add updated config for prefetch from vampir

qdp_project/src/Benchmark.cpp

@@ -53,7 +53,7 @@ void scan_b(size_t gid, size_t tid) {
     if constexpr (PERFORM_CACHING) {
         for (size_t i = 0; i < RUN_COUNT; i++) {
             const size_t chunk_index = get_chunk_index(gid, i);
-            uint64_t* chunk_ptr = get_chunk<TC_SCANB>(DATA_B_, chunk_index, i);
+            uint64_t* chunk_ptr = get_chunk<TC_SCANB>(DATA_B_, chunk_index, tid);
 
             for (size_t j = 0; j < SUBCHUNK_COUNT; j++) {
                 uint64_t* sub_chunk_ptr = &chunk_ptr[j * SUBCHUNK_SIZE_ELEMENTS];
@@ -261,7 +261,7 @@ int main() {
             fout
                 << i - WARMUP_ITERATION_COUNT << ";"
                 << nanos << ";" << seconds << ";"
-                << DATA_DST_[0] << ";"
+                << result_actual << ";"
                 << scana_run << ";" << scana_wait << ";" << scanb_run << ";" << scanb_wait << ";" << aggrj_run << ";" << aggrj_wait << ";"
                 << process_cache_hitrate() << ";"
                 << std::endl;

qdp_project/src/Configuration.hpp

@@ -5,11 +5,11 @@ constexpr uint32_t WARMUP_ITERATION_COUNT = 5;
 constexpr uint32_t ITERATION_COUNT = 5;
 
 #ifdef MODE_SIMPLE_PREFETCH
-constexpr uint32_t GROUP_COUNT = 8;
-constexpr size_t CHUNK_SIZE_B = 8_MiB;
-constexpr uint32_t TC_SCANA = 4;
+constexpr uint32_t GROUP_COUNT = 16;
+constexpr size_t CHUNK_SIZE_B = 64_MiB;
+constexpr uint32_t TC_SCANA = 2;
 constexpr uint32_t TC_SCANB = 1;
-constexpr uint32_t TC_AGGRJ = 4;
+constexpr uint32_t TC_AGGRJ = 1;
 constexpr bool PERFORM_CACHING = true;
 constexpr bool STORE_B_IN_HBM = false;
 constexpr char MODE_STRING[] = "simple-prefetch";
@@ -83,4 +83,4 @@ static_assert(TC_AGGRJ % (TC_SCANB > 0 ? TC_SCANB : TC_AGGRJ) == 0);
 static_assert(TC_AGGRJ >= TC_SCANB);
 static_assert(RUN_COUNT > 0);
 static_assert(WL_SIZE_B % 16 == 0);
-static_assert(CHUNK_SIZE_B % 16 == 0);
+static_assert(CHUNK_SIZE_B % 16 == 0);