Combat pedals. Solid and quality manufacturing throughout, however, there are some minor tweaks that should be done right from the start. Factory preset force levels on X65F are way too low which results in infamous centering problem - once tweaked to sensible and realistic force level, centering problems disappear. After force tweaking, stick and throttle need to be firmly fixed to the desk - then you start to really appreciate stiff throttle response and high force setting on the stick ;). Saitek pedal toe breaks need major axis tweaking in DCSW controller setup to function properly. I'm very pleased with the setup now - flexible, smooth and transparent in use. I even like Saitek programming software and I do all my button programming with it. Also, I have had zero issues with the latest (beta) drivers and software (knock on wood!). And yes, it took just a couple of days to adjust to force sensing technology. I don't think about it anymore ...

Well, they don't give you a new car every time you claim warranty issue, do they? :)

TrackIR is actually supreme value for money if you count in what are you getting - for $130 it's definitely the best accessory you can get for flight simulation software. I'd say that you can master any flight sim with simple, sturdy joystick, keyboard and TrackIR; however, with expensive HOTAS only and no TrackIR you miss completely on the immersion factor and you are severely handicapped in dogfight and busy AG missions. Regarding Natural Point bashing on the net - I don't get it. If they are braking monopoly laws, let the government deal with it! All this patent whining from also runs in head tracking business reminds me of that fine Aesop Fable - "Fox and the Grapes" :)

What about triple touch screens? Is it possible to have three touch screens working in Eyefinity / Surround? Anybody tried setup like that? It seems the best possible solution in combination with TrackIR (you can temporarily disable trackir with predefined key if you can't hold it steady enough)

I don't know if this was proposed before (I've done the search and nothing came out): Give us a Wheel Brake Axis in FC3 aircraft please!! :joystick: It's really shame that we cannot utilize hardware (rudder pedals with toe brakes) to its full capability in FC3.

I urge you to reconsider laptop as a choice for DCSW (or any other hardware demanding game for that matter), if at all possible. Especially in the price range you mentioned. I'm on the laptop (out of necessity, not the choice) ATM so I'm speaking from the bitter experience. So what's wrong with your average laptop for DCSW build then, you may ask? 1) Heat - cooling your overheating, ordinary sized laptop is a major engineering task. Not possible in your business oriented laptop. You need real gaming laptop for proper heat dissipation due to several hours of intensive gaming - only problem is that the decent laptops (meant for gaming) are $2000 and up. 2) Performance vs Price - this is huge. Although laptops are great for business and general computing - getting performance needed for the newest games is extremely expensive. Again, you are looking at $2000+ laptop to even approach $1000 desktop build. Screens are tiny even on the most expensive laptops in comparison to usual desktop monitor, so for decent flight simulator experience you need at least 23" external monitor. Another $400-500 minimum on the $2000 already mentioned ... 3) Upgradability - practically non-existent or extremely expensive. The only advantage i could think off is portability, but if you build your desktop system in increasingly popular LAN boxes (cases meant for easy transportation to LAN parties) you are portable as well. Of course, if you really want the laptop for gaming, go for it - for the price you mentioned, don't expect high quality graphics and the long term thermal stability is definitive gamble (google heat issues and mainstream Dell laptops, for example ;))