Target group
- Data scientists in research
- Data scientists in industry
Your requirements
- Energy-efficient inference with pre-trained machine learning models
- Use of high-performance computing (HPC) resources
Our offer
Three servers with FPGA accelerator cards from AMD (Xilinx) are provided at the Hanover site. Each offers 512 GB RAM and 208 threads and is equipped with the following FPGA accelerators:
- 3 × AMD VCK5000 Versal
- 6 × AMD Alveo V70
- 6 × AMD Alveo V70 The platform is suitable for developing specialised software for AI inference and allows direct comparisons with other architectures and implementations. The systems can be used with direct access by arrangement.