Too long wait between turns

[Syagrius]

I have downloaded the Operation Zitadelle operation from the scenario depot. The operation is nice, however it takes forever between turns. My computer is brand new and I have a good graphic card. When playing similar large operations like the in-game "Blitzkrieg" the game runs perfectly. Any advices or comments on this?

Thanks

[Cuirassier]

Play smaller scenarios.

[Syagrius]

Yes, I know, however the "Blitzkrieg" operation runs fine, so I was thinking there might be a problem with "Zitadelle"

[Sivodsi]

CMX1 is very heavy on processor use, and this is especially so with the huge scenarios. There's a hell of a lot going on under the hood - even though the graphics are nothing compared to what the latest games have.

Also, modern multi-processor chips are not of much use because the game is not optimized to take advantage of them - hey, its an old game. I remember a couple of years ago having to wait for 7 minutes + for a move to load for one large scenario (involving Elephants as I recall), and I know that others have played on with their PBEMs despite having to wait well over 10 minutes.

So, how long do you wait for your turns?

(and Cuirassier, be nice to new posters, otherwise I'll be forced to deal with you on the battlefield!)

[Redwolf]

Can be a brand new piece of junk

[Bannon DC]

Could be a high level of trenches, foxholes, craters, burning tiles, etc. The more the landscape is changed the longer it takes. Could also have a lot more units on map than the other scenario. Or all of the above.

[MeatEtr]

Also worth mentioning is all the high res mods that can take a hit on your fps. Although turn crunching is like 99% CPU driven. But if your moving around the map like a slug then the mods are hurting your performance.

But as far as the overall turn crunching goes, newer PCs are hitting the CMx1 engine roof. In other words, for example, take 2 PCs with the exact same setup, except one has a Duo(2x 2.3GHz) CPU and the other has a Quad(4x 2.8GHz) CPU. You'll get the same CM turn crunching performance out of them.

[John Kettler]

Syagrius,

Welcome aboard! Think of it as an opportunity to cultivate patience. Just wait until you play Tigers in the Valley or Royal Opponent! Then you'll really understand long turn times, especially with forces on the move, firing from clear across the field, lots of things burning, and other interesting factors. The simple truth is you're asking your computer to do an enormous amount of work, requiring many times the number of polygons to be drawn than were needed for FPS games produced in the same time period. The "busier" your battlefield, the more work for your computer, and you're talking about drawing large expanses of ground, the foliage, any cratering, any fires, the smoke, unit movement, firing effect computation, etc. Bottom line? You are working your computer very hard, probably harder, in aggregate, than at any other time in its service life. That's why things take so long.

Regards,

John Kettler

[tankibanki]

The waiting between turns has next to nothing to do with graphics. Between the turns CM does not calculate any graphics. In "Blue bar time" it only determines the action. The graphics are calculted in real time. The main inflencing factor for the waiting time is the number of units on each side. For example: If you double the the number of units on each side the calculations for LOS increases by four instead of doubling, as you might have expected.

Now take a scenario with 100 units on each side and there's no surprise that it'll take a while to calculate the outcome of a turn.

[John Kettler]

tankibanki,

Trust me when I tell you this isn't correct. If you don't, run some tests. If you take a large scenario with a pristine battlefield and issue orders, you'll get some time X. Now, shell parts of it vigorously enough to cause tiles to catch fire, buildings to blow up and vehicles to die, then watch what happens. The time will soar, even with fewer units as a result of the shelling. There's nothing free in cybernetics. If you alter the baseline conditions by wrecking the terrain and besmirching the sky with smoke, you are going to need longer to calculate turns than you would otherwise. Likewise, battles in forested areas, holding size and unit count constant, will require longer turns than will battles in the barren desert, barring, of course, a sandstorm. If you really want time for your coffee, though, play a large scenario, with lots of units, in an area with significant tree cover--in fog. I have, and it only gets worse when combat effects start to be applied to the situation. Blue bar time is when all the firing is assessed, and that requires first calculating all the lines of sight in every such interaction, a huge computational load which far eclipses mere movement.

Regards,

John Kettler

[Syagrius]

Ok thanks for the welcome J. Kettler and the feedback guys. Now I know that it is quite normal to wait for a while in some scenarios. Its true that there is a lot of trenches and bunkers in Zitadelle.

[John Kettler]

Syagrius,

The more the computer has to keep track of, the longer it's going to take to process turns. This needn't be a bad, thing, though. Just think of it as BFC's way of making sure you stay hydrated, fed and have time for personal hygiene, even in the midst of fierce combat! What other wargame firm is so solicitous of its clients?

Regards,

John Kettler

[Redwolf]

Keep in mind a modern CPU is most likely a CPU with many slow cores, instead of one fast CPU. Since CM is single-threaded you only use one of the slower ones that way.

[SKELLEN]

I have downloaded the Operation Zitadelle operation from the scenario depot. The operation is nice, however it takes forever between turns. My computer is brand new and I have a good graphic card. When playing similar large operations like the in-game "Blitzkrieg" the game runs perfectly. Any advices or comments on this?

Thanks

Wait until you try 'To the Volga' if you are brave enough. Processing times can be 10 to 30 mins or more.:eek:

Admittedly I never intended to complete the above but I was simply curious about the size of the OP.

There are actually two Ops (although being from CMAK) that I can't even run on my computer, but you would think if I can run 'To the Volga' then surely any other OP/Scenario wouldn't pose a problem - :confused:

[tankibanki]

tankibanki,

Trust me when I tell you this isn't correct. If you don't, run some tests. If you take a large scenario with a pristine battlefield and issue orders, you'll get some time X. Now, shell parts of it vigorously enough to cause tiles to catch fire, buildings to blow up and vehicles to die, then watch what happens. The time will soar, even with fewer units as a result of the shelling.

John Kettler

This is simply not true. Look at the two example files I attached. It's exacly the same scenario except that the first one is as flat as a snooker table, the second one has got hills, lots of burning tiles, etc.

The times on my PC:

version 1: AI planning time 3sec, Blue bar time 6sec

version 2: AI planning time 3sec, Blue bar time <2sec

And this is not surprising, because in the flat example, many more lines of sight exist between units, so there have to be many more calculations which unit will attack which one with which weapon (In this case no one attacks because I picked units which can't harm each other at that distance.

Naturally, burning tiles can increase the blue bar time, but only due to their effect on the action in the turn, e.g. calculations whether the adjacent tiles start burning as well, pathfinding, etc ... but the effect is minimal compared to those calculations related to the number of units and most importantly, it doesn't have to do with the graphics of the burning tiles.

timetest-scenarios.zip