Testimonials

The KIAC computing facility has been a valuable resource for our research, providing the computational power needed for large-scale deep learning experiments. Much of our model training, hyperparameter tuning, and experimental evaluation was carried out on the KIAC cluster, allowing us to run extensive experiments efficiently that would have otherwise taken a considerable amount of time on local machines.

The facility has played an important role in accelerating our research progress. In particular, we are currently preparing a manuscript for submission to a top-tier international conference, namely The IEEE/CVF Winter Conference on Applications of Computer Vision 2027, and a significant portion of the experimental results presented in this work were generated using the KIAC cluster.

We are grateful to the KIAC team for maintaining such a reliable and accessible computing infrastructure. Their support has been instrumental in enabling and advancing our research efforts.

Siladittya Manna
Research Associate,
(working with Prof Anirban Chakraborty),
Department of Computational and Data Sciences, IISc

The KIAC computing facility, Mahamathi has proven to be invaluable in pursing my research, enabling me to explore directions and rapidly iterate over experiments that would have otherwise been practically infeasible.

Specifically, the Mahamathi cluster has allowed me to perform large-scale batch inference experiments with generative models for our paper, ‘Don’t Settle at the Mode! Mitigating Feature Collapse in Pretrained Flow Models via Feature Self-Guidance’, which has been accepted at the European Conference of Computer Vision (ECCV) 2026.

In addition, utilising the H200 cluster on Mahamathi has enabled large-scale training of generative models for 3D editing in real images. This work has resulted in a preprint submission on arXiv, ‘Thinking in Boxes: 3D Editing in Real Images Made Easy’.

I look forward to continuing to utilise the KIAC compute facility for my upcoming research experiments, and I am incredibly grateful to Kotak Mahindra Bank Limited for providing these resources.

Pradhaan S Bhat
Research Intern,
(working with Prof Venkatesh Babu Radhakrishnan),
Department of Computational and Data Sciences, IISc

The KIAC computing facility has significantly accelerated my research capabilities. The cluster’s robust SLURM-based environment, coupled with access to top-tier GPUs provided the immense computational power required for my work.

In my research focusing on flow/diffusion models, I have relied almost exclusively on the ADA and A100 cards. Their performance has been outstanding, allowing me to execute high-compute jobs efficiently and without bottlenecks. Furthermore, the facility is exceptionally well-maintained. The IT and support staff deserve special mention for their rapid response times, and their quick turnaround ensures minimal downtime for our projects. I am immensely grateful to Kotak Mahindra Bank Limited for providing these resources, and I look forward to relying on the KIAC facility for my upcoming experiments.

Dasaraju Karthik Nandan
MTech (Research) student,
(working with Prof Venkatesh Babu Radhakrishnan),
Department of Computational and Data Sciences, IISc

I have been using the KIAC computing infrastructure extensively during both my MTech and current JRF tenure at IISc. The A6000-based KIAC cluster supported my initial research work, while the KIAC and Mahamati clusters, with H200, Ada, A100, and A6000 GPUs, have been very useful for our deep learning-based plug-and-play reconstruction experiments.

These resources have significantly helped our training and inference workloads and contributed to our recent publications at ICML and ECCV 2026.

Trishit Mukherjee
Junior Research Fellow,
(working with Prof Kunal Narayan Chaudhury),
Department of Electrical Engineering, IISc

I am sincerely grateful to KIAC for providing access to state-of-the-art GPU cluster resources, which have played a crucial role in supporting my research projects. These resources helped me design, train, and evaluate models efficiently for projects that eventually led to prestigious publications. Without the KIAC GPU cluster, several of these projects would have taken significantly longer to complete.

I would also like to specially thank the KIAC administration team, in particular Ms Neetha Ashtakar, for their constant support and active management of the cluster. The team has been extremely responsive in resolving issues, managing resources efficiently, and making thoughtful decisions that benefit the entire user community.

I strongly believe that the KIAC cluster has made a significant contribution to the progress IISc is making in AI and machine learning research. I hope KIAC continues to support and strengthen the research ecosystem at IISc in the future.

Arghya Sinha
PhD student,
(working with Prof Kunal Narayan Chaudhury),
Department of Computational and Data Sciences, IISc