Believe it or not running on a single core is often faster then spreading the work out across multiple cores. Almost every single application out there is multi-threaded, you can check this in your task manager by telling it to display the number of threads in each process.
Try playing Supreme Commander on a single core. Then play on a dual core and watch your performance skyrocket. Then play on a quad core and watch your performance skyrocket some more. Then use Core Maximizer to better balance the threads amongst multiple cores and watch your performance go up some more.
And just because the application is multi threaded does not make it multi core aware.
There is no such things as “multi-processor aware”, there is only single threaded processes and multi-threaded processes. The OS handles all thread-core assignments, the only thing an application can do is ask the OS to put all the threads on a single core.
Only non-blocked separate threads can take advantage of separate processors and the . The heavy lifting of every current generation graphics engine, which is the major bottleneck in most games is still performed by a single thread or multiple-blocked threads. It is not possible to retrofit these engines to change that, it needs to be designed in from the ground up.
Actually the O/S isn't the only way to handle thread-core assignments.
Comments
Try playing Supreme Commander on a single core. Then play on a dual core and watch your performance skyrocket. Then play on a quad core and watch your performance skyrocket some more. Then use Core Maximizer to better balance the threads amongst multiple cores and watch your performance go up some more.
And just because the application is multi threaded does not make it multi core aware.
There is no such things as “multi-processor aware”, there is only single threaded processes and multi-threaded processes. The OS handles all thread-core assignments, the only thing an application can do is ask the OS to put all the threads on a single core.
Only non-blocked separate threads can take advantage of separate processors and the . The heavy lifting of every current generation graphics engine, which is the major bottleneck in most games is still performed by a single thread or multiple-blocked threads. It is not possible to retrofit these engines to change that, it needs to be designed in from the ground up.
Actually the O/S isn't the only way to handle thread-core assignments.
http://www.researchchannel.org/prog/displayevent.aspx?rID=19812&fID=569