Feb
21
Protected: logs and metrics made easy with OVH
There is no excerpt because this is a protected post.
Sep
19
Crystals – building made easy
Crystals is fully open source since a few years but building from source was always a tedious task. After many years, improvements have been made. The old Linux Makefiles and windows bat files have been replaced with cmake. The same … Continue reading
Jul
09
Compression and crystallograpic 2D images
Since the apparition of area detectors, the storage of the frames resulting of the experiment has always been troublesome. The apparition of bigger and faster detectors increased the problem where a lots of big images are produced. Several methods exist … Continue reading
Sep
06
eigenvalue decomposition benchmark
After looking at the Cholesky factorization problem, time to have a look at the eigenvalue decomposition of a symmetric matrix. Depending on the function and the blas library used, lots of differences appear. The functions which are benchmarked here are: … Continue reading
Sep
01
Lapack and packed storage
Crystals is using packed storage to store symmetric matrices. Packed storage is using a linear array to store only the unique data (Netlib packed storage). There is a huge penalty on the calculation by doing so in some cases. In … Continue reading
Aug
15
reading/writing byte offset compression in cbf
The compression scheme “byte offset” is commonly used in frames encoded in cbf. The Pilatus hybrid pixel detectors are using this format. The algorithm is descrived here: http://www.bernstein-plus-sons.com/software/CBF/doc/CBFlib.html#3.3 and a library already exist to manipulate cbf files called cbflib. The … Continue reading
May
17
Parallel structure factors calculation
As part of my new position in Oxford, I have a new computer: An 8-cores Intel Xeon CPU E5-2665 with 16GB of memory. Enough cores to try if my library is scaling well in multi-threaded environments. The result is really … Continue reading
Apr
14
autovectorization: “not vectorized: not suitable for gather”
The cryptic warning “not vectorized: not suitable for gather” is from a report of gfortran on auto-vectorization. For a long time, gfortran was reporting this failed optimization on a very simple mathematical operation. I managed to find the cause today … Continue reading
Feb
10
Fortran weakness: byte wise I/O
In the Crm² lab, in Nancy (France), we have prototype of an hybrid pixel detector from the company Imxpad. The data from the detector are raw ASCII files that I need to convert into files suitable for integration software. We … Continue reading
Dec
17
CPU evolution, an increase in complexity
Moore’s law has driven computing power for years and will probably continue for sometime even though it ends is coming. However, if during the first 20 years the increasing power was directly related to the frequency, it’s no longer the … Continue reading