Meauring cost of extracted decision procedures

jaccokrijnen / plutus-cert

0 stars 2 forks source link

Open jaccokrijnen opened 1 year ago

jaccokrijnen commented 1 year ago

[ ] test performance of "golden tests", possibly nofib
- see plutus-ir/test , see TransformSpec.hs
- Running cabal test plutus-ir-test
[ ] test against the marlowe smart contract.
- See https://github.com/input-output-hk/plutus/blob/master/plutus-benchmark/marlowe/exe/Main.hs
[ ] Compare lines-of-code
- Both generated and hand-written, compared to compiler code base

jaccokrijnen commented 1 year ago

Running times (with extracted decision procedures) 8.40s 10.71s 9.36s 8.81s 9.04s

AVG: 9.264s

jaccokrijnen commented 1 year ago

Runinng times (without extracted decision procedures) 8.78s 12.86s 12.73s 11.08s 8.74

AVG: 10.83s

jaccokrijnen commented 1 year ago

Haskell lines of code

generated: 902
glue code: 275
plutus: 74035
plutus/plutus-tx-plugin: 5633
plutus/plutus-tx: 7531
- plutus/plutus-core/plutus-ir: 7388
- plutus/plutus-core/plutus-core: 20941

0.3% glue code 1% generated code