Collective Knowledge Aggregator proof-of-concept
Crowdsourced experiments CK project Partners Open AI powered by CK Reusable AI artifacts Get CK

Distinct solutions after online classification (auto/crowd-tune LLVM compiler flags (minimize execution time))

Scenario UID2aaed4c520956635 (experiment.tune.compiler.flags.llvm.e)
Data UID05b91165b6440b76
Discuss (optimizations to improve compilers,
semantic/data set/hardware features
to improve predictions
, etc):
GitHub wiki, Google group
Download:[ All solutions in JSON ], [ Solutions' classification in JSON ]
Reproduce all (with reactions):ck replay 2aaed4c520956635:05b91165b6440b76
CompilerLLVM 3.8.0
CPUQualcomm Technologies, Inc MSM8996
Objectivemin
Improvement key IK1Main kernel execution time speedup [min]
Improvement key IK2Code size improvement

Improvements (<4% variation) Distinct workload for highest improvement
# Solution UID IK1 IK2 New distinct optimization choices Ref Best species Worst species Touched Iters Program CMD Dataset Dataset file CPU freq (MHz) Cores Platform OS Replay
S1 90892fa8e32a5148 4.40 1.00 -O3 -fno-assume-sane-operator-new -fvisibility-inlines-hidden -fslp-vectorize -mlong-calls -fmodules-prune-after=28 -fmax-type-align=10 -O3 1 0 2 1 cbench-bzip2 decode bzip2-0001 data.bz2 1555.2, 1555.2, 1555.2, 1555.2 1 SAMSUNG SM-G930V Android 7.0
S2 7ffac749700dcd0f 1.63 1.00 -O3 -fvisibility-inlines-hidden -fmax-type-align=5 -ffp-contract=off -O3 1 0 2 1 milepost-codelet-mibench-consumer-lame-src-takehiro-codelet-5-1 default bzip2-0001 data.bz2 1036.8, 1036.8, 1036.8, 1036.8 1 SAMSUNG SM-G930V Android 6.0.1
S3 a3091b7d9edf456f 1.31 1.00 -O3 -fshort-enums -mrelax-all -O3 1 0 2 1 milepost-codelet-mibench-consumer-lame-src-newmdct-codelet-3-1 default cdataset-patricia-0001 data.txt 1324.8, 1324.8, 1324.8, 1324.8 1 SAMSUNG SM-G930V Android 6.0.1
S4 1cbc540d4b84ba16 1.29 1.00 -O3 -fno-constant-cfstrings -fstrict-enums -fwrapv -fslp-vectorize -mno-long-calls -mstack-alignment=25 -fmax-type-align=31 -ffp-contract=fast -O3 1 0 2 1 milepost-codelet-mibench-consumer-tiff2rgba-src-tif-predict-codelet-4-1 default number-0001 1036.8, 1036.8, 1036.8, 1036.8 1 SAMSUNG SM-G930V Android 6.0.1
S5 e864396b9f14853b 1.28 1.00 -O3 -fno-assume-sane-operator-new -fstrict-enums -fslp-vectorize-aggressive -fmodules-prune-interval=118 -O3 1 0 2 1 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default cdataset-dijkstra-0001 data.txt 1190.4, 1190.4, 1190.4, 1190.4 1 SAMSUNG SM-G930V Android 6.0.1
S6 629c1e5a4a0790ee 1.14 1.00 -O3 -freroll-loops -trigraphs -fwrapv -ftemplate-backtrace-limit=637 -O3 1 0 2 1 milepost-codelet-mibench-office-ghostscript-src-gdevpbm-codelet-1-1 default DGEMM_NT DGEMM_NT_2x1.json 1190.4, 1190.4, 1190.4, 1190.4 1 SAMSUNG SM-G930V Android 6.0.1
S7 0f951e112d54980a 1.10 1.00 -O3 -fno-builtin -fomit-frame-pointer -mstrict-align -fslp-vectorize -fstrict-vtable-pointers -fmodules-prune-interval=98 -O3 1 0 2 1 milepost-codelet-mibench-telecomm-adpcm-c-src-adpcm-codelet-1-1 default image-pgm-from-pamela-100 data.pgm 1036.8, 1036.8, 1036.8, 1036.8 1 SAMSUNG SM-G930V Android 6.0.1



[ Participated users, platforms, OS, CPU, GPU, GPGPU, NN ] [ How to participate ] [ Slides ] [ Paper ] [ Android app ] [ dividiti ] [ Collective training set ] [ Unified AI ]
View entry in raw format

Developed by dividiti,
cTuning foundation,
and the community
          
Implemented as a CK workflow
                     
   
   
                      Hosted at