of synchronous applications severely (by up to 60%), contradicting previous believes that garbage collections should be done as independently as possible. This is because an independent local collection makes the collecting processor unresponsive," making processors waiting for a reply from the collecting processor idle. For asynchronous applications with plenty of intra-node parallelism, independent collections perform better than synchronous collections, but the difference is small at least in our experiments. A more advanced strategy which adaptively selects a right strategy is also implemented and shown to be effective, though it is not significantly better than a simpler always-synchronous" approach in the current experimental conditions.

On top of these runtime systems, a new programming language ABCL/f is designed and implemented. Several non-trivial applications written by the author and others are used for experiments. Both sequential performance and speedup of the applications are reported.

2

Acknowledgement

First of all, I express my biggest gratitude to my supervisor, Professor Akinori Yonezawa, for leading me to this very exciting area in computer science|programming language design, implementation, and parallel programming. Your encouragement was the greatest source of motivation to do my best, when I didn't know how to work as a researcher. It was the greatest event of me to become a research associate of your laboratory and enjoy extra time working in this amazing group.

Since I became a member of this group, Professor Satoshi Matsuoka has been the best advisor of my work. He has a broad range of background in this field and has been improving the quality of my job over five years.

One of the most fortunate things for a researcher is to be involved in a group with a lot of high ability persons. Dr. Masahiro Yasugi has been the greatest source of my inspiration ever since I begin to work. I learned from you how implementers should work and how to make systems faster. Professor Naoki Kobayashi, formerly a research associate of the Yonezawa Laboratory, has always been stimulating me with his keen intelligence. Luckily, after you take the current position, I still have joint meetings with your laboratory, to be timely informed of your activities. Since I begin to learn computer science, Hidehiko Masuhara has been the best friend to discuss with. You have been patiently listening to my incomplete questions or vague ideas and inspiring me with comments hitting the point.

As a research associate, there is nothing more exciting than working and discussing with motivated and hard-working students. I especially thank to Toshio Endo, Yoshihiro Oyama, and Takashi Ninomiya, for their good work and patience. You make my research life here far more fun and exciting than it would be without you. It is really difficult to motivate myself without working with other motivated people! I also thank Atsushi Igarashi,