Quad/Triple Core CPUs..Will Empire be optimised?

[hellenes]

IMO this is a VERY important point. The RTW engine was designed and optimised for single gore CPUs...Thus the whole 10000 men technical limit.
With the establishment of dual cores and the rising quad/triple cores this number (apart from graphics where it can be easily scaled too, though there are GPUs that can take huge loads of polygons) can obviously be pushed much further....
Now a question arises: Will ETW be optimised to take advantage of strong CPUs? So people with quad/triple core processors actually have their money's worth? And more realistic and impressive troop count?

[Husar]

I would hope so.
If game developers haven't noticed that trend yet, their games will fall behind technically in not too much time I guess.

[rajpoot]

The 10000 man limit is still there :(
If CA have taken into consideration the new CPUs, they certainly haven't used it in unit sizes.

[Martok]

Indeed, CA has already said that battles will be restricted to the usual 10,000 men. Perhaps they do so in order to not keep the game within reach of only moderately-powerful machines like mine? I don't know really know how that works.

[hellenes]

Indeed, CA has already said that battles will be restricted to the usual 10,000 men. Perhaps they do so in order to not keep the game within reach of only moderately-powerful machines like mine? I don't know really know how that works.

If one can choose unit sizes whats the problem? Why not utilize the technology available? Why stay stuck in 2001?

[Polemists]

The problem I think is the fact that most people do not have quad or triple core pcs. Even amongst the gamer market, your hardly talking mainstream. Empire is aiming for massive sales and while games like Crysis look good, it's games with the graphics equivilent of Civ 4, EU III, and MTW 2 that dominate the pcs of turn based gaming.

The day may come when it's 20,000 but for the time being they are sticking with 10, and that's more then enough for me. I don't need to see 100,000 v 100,000. Even with 10,000 v 10,000 I already have to much action to actually watch or enjoy anything, so I usually go with 2,000 v 2,000.

[Nelson]

If I can dial back the options and still have a nice experience that’s fine. I hate to do it mind you, but if it’s a good game worth playing a year or two down the road after I eventually upgrade I’ll see the thing in all its’ glory.

[hellenes]

The problem I think is the fact that most people do not have quad or triple core pcs. Even amongst the gamer market, your hardly talking mainstream. Empire is aiming for massive sales and while games like Crysis look good, it's games with the graphics equivilent of Civ 4, EU III, and MTW 2 that dominate the pcs of turn based gaming.

The day may come when it's 20,000 but for the time being they are sticking with 10, and that's more then enough for me. I don't need to see 100,000 v 100,000. Even with 10,000 v 10,000 I already have to much action to actually watch or enjoy anything, so I usually go with 2,000 v 2,000.

A triple core CPU costs £64...thats a JOKE....And whats the harm of optimisation? Its even better for people with weak machines... Look at Supreme commander...it utilises all cpus and still can run (on a tiny scale compared to tw) on old machines...

[Alexander the Pretty Good]

According to the latest Steam tech survey, dual core computers made up 36.72% of respondants, while 4.36% have quads. It certainly isn't a deeply scientific survey, but it certainly feels right to me.

I would like the game to have some optimization for multiple-cored CPUs, but it would be icing on the cake.

[hellenes]

According to the latest Steam tech survey, dual core computers made up 36.72% of respondants, while 4.36% have quads. It certainly isn't a deeply scientific survey, but it certainly feels right to me.

I would like the game to have some optimization for multiple-cored CPUs, but it would be icing on the cake.

Virtually all modern sold machines have dual cores and a HUGE chunk has Triple cores....Its the GPU part that always lacks....

[Alexander the Pretty Good]

However, as seen from the survey, the market has plenty of older models without dual cores - and CA has to weigh the cost of extra effort in optimization versus the benefit of extra sales by people who view optmization support as a deal breaker or maker. I imagine the latter isn't a huge market...

[rajpoot]

Nice survey, makes me feel good about my system.

And if it's accurate, the most surprising point I see is that half of the people actually have single core CPUs still.

[Alexander the Pretty Good]

The survey is from Steam users who agreed to respond. I don't know what that says about the PC gaming culture as a whole. I would wager lightly that it might be biased towards those with better than average systems - bragging rights and all that. But I don't know.

[Megas Methuselah]

Based from what I learned in my Statistics class, I believe you're right, Alexander the Pretty Good. It is a biased sample, with a few confounding variables, as well. Single and Dual cores probably are more numerous than the survey may suggest.

[Marius Dynamite]

I don't think more units on the field is necessary for this game. I would imagine armies set out in very long lines. It will make them seem a lot bigger even though they may be the same number as in M2. Consequently the Battlemap size will have to be a lot bigger.

The downside of this is if Sprites are not done very well the games could end up looking awful as I felt M2TW did. Zoomed in the game looked brilliant, zoom out a little and all the white horses turn brown. I can already see myself looking down the line of musketeers and seeing a 3D guy beside a 2D Guy :(

[hellenes]

I don't think more units on the field is necessary for this game. I would imagine armies set out in very long lines. It will make them seem a lot bigger even though they may be the same number as in M2. Consequently the Battlemap size will have to be a lot bigger.

The downside of this is if Sprites are not done very well the games could end up looking awful as I felt M2TW did. Zoomed in the game looked brilliant, zoom out a little and all the white horses turn brown. I can already see myself looking down the line of musketeers and seeing a 3D guy beside a 2D Guy :(

Sadly only few people dont care about graphics....

[JeromeGrasdyke]

You will definitely see good benefits from triple / quad core CPU's, and also from SLI dual-GPU machines.

But as people have mentioned above this is a scaling problem... ETW does need to run as well as possible on the low-spec machines, and that means that the core game model - the part of the code that moves the men, picks animations, does collision, does the rules calculations, ai and so on - has to work within those limits. It's one main reason why we made the decision not to move the 10k man 'goal'.

The other big reason is that we wanted to do more with the men, and actually lay the groundwork for men as more complicated entities in the world. In RTW, the men were -in code terms- pretty dumb and unsophisticated objects. It was very efficient, but getting individual men to do complicated things such as dismounting, or handling interruptions while loading a catapult, was cumbersome and time-consuming for programmers. In ETW, men are more complex, more agent-like and so can display a much greater range of behaviours.

It was a choice we made, spending some more CPU power here to give us something we can work with and extend more easily in the long run, so we can give you not just more soldiers, but better and more realistic soldiers... I'm confident it was the right decision ;)

[Husar]

I think it's more the other way around, those guys who know a lot and care a lot about computers won't partake in the survey because "omg, my privacy, I will feel raped!" while those casual users who have no clue and slug around on their ages old single cores playing Peggle all day will just click yes and not care.

[Polemists]

There CA said it best. If you have triple or quad you get a nice game, and if you don't, you can still play it just fine.

Everyone's happy :P

[Belgolas]

THanks CA.

Anyways I can play with more then 10,000 men on the battlefield with playable frame rate in M2TW but not in RTW. That shows that M2TW is more optimised. So I trust that CA will make E:TW more optimised so maybe I can have more then 10,000 men. Although to get a battle that large is so rare that it pretty much only happens in custom battles.

OH and for lan games please allow AI controled armies!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
M2TW had this for lan battles and it was fun having my brother and I take down a few enemy armies.

[rajpoot]

You will definitely see good benefits from triple / quad core CPU's, and also from SLI dual-GPU machines.

But as people have mentioned above this is a scaling problem... ETW does need to run as well as possible on the low-spec machines, and that means that the core game model - the part of the code that moves the men, picks animations, does collision, does the rules calculations, ai and so on - has to work within those limits. It's one main reason why we made the decision not to move the 10k man 'goal'.

The other big reason is that we wanted to do more with the men, and actually lay the groundwork for men as more complicated entities in the world. In RTW, the men were -in code terms- pretty dumb and unsophisticated objects. It was very efficient, but getting individual men to do complicated things such as dismounting, or handling interruptions while loading a catapult, was cumbersome and time-consuming for programmers. In ETW, men are more complex, more agent-like and so can display a much greater range of behaviours.

It was a choice we made, spending some more CPU power here to give us something we can work with and extend more easily in the long run, so we can give you not just more soldiers, but better and more realistic soldiers... I'm confident it was the right decision ;)

All right, that sounds pretty promising Thank you.

[Martok]

You will definitely see good benefits from triple / quad core CPU's, and also from SLI dual-GPU machines.

But as people have mentioned above this is a scaling problem... ETW does need to run as well as possible on the low-spec machines, and that means that the core game model - the part of the code that moves the men, picks animations, does collision, does the rules calculations, ai and so on - has to work within those limits. It's one main reason why we made the decision not to move the 10k man 'goal'.

The other big reason is that we wanted to do more with the men, and actually lay the groundwork for men as more complicated entities in the world. In RTW, the men were -in code terms- pretty dumb and unsophisticated objects. It was very efficient, but getting individual men to do complicated things such as dismounting, or handling interruptions while loading a catapult, was cumbersome and time-consuming for programmers. In ETW, men are more complex, more agent-like and so can display a much greater range of behaviours.

It was a choice we made, spending some more CPU power here to give us something we can work with and extend more easily in the long run, so we can give you not just more soldiers, but better and more realistic soldiers... I'm confident it was the right decision ;)

Good to know Jerome. Thanks!

I myself have an AMD dual-core (2.2 Ghz), so I'm guessing I probably won't be affected much either way?

[hellenes]

You will definitely see good benefits from triple / quad core CPU's, and also from SLI dual-GPU machines.

But as people have mentioned above this is a scaling problem... ETW does need to run as well as possible on the low-spec machines, and that means that the core game model - the part of the code that moves the men, picks animations, does collision, does the rules calculations, ai and so on - has to work within those limits. It's one main reason why we made the decision not to move the 10k man 'goal'.

The other big reason is that we wanted to do more with the men, and actually lay the groundwork for men as more complicated entities in the world. In RTW, the men were -in code terms- pretty dumb and unsophisticated objects. It was very efficient, but getting individual men to do complicated things such as dismounting, or handling interruptions while loading a catapult, was cumbersome and time-consuming for programmers. In ETW, men are more complex, more agent-like and so can display a much greater range of behaviours.

It was a choice we made, spending some more CPU power here to give us something we can work with and extend more easily in the long run, so we can give you not just more soldiers, but better and more realistic soldiers... I'm confident it was the right decision ;)

Oh I see....
I just have a final question: Is the core engine limited and overflows the CPU no matter how strong it is over 10k men? I mean lets say a 8400C2D can operate 10k men. Now will a €1000 Quad Extreme be able to operate 40k? (Cause roughly thats the power scale) Or will the engine code eat CPU cycles for dinner?

[rajpoot]

I don't think you'll be able to get past the 10k limit even if you have the new experimental CPU Intel people are working on..........having a better, like a quad, (atleast by what I've understood) will give you better performance, not more men.

I myself have an AMD dual-core (2.2 Ghz)

Same here, and I love that thing, when I was buying it, I wasn't so sure, it was the Intel Core2 Duo or AMD Dual Core, and I sure am glad I chose the latter.

[Belgolas]

I don't think you'll be able to get past the 10k limit even if you have the new experimental CPU Intel people are working on..........having a better, like a quad, (atleast by what I've understood) will give you better performance, not more men.



Same here, and I love that thing, when I was buying it, I wasn't so sure, it was the Intel Core2 Duo or AMD Dual Core, and I sure am glad I chose the latter.

First off having play many many battles in M2TW with over 10K men WITH playable frame rate proves that wrong.

Second C2D is WAY faster at gaming then any product AMD has right now. Intel has the best bang for your buck right now. And I sadly don't think AMD will have much to compete against Core i7 either. AMD needs to get into the game again so we can get cheaper CPU's and better CPU's out faster.

[hellenes]

First off having play many many battles in M2TW with over 10K men WITH playable frame rate proves that wrong.

Second C2D is WAY faster at gaming then any product AMD has right now. Intel has the best bang for your buck right now. And I sadly don't think AMD will have much to compete against Core i7 either. AMD needs to get into the game again so we can get cheaper CPU's and better CPU's out faster.

Amd quads cost as much as C2D line...so bang for your buck is still AMD...Intel was/is and always will be horribly overpriced....

[Derfasciti]

Wait... 10,000? My games in MTW2 and whatnot seem to allow battles of at tops 2,000 vs 2,000 and that's with unit size on huge.

[Megas Methuselah]

Reinforcements. Multiple armies per side.

[Hax]

Slight 56k warning:

Spoiler Alert, click show to read: 

No but seriously, it's good to hear this. I run a Q6000 Quadcore myself, and hearing that it won't need the überpc of the future to run properly is good.

[rajpoot]

First off having play many many battles in M2TW with over 10K men WITH playable frame rate proves that wrong.

Keeping the CPU debate aside, I still don't get how you managed over 10000 men. The custom battles have what, 8 armies in all, on huge scale, with full stacks..........I had done some math, I could just reach to 10000 in theory. Perhaps you could elaborate more?