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 GCC compiler flags (minimize execution time))

Scenario UID8289e0cf24346aa7 (experiment.tune.compiler.flags.gcc.e)
Data UIDd24a4fde9f120e10
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 8289e0cf24346aa7:d24a4fde9f120e10
CompilerGCC 4.9.2
CPUBCM2709
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 b669e32307c658b1 9.00 0.31 -O3 -flto -O3 6 3 123 170 milepost-codelet-mibench-automotive-bitcount-src-bitcnts-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S2 f0542a47f19b56c6 6.37 0.31 -O3 -fno-inline -flto -O3 1 1 190 228 milepost-codelet-mibench-telecomm-fft-fftmisc-codelet-5-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S3 501f4ba740bea202 2.78 0.77 -O3 -fno-if-conversion2 -funroll-loops -O3 2 1 134 181 milepost-codelet-mibench-consumer-lame-src-quantize-pvt-codelet-6-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S4 99b9c25ad1975add 2.49 0.99 -O3 -fpeel-loops -ftracer -O3 1 3 81 130 milepost-codelet-mibench-consumer-lame-src-quantize-pvt-codelet-6-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S5 2ccbf6c1f0324486 1.62 0.84 -O3 -floop-nest-optimize -fno-sched-interblock -fno-tree-copy-prop -funroll-all-loops -O3 4 1 118 166 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S6 dfd7bb4c138d57af 1.62 0.84 -O3 -funroll-loops -O3 2 3 210 243 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S7 b13457eb65c0b743 1.62 0.84 -O3 -floop-strip-mine -funroll-loops -O3 1 1 150 194 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S8 b9a130514a0c890b 1.62 0.84 -O3 -fno-inline -fno-merge-all-constants -fno-tree-ccp -funroll-all-loops -O3 2 3 166 208 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S9 30d874f89911c336 1.62 0.84 -O3 -fno-tree-loop-if-convert -funroll-all-loops -O3 3 2 214 245 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S10 2367ded45bc915d8 1.61 0.84 -O3 -fno-section-anchors -fselective-scheduling2 -fno-tree-forwprop -funroll-all-loops -O3 2 2 183 222 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S11 15989ae392f54913 1.59 0.83 -O3 -fno-ivopts -funroll-loops -O3 4 1 207 242 milepost-codelet-mibench-automotive-susan-s-src-susan-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S12 94cc88432dac4d68 1.51 0.77 -O3 -fno-tree-ch -funroll-all-loops -O3 1 1 119 166 milepost-codelet-mibench-security-pgp-e-src-mpilib-codelet-3-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S13 4ac69bb6998110ac 1.51 0.77 -O3 -fno-move-loop-invariants -fno-tree-ch -funroll-loops -O3 1 2 138 184 milepost-codelet-mibench-security-pgp-e-src-mpilib-codelet-3-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S14 df142a586c394a85 1.40 1.05 -O3 -fira-algorithm=priority -fno-ivopts -O3 1 2 174 215 milepost-codelet-mibench-consumer-jpeg-c-src-jfdctint-codelet-2-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S15 04e8098c30dcd57e 1.39 1.06 -O3 -fno-ivopts -O3 2 4 220 248 milepost-codelet-mibench-consumer-jpeg-c-src-jfdctint-codelet-2-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S16 82e3849ef4bfcbe0 1.36 1.01 -O3 -fno-sched-spec -fno-tree-ch -O3 1 2 140 185 milepost-codelet-mibench-automotive-susan-e-src-susan-codelet-10-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S17 6d1b29c25d254c1e 1.34 1.06 -O3 -fno-ivopts -fselective-scheduling -fwhole-program -O3 1 1 202 238 cbench-automotive-susan corners image-pgm-clean-gray-square-600-450-8 data.pgm 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S18 3b8835387e3ef430 1.30 1.00 -O3 -fno-omit-frame-pointer -fno-tree-loop-optimize -O3 1 4 94 143 milepost-codelet-mibench-consumer-tiffmedian-src-tiffmedian-codelet-3-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S19 289b999fde018959 1.30 1.00 -O3 -fno-auto-inc-dec -ffinite-math-only -O3 1 2 54 104 milepost-codelet-mibench-consumer-tiffmedian-src-tiffmedian-codelet-3-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S20 68c75f44b7aa20b4 1.25 1.03 -O3 -fno-guess-branch-probability -fira-loop-pressure -fno-toplevel-reorder -O3 1 5 213 244 milepost-codelet-mibench-automotive-susan-e-src-susan-codelet-10-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S21 b4147872aeb87292 1.20 1.01 -O3 -fselective-scheduling2 -fno-tree-pre -O3 2 2 158 201 milepost-codelet-mibench-telecomm-adpcm-c-src-adpcm-codelet-1-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S22 5cc84565aa45fd16 1.19 0.85 -O3 -fgcse-sm -fno-move-loop-invariants -fno-tree-forwprop -funroll-all-loops -fno-web -O3 1 0 5 226 zlib encode image-pgm-clean-gray-square-600-450-8 data.pgm 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32
S23 d899b4a94364aaa3 1.18 1 -O3 -fno-schedule-insns -fselective-scheduling2 -O3 1 2 78 128 milepost-codelet-mibench-consumer-tiffmedian-src-tiffmedian-codelet-3-1 default 1200, 1200, 1200, 1200 4 Raspberry Pi 3 Linux 4.4.32



[ 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