What specifically slows the game down?

Sorry if this has been asked before, I've looked but can't find it. I'm playing FM12 on a laptop, and I like to play the game with not many leagues (often just the top five divisions in England), but a large number of players loaded (usually a custom database with all EU players included, something like 200,000 in all). The game runs quite slowly, and I assume it's the player database size causing that. Does this mean adding more leagues wouldn't have much effect on the speed of the game?

As it happens this laptop has other issues and it's time to get a new one. To play with a large player database, is more memory more important, or a faster processor? I'mnot sure which has the greater effect on the speed of the game. Any pointers would be great, thanks.

Try to delete unneeded scout reports, that can help a bit. But in your case the amount of players seems to be the problem. You shouldn't load any more leagues as there is a bug within the feature that lets the amount of generated players explode. If you want to play faster right away, you might need to use FMRTE to retire unneeded players.

If you have enough RAM (4-6 GB is enough), then the only way you can speed up FM is to cut back on leagues and database in play, or get a more powerful processor. A SSD will help to, making the gradual slowing down of the game less noticable, or even negotiate it completely.

Any idea whether the memory or the processor is more important to this type of player-heavy setup though? I'll get a new laptop rather than tweak this one, it needs replacing for other reasons in any case.

Any idea whether the memory or the processor is more important to this type of player-heavy setup though? I'll get a new laptop rather than tweak this one, it needs replacing for other reasons in any case.

There is a benchmark thread here on the forums which I can't find. It concluded (more or less) with that the most important thing for performance is the speed of processing. The more mhz the better. The rest didn't matter that much.

There is a benchmark thread here on the forums which I can't find. It concluded (more or less) with that the most important thing for performance is the speed of processing. The more mhz the better. The rest didn't matter that much.

The first part of your answer is true - the speed of processing is the most important. But then you go on to say: The more MHz the better ... which is not necessarily true. How fast a processor is at processing is not only about core speeds. I think you will find that the most modern processors from AMD and Intel will not perform equally good, even if they have the same core speeds - for example. Also, we now have 2-core and 4-core (and 6-core) processors ... a 2-core with higher core speeds than a 4-core will not necessarily perform better, not even with software that is not particularly optimised for 4 or more cores. There's also other qualities about the architecture of a processor that means that core speeds is not alone in determining the real acheived processing speeds.

In general you can say that; each step up in the current line of processors (from one of the two manufacturers), will give you a stronger processor.

So, I hear you saying ... is the slowest Core i7 quad core from Intel always better than their strongest Core i5 dual core? In many cases - no. In many cases the strongest Core i5 will be at least as good as that i7. But also, in many cases the i7 actually will be the better choice, in terms of real acheived processing power. I'll leave the question of value-for-money to another debate, I'm not talking about that.

My point is: The statement "more MHz is better" - is not necessarily true.

If you have enough RAM (4-6 GB is enough)

4-6 GB of RAM is pointless, since SI don't let you use more than 3.

4-6 GB of RAM is pointless, since SI don't let you use more than 3.

But you will want to let the rest of what's going on in your computer, the running of your OS (Operating System, meaning your Windows, MacOS, Linux etc.) and so on, also have some RAM available to it, don't you? You may even want to do other things simultaneously while playing FM? All of these other things (and there's usually a lot going on in the background) needs some RAM to run properly. When I start up my computer, the OS and all the bacground processes and running services alone takes up about 1.2 to 1.3 GB of RAM. With for example 2 GB of RAM fitted, you will then have 0.7 to 0.8 GB RAM available to run FM. With 3 GB RAM you will have ... and so on. What happens then is that the OS will start some whealin' and dealin', trying to allocate enough RAM for everything. This slows everything down. In addition, the OS will have to make a substitute because of the lack of RAM, and put some of what the RAM should be holding, on to the page file (or swap file as it's also known as). That means it will use a designated area of your harddisk as a substitute for RAM. This is a lot slower.

Also, I'm quite sure that FM will go 64 bit in the near future, and will then allow the software to have a lot more RAM available for use.

The first part of your answer is true - the speed of processing is the most important. But then you go on to say: The more MHz the better ... which is not necessarily true. How fast a processor is at processing is not only about core speeds. I think you will find that the most modern processors from AMD and Intel will not perform equally good, even if they have the same core speeds - for example. Also, we now have 2-core and 4-core (and 6-core) processors ... a 2-core with higher core speeds than a 4-core will not necessarily perform better, not even with software that is not particularly optimised for 4 or more cores. There's also other qualities about the architecture of a processor that means that core speeds is not alone in determining the real acheived processing speeds.

In general you can say that; each step up in the current line of processors (from one of the two manufacturers), will give you a stronger processor.

So, I hear you saying ... is the slowest Core i7 quad core from Intel always better than their strongest Core i5 dual core? In many cases - no. In many cases the strongest Core i5 will be at least as good as that i7. But also, in many cases the i7 actually will be the better choice, in terms of real acheived processing power. I'll leave the question of value-for-money to another debate, I'm not talking about that.

My point is: The statement "more MHz is better" - is not necessarily true.

It is (more or less) for FM, because the benchmark thread showed that the amount of cores meant little. My 2.6ghz dual core processed the same savegame just as fast as a 2.6ghz quad core. Of course, the test wasn't scientific or anything, but overall I think that it showed that memory and the amount of cores meant less than "desired".

The first part of your answer is true - the speed of processing is the most important. But then you go on to say: The more MHz the better ... which is not necessarily true. How fast a processor is at processing is not only about core speeds. I think you will find that the most modern processors from AMD and Intel will not perform equally good, even if they have the same core speeds - for example. Also, we now have 2-core and 4-core (and 6-core) processors ... a 2-core with higher core speeds than a 4-core will not necessarily perform better, not even with software that is not particularly optimised for 4 or more cores. There's also other qualities about the architecture of a processor that means that core speeds is not alone in determining the real acheived processing speeds.

In general you can say that; each step up in the current line of processors (from one of the two manufacturers), will give you a stronger processor.

So, I hear you saying ... is the slowest Core i7 quad core from Intel always better than their strongest Core i5 dual core? In many cases - no. In many cases the strongest Core i5 will be at least as good as that i7. But also, in many cases the i7 actually will be the better choice, in terms of real acheived processing power. I'll leave the question of value-for-money to another debate, I'm not talking about that.

My point is: The statement "more MHz is better" - is not necessarily true.

Don't forget to consider the CPU cache aswell. It has nothing to do with the CPU clock speed, but it can drastically improve performance.

This has been excellent help, thanks everyone. I think I've found the benchmarking thread talked about, here.

It is (more or less) for FM, because the benchmark thread showed that the amount of cores meant little. My 2.6ghz dual core processed the same savegame just as fast as a 2.6ghz quad core. Of course, the test wasn't scientific or anything, but overall I think that it showed that memory and the amount of cores meant less than "desired".

That test was likely not done with the full detail match sim enabled for all leagues in the save. The full detail match sim grabs every core and thread it can. Enabling it makes a big difference to how FM performs depending on how many competitions you set it up. The competitions you compete in yourself are always simulated in full by default, anything else is up to you to decide.

That test was likely not done with the full detail match sim enabled for all leagues in the save. The full detail match sim grabs every core and thread it can. Enabling it makes a big difference to how FM performs depending on how many competitions you set it up. The competitions you compete in yourself are always simulated in full by default, anything else is up to you to decide.

Yes that is correct. The test was a "normal" all-leagues-loaded game without extra detail level added. Someone tested with full match detail and then cores and memory made more of a difference.

All this does seem to be about the number of leagues loaded though, plus the level of detail, graphics and so on. Is it likely that the number of players loaded (database size) is heavier on memory, or is it all the same and therefore (very generally speaking) processor is more important?

I want to load as many players as I can, but I very rarely play a league that's outside England and Scotland so don't need to load more leagues than those.

That is of course perfectly allright Cid. But the reason some of us like to have as many leagues in play as possible, is in order to get as realistic a "football world" as possible - for example the question of which teams manage to qualify for the CL, from each European nation - and not necessarily in order to be able to play in all these leagues. But compromises have to be done of course, if your computer isn't absolutely up for it - or if you are rather impatient about the speed FM will run at with many leagues loaded

As for your question: Large database will also tax the processor heavily, because all these players have to get their rise/decline in stats computed, allmost continously. Again; enough RAM is enough, your processor is the key. In fact, I don't know of any game that is so heavily reliant on processor power, as FM is. Many are, of course, but those usually are very dependent on graphics performance as well. Im FM, it's all about the processor, the 3D graphics can easily be handled by rather modest graphics chips or cards.

200,000 players is a hell of a lot of players, especially for a game with not many leagues active.

i tend to get into trouble on my little i5 cpu lappy with anything more than 250,000 players after about 10 years

Hello everyone. I want to share my experiment of having a lot of players in the database.

At the beginning i started a game with 21 leagues loaded from 7 countries. I chose to load every african players. The game started with 50000 players on.

After a while i decided to add more leagues to my game. I added Australia, Korea, China, Mexico, Brazil, Argentina and some more but i never had more than 10 countries loaded at the same time. I realized then that the game was slower and slower. After playing in Asia, Africa and America i removed almost everything and kept only 5 leagues in 5 countries. And the game was still very very slow. I launched Genie Scout and i saw that 350000 players were in the game. Like you guys i think that this impressive amount of players is making my game very slow. It takes me 1 hour to simulate one week in the game.

I saw on this topic that someone is suggesting to retire every useless player via FMRTE. Is that a good way to solve the problem ? I'm a little affraid that using FMRTE could generate some bugs in my game.

Hello everyone. I want to share my experiment of having a lot of players in the database.

At the beginning i started a game with 21 leagues loaded from 7 countries. I chose to load every african players. The game started with 50000 players on.

After a while i decided to add more leagues to my game. I added Australia, Korea, China, Mexico, Brazil, Argentina and some more but i never had more than 10 countries loaded at the same time. I realized then that the game was slower and slower. After playing in Asia, Africa and America i removed almost everything and kept only 5 leagues in 5 countries. And the game was still very very slow. I launched Genie Scout and i saw that 350000 players were in the game. Like you guys i think that this impressive amount of players is making my game very slow. It takes me 1 hour to simulate one week in the game.

I saw on this topic that someone is suggesting to retire every useless player via FMRTE. Is that a good way to solve the problem ? I'm a little affraid that using FMRTE could generate some bugs in my game.

I removed several leagues (like 5-6) and after that the game seems to have sped up a little.