Merge branch 'main' into hops

Author: gilbertmike

accelforge/mapper/FFM/_make_pmappings/make_pmappings_from_templates/make_tile_shapes.py

@@ -906,8 +906,9 @@ def get_padded_choices(
         if subs in choices_padded:
             choices_padded[s] = choices_padded[subs]
         else:
-            choices_padded[s] = ones * subs
-
+            choices_padded[s] = np.full(
+                choices_enumerated.shape[0], subs, dtype=choices_enumerated.dtype
+            )
     for k, v in choices_padded.items():
         assert isinstance(v, np.ndarray), f"{k} is not a numeric array: {v}"
 
@@ -1781,10 +1782,10 @@ def update_symbol2goal(
                         # Haven't done any pruning, so valid is the # of total choices
                         valid = [choices_enumerated.shape[0]]
 
-                    porp = sum(valid) / max(1, choices_enumerated.shape[0])
+                    porp = int(np.sum(valid)) / max(1, choices_enumerated.shape[0])
                     job.log_porp_pmappings_kept(
                         f"{objective.name}",
-                        sum(valid) / max(1, prev_size),
+                        int(np.sum(valid)) / max(1, prev_size),
                     )
                     DEBUG and log_message(
                         f"Valid check", f"{objective.name}", f"porp={porp:.2%}"
@@ -1832,10 +1833,10 @@ def update_symbol2goal(
                         # Everyone valid (for counting purposes)
                         valid = [choices_enumerated.shape[0]]
 
-                    porp = sum(valid) / max(1, choices_enumerated.shape[0])
+                    porp = int(np.sum(valid)) / max(1, choices_enumerated.shape[0])
                     job.log_porp_pmappings_kept(
                         f"{objective.name}",
-                        sum(valid) / max(1, prev_size),
+                        int(np.sum(valid)) / max(1, prev_size),
                     )
                     DEBUG and log_message(
                         f"Valid check", f"{objective.name}", f"porp={porp:.2%}"
@@ -1980,7 +1981,7 @@ def update_symbol2goal(
                 prev_size = choices_enumerated.shape[0]
                 choices_enumerated = choices_enumerated[keep]
                 job.log_porp_pmappings_kept(
-                    f"Pareto", sum(keep) / choices_enumerated.shape[0]
+                    f"Pareto", int(np.sum(keep)) / choices_enumerated.shape[0]
                 )
                 DEBUG and log_message(
                     "pareto", f"size {prev_size} -> {choices_enumerated.shape[0]}"

tests/hwcomponents_expected.json

@@ -15,67 +15,67 @@
       }
     },
     "GlobalBuffer": {
-      "area": 6.85753842326923e-06,
-      "leak_power": 0.00025518689743667236,
+      "area": 6.8574526567307686e-06,
+      "leak_power": 0.00025508039081958506,
       "actions": {
         "read": {
-          "energy": 9.056648304726114e-11,
+          "energy": 9.056430298994264e-11,
           "latency": 1.0572650390625e-10
         },
         "write": {
-          "energy": 9.122564946135314e-11,
+          "energy": 9.122346940403464e-11,
           "latency": 2.0728900390625002e-10
         }
       }
     },
     "InputScratchpad": {
-      "area": 4.2167274519230763e-10,
-      "leak_power": 4.807610371689463e-07,
+      "area": 3.8736612980769225e-10,
+      "leak_power": 4.3815839033400895e-07,
       "actions": {
         "read": {
-          "energy": 2.3605916811887966e-14,
+          "energy": 2.2733893884485344e-14,
           "latency": 1.4807425e-10
         },
         "write": {
-          "energy": 3.242682392836752e-14,
+          "energy": 3.1554801000964905e-14,
           "latency": 2.4963675e-10
         }
       }
     },
     "WeightScratchpad": {
-      "area": 3.585853961538461e-09,
-      "leak_power": 8.302724620034194e-07,
+      "area": 3.534394038461538e-09,
+      "leak_power": 7.663684917510134e-07,
       "actions": {
         "read": {
-          "energy": 4.0526689360198025e-14,
+          "energy": 4.0152965248454046e-14,
           "latency": 4.4058267857142854e-11
         },
         "write": {
-          "energy": 6.272942751096912e-14,
+          "energy": 6.235570339922515e-14,
           "latency": 7.307612500000001e-11
         }
       }
     },
     "OutputScratchpad": {
-      "area": 4.904405576923076e-10,
-      "leak_power": 4.840417246689463e-07,
+      "area": 4.5613394230769223e-10,
+      "leak_power": 4.4143907783400897e-07,
       "actions": {
         "read": {
-          "energy": 2.7119425256184493e-14,
+          "energy": 2.6247402328781873e-14,
           "latency": 1.4807425e-10
         },
         "write": {
-          "energy": 3.815933349282664e-14,
+          "energy": 3.728731056542402e-14,
           "latency": 2.4963675e-10
         }
       }
     },
     "MAC": {
-      "area": 1.3973971153846152e-09,
-      "leak_power": 9.732325476491557e-06,
+      "area": 1.1748028846153846e-09,
+      "leak_power": 8.851870775236185e-06,
       "actions": {
         "compute": {
-          "energy": 1.408832599217931e-12,
+          "energy": 1.2813796807700769e-12,
           "latency": 3.25e-09
         }
       }
@@ -97,57 +97,57 @@
       }
     },
     "GlobalBuffer": {
-      "area": 2.3952215381843155e-07,
-      "leak_power": 9.523558182536147e-06,
+      "area": 2.3951167124150845e-07,
+      "leak_power": 9.50865964351506e-06,
       "actions": {
         "read": {
-          "energy": 8.295549237325948e-14,
+          "energy": 8.294876234356374e-14,
           "latency": 9.852990595611286e-13
         },
         "write": {
-          "energy": 9.035066030244479e-14,
+          "energy": 9.034393027274906e-14,
           "latency": 1.537023197492163e-12
         }
       }
     },
     "InputBuffer": {
-      "area": 5.399594303496503e-09,
-      "leak_power": 1.2188733186998888e-06,
+      "area": 5.391208241958042e-09,
+      "leak_power": 1.2069544874830208e-06,
       "actions": {
         "read": {
-          "energy": 9.660840031396417e-15,
+          "energy": 9.645592307867024e-15,
           "latency": 3.0015596590909094e-12
         },
         "write": {
-          "energy": 2.070367107287105e-14,
+          "energy": 2.0688423349341654e-14,
           "latency": 4.56405965909091e-12
         }
       }
     },
     "WeightBuffer": {
-      "area": 1.0704373813688813e-07,
-      "leak_power": 1.2624919331968711e-05,
+      "area": 1.0703430381765736e-07,
+      "leak_power": 1.2611510646849734e-05,
       "actions": {
         "read": {
-          "energy": 6.682829613803192e-14,
+          "energy": 6.681971929354662e-14,
           "latency": 8.655635653409092e-13
         },
         "write": {
-          "energy": 6.873944068448839e-14,
+          "energy": 6.873086384000309e-14,
           "latency": 1.6468135653409094e-12
         }
       }
     },
     "AccumulationBuffer": {
-      "area": 2.208693348951049e-09,
-      "leak_power": 6.153967734550903e-07,
+      "area": 2.2003072874125873e-09,
+      "leak_power": 6.034779422382222e-07,
       "actions": {
         "read": {
-          "energy": 7.099636607719463e-15,
+          "energy": 7.074774077633321e-15,
           "latency": 4.3586288361320215e-12
         },
         "write": {
-          "energy": 1.3851866895697208e-14,
+          "energy": 1.3827004365611064e-14,
           "latency": 6.906399536768965e-12
         }
       }
@@ -167,11 +167,11 @@
       }
     },
     "MAC": {
-      "area": 1.7436548076923076e-10,
-      "leak_power": 1.3613936377164362e-06,
+      "area": 1.4715951923076922e-10,
+      "leak_power": 1.2382323818087997e-06,
       "actions": {
         "compute": {
-          "energy": 1.9707270804040344e-13,
+          "energy": 1.7924412301183831e-13,
           "latency": 8.000000000000001e-10
         }
       }