I think the task of specifying whether a PC is going to be able to run an application properly is almost impossible. The best you can do is offer some kind of demo and see whether its acceptably fast or not.
The CPU power is just one of many factors in a system capability. You can have an extremely good CPU and a duff graphics card or even a slow bus to the ram or any number of other issues which will hit your performance.
You can run any app with any number of threads on a single core and it will time slice them and perform them all in some serialized sequence. I would agree that it might be helpful to state whether having extra cores would help out but that alone is not enough.
Sensible games programmers should write their programs which time how long certain processes are taking and adjust the processing to fit. Often they leave this up to the user to specify detail factor but the tree depth for AI search might be pruned lower if your CPU cant complete it within a certain time or the view depth for scenes might be adjusted or the model poly count might be lowered. Thats why benchmarks are not good indicators of games performance because the game might be doing different amounts of work depending on the system capabilities.
The best you can do is to run a demo or find a benchmark for that game that uses the same system hardware that you have. You can get a good idea but you will never be sure. It is the downside to all this lovely flexibility and upgradability that you wont be able to know how your essentially custom rig will handle any application - and its the bane of PC developers that must ensure that apps must work work on a variety of hardware configurations of which the vast majority you will never be able to test against.
The CPU power is just one of many factors in a system capability. You can have an extremely good CPU and a duff graphics card or even a slow bus to the ram or any number of other issues which will hit your performance.
You can run any app with any number of threads on a single core and it will time slice them and perform them all in some serialized sequence. I would agree that it might be helpful to state whether having extra cores would help out but that alone is not enough.
Sensible games programmers should write their programs which time how long certain processes are taking and adjust the processing to fit. Often they leave this up to the user to specify detail factor but the tree depth for AI search might be pruned lower if your CPU cant complete it within a certain time or the view depth for scenes might be adjusted or the model poly count might be lowered. Thats why benchmarks are not good indicators of games performance because the game might be doing different amounts of work depending on the system capabilities.
The best you can do is to run a demo or find a benchmark for that game that uses the same system hardware that you have. You can get a good idea but you will never be sure. It is the downside to all this lovely flexibility and upgradability that you wont be able to know how your essentially custom rig will handle any application - and its the bane of PC developers that must ensure that apps must work work on a variety of hardware configurations of which the vast majority you will never be able to test against.
