566
Views
1
CrossRef citations to date
0
Altmetric
COMPUTER SCIENCE

Determination of the optimal shape of matrix elements partitioning on three abstract heterogeneous processors

ORCID Icon, ORCID Icon, ORCID Icon & ORCID Icon | (Reviewing editor)
Article: 1769948 | Received 28 Jul 2019, Accepted 01 May 2020, Published online: 25 May 2020

Figures & data

Figure 1. Candidate shapes defined as potentially optimal for three-processor systems.

Figure 1. Candidate shapes defined as potentially optimal for three-processor systems.

Figure 2. The serial communication with barrier algorithm for three-processor systems.

Figure 2. The serial communication with barrier algorithm for three-processor systems.

Figure 3. The parallel communication with barrier algorithm for three-processor systems.

Figure 3. The parallel communication with barrier algorithm for three-processor systems.

Figure 4. The serial communication with overlap algorithm for three-processor systems.

Figure 4. The serial communication with overlap algorithm for three-processor systems.

Figure 5. The parallel communication with overlap algorithm for three-processor systems.

Figure 5. The parallel communication with overlap algorithm for three-processor systems.

Figure 6. Сlass diagram of “parallel matrix computation on heterogeneous processors” web-application.

Figure 6. Сlass diagram of “parallel matrix computation on heterogeneous processors” web-application.

Figure 7. Example of Program Computations for SCB algorithm with different input values of Pr, Rr, β1, β2,β3.

Figure 7. Example of Program Computations for SCB algorithm with different input values of Pr, Rr, β1, β2,β3.

Figure 8. Example of program computations for SCB and PCB algorithms with different input values of Pr, Rr, β1, β2,β3.

Figure 8. Example of program computations for SCB and PCB algorithms with different input values of Pr, Rr, β1, β2,β3.