4.3.3 Parallelizing Loop 2

4.4.3 Parallelizing PackA

4.4.2 Getting around Ahmdahl's Law

4.4.3 Parallelizing PackA old

4.4.2 Ahmdahl's law

4.2.3.3 Hello World 2

4.2.3.2 Hello World 1

4.3.6 Parallelizing Loop 5: smoothing

4.3.6 Parallelizing Loop 5

4.3.3 Parallelizing Loop 2

4.3.4 Smoothing performance

4.3.4 Parallelizing Loop 3

4.3.2 Parallelizing Loop 1

4.3.1 Lots of loops to parallelize

4.3.1 Single thread performance

4.1.1.3 Performance of a multithreaded reference implementation

4.2.3.4 Hello World 3

4.2.3.1 Hello World

4.1.1.2 Multithreaded JPI loop ordering graphs

4.1.1.1 Multithreading IJP loop graphs

4.1.1.1 Multithreading the IJP loop ordering

3.2.5 Five loops around the micro-kernel Take 2

3.3.4 Pack A Take 2

3.3.3 Packing B

1.3.5.1 Matrix-vector multiplication via axpys