| Abstract | .... 2 | |
| 1 | Introduction | .... 2 |
| 2 | KL1 Execution Overview | .... 3 |
| 2.1 | Brief Introduction to KL1 | .... 3 |
| 2.2 | Goal Reduction Feature in KL1 | .... 4 |
| 3 | Execution to Parallel Reduction | .... 8 |
| 3.1 | Consideration of PIM Cluster Architecture | .... 8 |
| 3.2 | Locality | .... 9 |
| 3.3 | Data Structure Allocation | .... 10 |
| 3.4 | Depth-first Scheduling | .... 11 |
| 3.5 | Load Balancing when Using Program Locality | .... 11 |
| 4 | KL1 Parallel Execution Model | .... 12 |
| 4.1 | Overview of this Model | .... 12 |
| 4.2 | ODP Distribution | .... 15 |
| 5 | Evaluation | .... 16 |
| 6 | Conclusion and Further Work | .... 18 |
| Acknowledgment | .... 19 | |
| References | .... 19 |
ICOT研究論文(TR)一覧に戻る / Back to the list of ICOT Technical Reports