E N

A B

LI N

G T

E C

H N

O LO

G IE

S

1 6 9 I H I G H L I G H T S 2 0 2 3

This chapter highlights the importance of having a powerful toolkit of solutions common across all beamlines. Specific computational demands posed by AI and ML highlight the necessity for robust computing clusters and specialised AI/ML processors like GPUs, which are integral to the successful implementation of these technologies. The exploration of advanced GPU solutions such as the NVIDIA GPU-H100 and of the impressive AMD GENOA 9654 with its 192 cores is indicative of the facility s commitment to staying at the cutting edge. The development of the blissdata API in BLISS will provide a robust solution for data-processing applications needing to access data online (i.e., before it hits the disk), to send feedback and control experiments from intelligent clients written in Python, e.g., using AI/ML techniques to drive data acquisition.

In summary, the ESRF considers the potential of AI/ML to be very promising and is investing in AI and ML technologies while, at the same time, continuing the development of generic toolkits like BLISS and EWOKS to make the most of its data. While the ESRF believes that AI/ML will play a role in the future, the main enabling factor in shaping the future of its operations will be human resources to implement the scientific workflows required to process the data efficiently and provide users with easy access to processed data.

J.-C. BIASCI AND C. NEVO