Large benches

• Add "large" Rodinia benchmarks for higher CPU utilization.

juno_samples/rodinia/srad/src/cpu.sch

@@ -40,10 +40,15 @@ let split = fork-split(loop2);
 let loop2_body = outline(split.srad_1.fj1);
 simpl!(loop2, loop2_body);
 
-inline(srad@loop2);
+fork-tile[32, 0, false, false](loop3);
+let split = fork-split(loop3);
+let loop3_body = outline(split.srad_2.fj1);
+simpl!(loop3, loop3_body);
+
+inline(srad@loop2, srad@loop3);
 delete-uncalled(*);
 
-fork-split(extract, compress, loop1, loop2_body, loop3);
-unforkify(extract, compress, loop1, loop2_body, loop3);
+fork-split(extract, compress, loop1, loop2_body, loop3_body);
+unforkify(extract, compress, loop1, loop2_body, loop3_body);
 
 gcm(*);