Changes

46 bytes added ,  17:15, 5 August 2008
m
no edit summary
Line 1: Line 1: −
[[Image:Wii_hw_diagram.png]]
+
[[Image:Wii_hw_diagram.png|thumb|right|Wii hardware diagram]]
 +
==Wii specs==
 
===CPU===
 
===CPU===
 
* IBM '[http://en.wikipedia.org/wiki/Broadway_%28microprocessor%29 Broadway]' 90 nm based on IBM's Power architecture.
 
* IBM '[http://en.wikipedia.org/wiki/Broadway_%28microprocessor%29 Broadway]' 90 nm based on IBM's Power architecture.
Line 13: Line 14:  
*One floating point unit (FPU) (supports single precision (32-bit) and double precision (64-bit))
 
*One floating point unit (FPU) (supports single precision (32-bit) and double precision (64-bit))
 
*The FPU supports paired single floating point (FP/PS)
 
*The FPU supports paired single floating point (FP/PS)
*The FPU supports paired single multiply add (ps_madd). Most FP/PS instructions can be issued in  
+
*The FPU supports paired single multiply add (ps_madd). Most FP/PS instructions can be issued in each cycle and completed in three cycles.
each cycle and completed in three cycles.
   
*Fixed-point to floating-point conversion can be performed at the same time as FPU register load and store, with no loss in performance.
 
*Fixed-point to floating-point conversion can be performed at the same time as FPU register load and store, with no loss in performance.
 
*The branch unit supports static branch prediction and dynamic branch prediction.
 
*The branch unit supports static branch prediction and dynamic branch prediction.
2,375

edits