diff --git a/Makefile b/Makefile
index cd6ddd8..269d167 100644
--- a/Makefile
+++ b/Makefile
@@ -1,8 +1,9 @@
-# Makefile for assignment 1, CPSC 521
+# Makefile for assignment 2, MPE version, CPSC 521
 
 CC = mpicc
 CFLAGS = -Wall -Wextra -std=c99
-LDFLAGS = -lm
+#LDFLAGS = -mpe=mpilog -lm -lmpe
+LDFLAGS = -lm -lmpe
 
 TARGET = nbody
 SOURCES = $(wildcard *.c)
diff --git a/nbody.c b/nbody.c
index 067b4f6..fefe072 100644
--- a/nbody.c
+++ b/nbody.c
@@ -1,6 +1,7 @@
 /* This program is written in C99. */
 
 #include <mpi.h>
+#include <mpe.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>  // for memcpy
@@ -60,6 +61,16 @@ void simulation_step(body_t *body_array, int granularity, int self, int bodies
 void send_body_array(body_t *body, int count, int dest, int tag);
 void recv_body_array(body_t *body, int count, int source, int tag);
 
+enum state_t {
+    STATE_INIT_1, STATE_INIT_2,
+    STATE_COMPUTE_1, STATE_COMPUTE_2,
+    STATE_COMMUNICATE_1, STATE_COMMUNICATE_2,
+    STATE_LOCAL_1, STATE_LOCAL_2,
+    STATE_FINAL_1, STATE_FINAL_2,
+    STATES
+};
+int state[STATES];
+
 int main(int argc, char *argv[]) {
     MPI_Init(&argc, &argv);
 #ifdef DO_TIMING
@@ -70,6 +81,23 @@ int main(int argc, char *argv[]) {
     MPI_Comm_rank(MPI_COMM_WORLD, &rank);
     MPI_Comm_size(MPI_COMM_WORLD, &size);
 
+    MPE_Init_log();
+    for(int i = 0; i < STATES; i ++) {
+        state[i] = MPE_Log_get_event_number();
+    }
+    if(rank == 0) {
+        MPE_Describe_state(state[STATE_INIT_1], state[STATE_INIT_2],
+            "Initialize", "red");
+        MPE_Describe_state(state[STATE_COMPUTE_1], state[STATE_COMPUTE_2],
+            "Shared compute", "blue");
+        MPE_Describe_state(state[STATE_COMMUNICATE_1], state[STATE_COMMUNICATE_2],
+            "Communicate", "yellow");
+        MPE_Describe_state(state[STATE_LOCAL_1], state[STATE_LOCAL_2],
+            "Local compute", "green");
+        MPE_Describe_state(state[STATE_FINAL_1], state[STATE_FINAL_2],
+            "Finalize", "orange");
+    }
+
     if(argc != 4) {
         if(rank == 0) printf(USAGE, argv[0]);
         return 1;
@@ -82,8 +110,10 @@ int main(int argc, char *argv[]) {
             data = parse_input(argv[3], size*granularity);
         }
 
+        MPE_Start_log();
         setup_simulation(data, rounds, granularity, rank, size);
         free(data);  // do nothing if NULL
+        MPE_Finish_log("cpilog");
     }
 
 #ifdef DO_TIMING
@@ -139,6 +169,7 @@ body_t *parse_input(const char *filename, int lines) {
 void setup_simulation(body_t *data, int rounds, int granularity,
     int self, int bodies) {
 
+    MPE_Log_event(state[STATE_INIT_1], 0, "start init");
     body_t *body_array = malloc(granularity * sizeof(*body_array));
 
     if(self == 0) {
@@ -158,9 +189,11 @@ void setup_simulation(body_t *data, int rounds, int granularity,
         }
         free(temp_body);
     }
+    MPE_Log_event(state[STATE_INIT_2], 0, "end init");
 
     perform_simulation(body_array, rounds, granularity, self, bodies);
 
+    MPE_Log_event(state[STATE_FINAL_1], 0, "start finalize");
     if(self == 0) {
         for(int j = 0; j < granularity; j ++) {
             printf("%.10f %.10f %f\n",
@@ -188,6 +221,7 @@ void setup_simulation(body_t *data, int rounds, int granularity,
         free(temp_body);
     }
     free(body_array);
+    MPE_Log_event(state[STATE_FINAL_2], 0, "end finalize");
 }
 
 void perform_simulation(body_t *body_array, int rounds, int granularity,
@@ -245,6 +279,7 @@ void simulation_step(body_t *body_array, int granularity, int self, int bodies
             continue;
         }
 
+        MPE_Log_event(state[STATE_COMMUNICATE_1], 0, "start send/recv");
         body_t *other = malloc(granularity * sizeof(*other));
         recv_body_array(other, granularity, prev, TAG_SIMULATE);
 #ifdef VISUALIZE_OUTPUT
@@ -253,7 +288,9 @@ void simulation_step(body_t *body_array, int granularity, int self, int bodies
         }
 #endif
         if(p + 1 < bodies) send_body_array(other, granularity, next, TAG_SIMULATE);
+        MPE_Log_event(state[STATE_COMMUNICATE_2], 0, "end send/recv");
 
+        MPE_Log_event(state[STATE_COMPUTE_1], 0, "start merge computation");
         for(int a = 0; a < granularity; a ++) {
             for(int b = 0; b < granularity; b ++) {
                 vector_t pos_vector;
@@ -270,8 +307,10 @@ void simulation_step(body_t *body_array, int granularity, int self, int bodies
             }
         }
         free(other);
+        MPE_Log_event(state[STATE_COMPUTE_2], 0, "end merge computation");
     }
 
+    MPE_Log_event(state[STATE_LOCAL_1], 0, "start local computation");
     for(int a = 0; a < granularity; a ++) {
         for(int b = a + 1; b < granularity; b ++) {
             vector_t pos_vector;
@@ -301,6 +340,7 @@ void simulation_step(body_t *body_array, int granularity, int self, int bodies
             body_array[a].pos.x, body_array[a].pos.y,
             body_array[a].velocity.x, body_array[a].velocity.y, body_array[a].mass);
     }
+    MPE_Log_event(state[STATE_LOCAL_2], 0, "end local computation");
 
     free(force_x);
     free(force_y);