diff --git a/montecarlo/montecarlo/main.cpp b/montecarlo/montecarlo/main.cpp
index d7c167a..607ede4 100644
--- a/montecarlo/montecarlo/main.cpp
+++ b/montecarlo/montecarlo/main.cpp
@@ -9,53 +9,76 @@
 #include <thread>
 #include <random>
 #include <chrono>
+#include <iomanip>
 
-void compute(std::atomic<long long> &inside, long long iterations){
-    long long to_add = 0;
+struct stats {
+    std::atomic<long long> iterations;
+    std::atomic<long long> points_inside;
+    std::chrono::time_point<std::chrono::steady_clock> start_time;
+    
+    stats(std::chrono::time_point<std::chrono::steady_clock> start){
+        iterations = 0;
+        points_inside = 0;
+        start_time = start;
+    }
+};
+
+void compute(std::atomic<long long> &inside, std::atomic<long long> &iterations){
     unsigned int seed = (unsigned int) std::chrono::system_clock::now().time_since_epoch().count();
     std::mt19937 e2(seed);
     std::uniform_real_distribution<double> dis(0.0, 1.0);
-    for(int i=0; i<iterations; i++){
-        double x = dis(e2);
-        double y = dis(e2);
-        if(x*x+y*y<=1){
-            to_add++;
+    while(1){
+        long long add_inside = 0;
+        for(int i=0; i<250000; i++){
+            double x = dis(e2);
+            double y = dis(e2);
+            if(x*x+y*y<1){
+                add_inside++;
+            }
         }
+        inside+=add_inside;
+        iterations+=250000;
+
     }
-    inside+=to_add;
 }
 
-double monte_carlo(long long iterations){
-    
-    std::atomic<long long> inside(0);
+void monte_carlo(stats &stats){
     
     std::vector<std::thread> threads;
     
     int threadcount = std::thread::hardware_concurrency();;
     
-    long long actual_n = iterations - iterations % threadcount;
-    
     for(int i=0; i<threadcount; i++){
-        threads.emplace_back(compute, std::ref(inside), iterations/threadcount);
+        threads.emplace_back(compute, std::ref(stats.points_inside), std::ref(stats.iterations));
     }
     
+    
+    
     for(auto& t : threads){
         t.join();
     }
+}
+
+void report_stats(stats &full_stats){
+    using namespace std::chrono_literals;
+    double pi = 4*(double)full_stats.points_inside/(double)full_stats.iterations;
+    auto time_ms = std::chrono::duration_cast<std::chrono::milliseconds>(std::chrono::steady_clock::now()-full_stats.start_time);
     
-    return inside * 4 / (double)actual_n;
+    double iterations_per_second = full_stats.iterations / time_ms.count() * 1000;
+    
+    std::cout << "ran for "<<time_ms <<"\n";
+    std::cout << "iterations: "<<full_stats.iterations << "\ninside: "<< full_stats.points_inside <<std::fixed<< std::setprecision(10)<< "\npi: "<<pi<<"\n";
+    std::cout << "speed: "<<std::scientific<<std::setprecision(2)<<iterations_per_second<<"/s\n";
+    std::this_thread::sleep_for(1000ms);
+    report_stats(std::ref(full_stats));
 }
 
 int main() {
-    // testing with 1,000,000,000 iterations for now
-    std::vector<double> pi_values;
-    double pi_average = 0;
-    for(int i=0; i<10; i++){
-        double pi = monte_carlo(1e9);
-        std::cout << "Computed pi #"<<i+1<<" result: "<<pi << "\n";
-        pi_values.push_back(pi);
-        pi_average+=pi/10;
-    }
-    std::cout << "iterations: "<<1e8<<", pi: "<<pi_average<<','<<"\n";
-    return 0;
+    
+    stats full_stats(std::chrono::steady_clock::now());
+    
+    std::thread rst(report_stats, std::ref(full_stats));
+    monte_carlo(full_stats);
+    
+    
 }