Automated-Parallelization-of-Source-Code-using-Program-Comprehension/Paper_2_Overview.txt at master · Spielerr/Automated-Parallelization-of-Source-Code-using-Program-Comprehension · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Paper 2:
Intra Function Parallelism using Program Comprehension

Intro to Paper:
- How Parallelism is useful
- Concept of Program Comprehension
- Martino
- Related Work
- How our approach is different
- Loop based Parallelism (Pluto)
- Using Program Comprehension as an approach to solve Parallelism


Refer Final Report

Program Comprehension

Mappings from sequential to parallel algos
- sort
- max
- min
- mean

add more Mappings

Write our own stubs

Flow charts for Program Comprehension Implementation
Explain:
- Reducing program to AST
- Path Based Representation of data
- Code to Vec Model to generate vector embeddings (Similar to Word to Vec)
- K Means Clustering(Modified)
- Training phase and prediction phase
- Threshold concepts
- Other Category (Maintaining semantics and correctness of the original program)
- Mapping program to label
- Dynamic Verification
- Replacing code with parallel equivalent

Results:
- Graphs
- Showing correctness is maintained
- Improvements with hardware (Better utilization of hardware)