Contribution Opportunities at the SBME Department

It is insufficient to teach our students only theoretical AI, or AI applications that are only developed and tested on powerful workstations. We are eager to enrich our students' background with the engineering principles of deploying ML on power-efficient, resource-constrained devices, such as this board. These boards will be employed in Embedded Systems and Biomedical Edge AI courses, among others, in addition to relevant graduation projects.

The Kria KV260 is a vision-AI starter kit (AMD K26 SOM) for real-time computer vision and edge inference.

Learn more: AMD Kria KV260 Vision AI Kit · Kria apps documentation

This year our new cluster supports up to 40 students and researchers. In 2027, we aim to phase in a new cluster that can serve up to 120 students, which requires adding more compute capacity. We chose this model as it will contribute significantly to this capacity.

The new cluster enriches node connectivity with an average of 5 Gbps bandwidth for cluster nodes and 20 Gbps bandwidth for data-serving nodes.

It is insufficient to teach our students only theoretical AI, or AI applications that are only developed and tested on powerful workstations. We are eager to enrich our students' background with the engineering principles of deploying ML on power-efficient, resource-constrained devices, such as this board. These boards will be employed in Embedded Systems and Biomedical Edge AI courses, among others, in addition to relevant graduation projects.

The PYNQ-Z2 pairs a Xilinx Zynq-7000 FPGA with Python (PYNQ), so students go from notebook to programmable silicon.

Learn more: PYNQ project · PYNQ documentation · TUL PYNQ-Z2 board

This year our new cluster supports up to 40 students and researchers. In 2027, we aim to phase in a new cluster that can serve up to 120 students, which requires adding more compute capacity. We chose this model as it will contribute significantly to this capacity.