GoHEP: what can Go do for science?
summary
GoHEP is a Go library collection (Gonum, Apache Arrow support, NPY/HDF5 IO, Jupyter kernel) plus a concurrent framework and a detector toolkit FADS (fast detector simulation) that ports the C++ Delphes logic. Benchmarks show a proof-of-concept analysis pipeline using GoHEP runs twice as fast and often with lower RSS than the original C++/Python mix; concurrency scales well until ~30 workers. Interop is via cgo and auto-generated Python bindings; physics output matches Delphes-level results for the tested cases.
AI-written summary · about
chapters (22)
- 0:00Introduction
- 0:46CERN
- 2:36Fortran
- 3:38Python
- 4:44Generators
- 8:27People
- 9:39Developers
- 11:07Size
- 11:52Application
- 12:56Frameworks
- 13:39Scalability
- 14:23Why Go
- 17:23Roadie
- 19:11Framework
- 20:57Framework application
- 22:21Framework interface
- 25:14CPU Killing
- 25:51More IO
- 26:20Khanoom packages
- 27:26Data science packages
- 29:01Conclusion
- 29:36Interoperability
related talks
04mGo and data, can they ever be friends?
33mData Snarfing with Go: Heka Good Time
39mGophers riding Elephants
35mState of the Gopher Nation
12mLeveraging Go for Efficient Infrastructure&Data Handling
45mGo(lang) to Python
47mUsing Go in DevOps
28mGo Make Something Real – The Potential for Go on the Factory Floor