
Computing precision
Leading up to the summer program, we’re hosting a special series of 18 webinars called « Gray Scott Thursdays » focus on Unit tests, CPU & GPU Architecture, Computing precision, Memory allocation, CPU & GPU computing with Modern C++, Rust, Fortran, Python and Sycl, also EVE and CUDA…
These weekly sessions will:
- Introduce key topics from the courses
- Give you an update of the most known languages and libraries
- Develop your skills and give you the chance to meet our incredible professors
Speakers
Description
We compute, idealy fast with simple obects (we call them ‘primitive types’) : floating point numbers. Float or double, some languages call them Real. This webinar will highligth some surprising aspects of their behaviour. These numbers which can be represented as floating point and induced errors or the other. Moreover, defferent use cases will be presented to spot traps of floating point computing, and find the perfect balance between performance, precision and correction. An important point is the difference of performance which appears on several hardwares when exotic numbers appears in the computation such as denormals, NaN, inf.