Releases: DC-analysis/ChipStream
ChipStream 0.9.14
This release was built with CUDA 13 and some of the older NVIDIA GPUs might not be supported.
According to https://en.wikipedia.org/wiki/CUDA#GPUs_supported, compute capability 7.5 is required. A list of supported GPUs can be found here: https://developer.nvidia.com/cuda/gpus/
If your GPU is supported, but ChipStream does not offer you the "Use GPU" option, try to update the GPU drivers: https://www.nvidia.com/de-de/drivers/
ChipStream 0.9.13
ChipStream 0.9.13 has increased performance compared to 0.9.9. The main improvements are due to the bump of dcnum from 0.28.3 to 0.29.1 (lazy-loading, extension of universal workers).
- enh: put
mainmethods in__name__ == "__main__"guard - enh: check for cuda availability instead of letting dcnum decide
- setup: bump dcnum to 0.29.1
- setup: update pyinstaller hooks
ChipStream 0.9.9
ChipStream 0.9.8
ChipStream 0.9.5
I implemented the first set of optimizations in dcnum to speed-up data processing 🚀.
In addition, the pipeline identifier changed because of a bug in dcnum (DC-analysis/dcnum#47). This means that you should reanalyze previous data if you would like to compare them to data analyzed with ChipStream 0.9.5.
ChipStream 0.8.0
ChipStream 0.8.0 fixes a bug that caused analysis processes to continue (instead of stopping) after the GUI was closed. In addition, ChipStream now writes measurement identifiers in basins which makes it easier for DCOR to identify basin resources within the same dataset.
- fix: when user quits GUI during processing, forcibly kill all processes
- docs: add installation section
- setup: add psutil dependency
- setup: bump dcnum to 0.25.11 (measurement identifiers in basins)
ChipStream 0.7.0
ChipStream 0.6.8
ChipStream 0.6.5
This is a maintenance release with improved error messages.
- enh: improve error message when input file is invalid (#13)