A Macrotask-level Unlimited Speculative Execution on Multiprocessors
(19/19)
The following description is based on the paper presented at
1995 ACM International Conference on Supercomputing head
in Spain on July 3rd to 7th,1995.
If you want to have the copy of the paper, please
let me know.
In conclusion, I'd like to indicate the followings:
Firstly, We propose the Unlimited speculative execution
scheme to generalize the speculation on multiprocessors and to achieve
high performance.
Why do we call unlimited ? Because the speculation depth is automatically
decided at runtime depending on the overhead of macrotask control.
Secondly, the proposed scheme is divided into two new schemes.
They have advantages in comparison with the conventional schemes as shown here.
The future work is to construct the HPF compiler including the proposed
speculation.
Thank you for your reading!
If you have any questions, I'll be happy to
answer them.
Hayato YAMANA
/ Questions and comments go to
yamana@etl.go.jp
You can write a mail in English or Japanese.