Hi,
i made this thread after testing different builds and observing that any kind of build with SSE3 wont work, and builds with SSE2 will run slower, on my AMD system but will work on my INTEL one (being the intel one "smaller" it run dolphin faster, on fps some times its +10, both with same configuration or a different one).
Wondering why this happen (taking into account that my CPU support SSE3 and all) i have researched around a for a while and keep finding different whinings, until i started to find actual information, this is the best article so far i have found:
Intel's "cripple AMD" function
The author explain why this happen, since when it happens and how there is no official solution yet.
Fortunately there is an unofficial fix for the compiler that works, you can access from the article itself (lecture recommended) or you can go trough this link:
Software optimization resources
In the first chapter Optimizing software in C++: An optimization guide for Windows, Linux and Mac platforms you need to go to 13.5 CPU dispatching in Intel compiler in there is explained not only how CPU dispatching works but the fix for the intel c++ compiler plus examples for implementing it.
I wont cry to you all, taking into account that you guys made builds extremely fast so that we can use them, but it will be a nice "touch" for us non-intel cpus users if this fix can be applied.
Thx
PD
All that aside, the guides seems to be pretty good and probably useful for you all out there.
---------------------
i didnt know where to make this thread so pls move if necessary
---------------------
i made this thread after testing different builds and observing that any kind of build with SSE3 wont work, and builds with SSE2 will run slower, on my AMD system but will work on my INTEL one (being the intel one "smaller" it run dolphin faster, on fps some times its +10, both with same configuration or a different one).
Wondering why this happen (taking into account that my CPU support SSE3 and all) i have researched around a for a while and keep finding different whinings, until i started to find actual information, this is the best article so far i have found:
Intel's "cripple AMD" function
The author explain why this happen, since when it happens and how there is no official solution yet.
Fortunately there is an unofficial fix for the compiler that works, you can access from the article itself (lecture recommended) or you can go trough this link:
Software optimization resources
In the first chapter Optimizing software in C++: An optimization guide for Windows, Linux and Mac platforms you need to go to 13.5 CPU dispatching in Intel compiler in there is explained not only how CPU dispatching works but the fix for the intel c++ compiler plus examples for implementing it.
I wont cry to you all, taking into account that you guys made builds extremely fast so that we can use them, but it will be a nice "touch" for us non-intel cpus users if this fix can be applied.
Thx
PD
All that aside, the guides seems to be pretty good and probably useful for you all out there.
---------------------
i didnt know where to make this thread so pls move if necessary
---------------------