From 9e0a2ce822de3006e8fec14c7d0991708d9b78b4 Mon Sep 17 00:00:00 2001
From: "petros.anastasiadis"
Date: Wed, 20 Sep 2017 16:17:22 +0300
Subject: [PATCH] Initial commit
---
External_Functions/README.md | 1 +
GPUs/README.md | 1 +
MPI/README.md | 4 ++++
OpenMP/README.md | 1 +
README.md | 8 ++------
Serial/README.md | 2 +-
6 files changed, 10 insertions(+), 7 deletions(-)
diff --git a/External_Functions/README.md b/External_Functions/README.md
index db1f30f..c15d12e 100644
--- a/External_Functions/README.md
+++ b/External_Functions/README.md
@@ -2,6 +2,7 @@
Basic helpfull functions used by some of the programs.
Function explanation and usage included in corresponding header(.h) files.
+
Modify at your own risk!
->input.c
diff --git a/GPUs/README.md b/GPUs/README.md
index 86661f1..17d67f9 100644
--- a/GPUs/README.md
+++ b/GPUs/README.md
@@ -1,6 +1,7 @@
/****************************************************/
A single GPU impementation of the Matrix-Vector algorithm with:
+
->cuBLAS(BLAS routines implemented on the GPU by NVIDIA)
07/09/2017: Completed
13/09/2017: Modified to use unified memory
diff --git a/MPI/README.md b/MPI/README.md
index 47d9bb9..c1b686d 100644
--- a/MPI/README.md
+++ b/MPI/README.md
@@ -1,7 +1,11 @@
/****************************************************/
A parallel MPI implementation of the matrix-vector multiplication algorithm
+
->MPI(Basic implementation using intel mpi for compilation)
19/09/2017: Completed
+->MPI-OpenMP(Hybrid implementation with MPI for data management between nodes and OpenMP for computations)
+20/09/2017: Completed
+
/****************************************************/
diff --git a/OpenMP/README.md b/OpenMP/README.md
index 01078da..516df02 100644
--- a/OpenMP/README.md
+++ b/OpenMP/README.md
@@ -1,6 +1,7 @@
/****************************************************/
Two OpenMP impementations of the Matrix-Vector algorithm:
+
->OpenMP (a simple naive parallel for implementation)
07/09/2017: Completed
diff --git a/README.md b/README.md
index 366249e..3ba48d8 100644
--- a/README.md
+++ b/README.md
@@ -7,17 +7,13 @@ Directory layout(tree -d):
Training
├── External_Functions
├── GPUs
-├── Multinode
-│ ├── Hybrid
-│ │ ├── Cuda-MPI
-│ │ ├── OpenMP-MPI
-│ │ └── PHI-MPI
-│ └── MPI
+├── MPI
├── OpenMP
├── Outputs
│ └── Debug
└── Serial
+
Loader contains required modules for all makefiles. Create executables with 'make' in corresponding directories.
/****************************************************/
diff --git a/Serial/README.md b/Serial/README.md
index 4e97e41..18907b1 100644
--- a/Serial/README.md
+++ b/Serial/README.md
@@ -1,8 +1,8 @@
/****************************************************/
A serial impementation of the Matrix-Vector algorithm:
-->Serial(Mostly used for time comparison and error checking)
+->Serial(Mostly used for time comparison and error checking)
05/09/2017: Completed
/****************************************************/
--
GitLab