Q. Multithreaded Architecture?
It is clear at the moment if we provide a lot of contexts to multiple threads then processors with numerous contexts are known as multithreaded systems. These systems are applied in a manner close to multitasking systems. A multithreaded processor will hang the current context and switch to other. In this manner, the processor would be busy most of the time in addition latency problems would also be optimized. Multithreaded architecture relies on context switching time between threads. The switching time must be very less in comparison to latency time.
The processor efficiency or its utilization can be measured in following manner:
U = P / (P + I + S)
Where
P = useful processing time for which processor is busy
I = Idle time when processor is waiting
S = Context switch time used for changing the active thread on the processor
The aim of any parallel system is to maintain U as high as possible. U will be high if I and S are very low/negligible. The concept of multithreading systems is to decrease I like S isn't increasing. If context switching time is more when weigh against to idle time then the objective of multithreaded systems is lost.