Algorithm developers are usually using double precision data types to be able to focus on the mathematical functionality of the algorithm. When this algorithm is implemented as a hardware module, the data accuracy must be reduced to minimum number of bits that still fulfills the system performance requirements. The process of converting the floating-point algorithm to bit-level optimized model is complicated and requires special knowledge. This webinar introduces a simple and robust quantization methodology based on value range analysis.
What You Will Learn
- What is fixed-point conversion a.k.a quantization
- Dynamic and static quantization methods
- Handling special cases
- Using Catapult Value Range Analysis feature for quantizing HLS design
Who Should Attend
- Algorithm developers
- HLS designers
- HW designers
- Verification Engineers
Meet the Speaker
Petri Solanti
Senior Application Engineer
Petri Solanti is a senior application engineer at Siemens, with an HLS and low-power tools focus. He is a designer and application engineer with over 25 years of experience in Electronics System-Level design tools and methodologies. His areas of interest include design methodologies from algorithm to RTL, system analysis and HW/SW co-design. Prior to Mentor, Mr. Solanti held application engineer positions at Cadence, CoWare, Synopsys and MathWorks. He received his MScEE degree from Tampere University of Technology, Finland.
Conteúdo informativo relacionado
Gerenciamento de certificação digital
Desde o menor planador e helicóptero até o maior avião comercial e jato militar, todas as aeronaves devem receber e manter a prova de que seu projeto e todas as cópias de produção são aeronavegadas.