IIDAI Future Research Agenda

Building on the research collaborations between IBM and Illinois over the past five years, the future technical agenda of the Institute will focus on the following:

Hybrid Cloud Platform for AI
We are focusing on scalable and efficient inference systems through three broad areas: optimizing communication and platform capabilities across multi-node, multi-accelerator setups based on inferencing patterns, SLOs, and infrastructure; exploring novel AI model architectures beyond State-Space Models and leveraging hardware-aware compilers for co-optimization; and advancing secure, verifiable, and accurate agentic automation to enable intelligent IT management.

Quantum-Centric Supercomputing
Quantum-Centric Supercomputing (QCSC) is a computing platform that integrates quantum and any other classical systems like HPC, GPU, TPU.  We are developing SW tools and architectures for QCSC such that users can use QCSC systems efficiently. Topics of interest focus on building seamless orchestration of quantum and classical computing resources, efficient workflow design between quantum and classical computation, and developing optimum programming models and library tools for new users that will join from both quantum and HPC/GPU/TPU communities.

Algorithms and Applications for Emerging Compute Platforms
We are developing algorithms and mathematical frameworks that exploit quantum computing, AI or quantum-HPC to enhance the efficiency of existing algorithms or solve scientific problems better than existing algorithms. The proposed algorithms or mathematical frameworks can use quantum computing, AI, and QCSC to solve mathematical problems, and require demonstration that the developed algorithms can solve certain scientific problems faster, cheaper or more accurately.  Industry use-cases of these algorithms and mathematical frameworks that can be demonstrated with a prototype are also within the focus.