Matt Cossins Academic Ecosystem Development Manager

Brief Bio:
Matt holds an MEng in Electrical and Electronic Engineering from the University of Nottingham and specialised in his thesis on neuromorphic AI & ML. He has held engineering roles at Capgemini (global technology consultancy), working on software/embedded research projects for multiple clients, and at cellXica, in Cambridge, working on embedded and RTL design for software-defined radio in 5G communications. At Arm, his work bridges the gap between engineering and academia for education and research projects.

Kieran Hejmadi Software and Academic Ecosystem Manager

Brief Bio:
Kieran holds MEng in Electronic Engineering with Nanotechnology from the University of Southampton. Kieran's career began at Sony, where he worked as an engineer in the image processing. Kieran then transitioned to the life sciences sector, joining the startup unicorn Oxford Nanopore. There, he was responsible for delivering machine learning and software and features for DNA sequencing spanning EdgeAI to server-class hardware.

Talk Title: Arm Kleidi : Scaling AI Deployments on Arm Neoverse Platform

Abstract:
Arm Kleidi is Arm’s new technology designed to unlock best-in-class on-CPU inference performance and price-performance for cloud deployments on Neoverse-based platforms. This talk will show how Kleidi AI micro-kernels and Arm Compute Library optimizations, combined with Neoverse features like BF16, INT8/INT4 and MMLA, accelerate real workloads behind the scenes in frameworks such as PyTorch. We’ll walk through concrete LLM and traditional ML case studies—such as Llama 3 INT4 text generation and XGBoost recommenders, highlighting gains in throughput, latency and TCO versus x86 and mid-range GPUs.