Sat, Nov 16, 2024 | Jumada al-Awwal 14, 1446 | DXB ktweather icon0°C

Accelerating emergent workloads in AI and genomics via industry-academia collaborations

Genomics and Artificial Intelligence (AI) are two newly emergent workloads that have raised new and big computational challenges to the world

Published: Wed 14 Feb 2024, 3:07 PM

  • By
  • Shagun Sharma

Top Stories

While genomics is revolutionizing the healthcare industry with the possibility of personalised treatment plans customised to every individual (Precision Medicine), AI is ubiquitous in 2024 from our personal devices to various LLM-powered tools. Genomics and AI permeate various aspects of human life, yet each industry has unique computing needs, ruling out a one-size-fits-all solution. Dr Harisankar Sadasivan, a trailblazer in computer science and engineering, showcases exceptional skills and spearheads groundbreaking research in identifying and optimizing these newly emerging workloads in genomics and AI. His contributions include developing several accelerated solutions to portable and programmable DNA sequencing and contributing to important accelerated GPU libraries that drive today’s most popular Large Language Models (LLMs) like ChatGPT and Co-Pilot. Dr Sadasivan's journey reflects a commitment to excellence, making him a visionary at the intersection of AI and healthcare innovation.

Professional experience and expertise: With an extensive eight-year career, Dr. Sadasivan's journey includes roles at Samsung Research and collaborations with industry giants like NVIDIA. His work at AMD focuses on open-sourced GPU library kernels for AI used by various multinational corporations and research labs. Additionally, he imparts his knowledge through teaching GPU Programming at the University of Washington, Seattle.

Revolutionising virus testing with SquiggleFilter: A notable achievement in Dr. Sadasivan's career is the design of SquiggleFilter, a portable hardware-software solution for rapid virus testing. Published in IEEE/ACM MICRO 2021, SquiggleFilter earned global recognition, receiving the "IEEE/ACM MICRO Top Picks 2022 with Honorable Mention" award. The collaboration with NVIDIA gave rise to DTWax, optimising SquiggleFilter for GPUs and extending its capabilities to tackle more complex challenges posed by larger pathogens.

Industry-academia collaboration: The partnership with NVIDIA exemplifies the transformative impact of industry collaborations. DTWax, born out of this collaboration, not only optimised algorithms for GPUs but also embraced open-sourcing, contributing to the tech community's collective knowledge pool. Dr. Sadasivan's work underscores the profound impact that collaborative efforts can have in driving technological progress.

Professional recognition and accolades: Dr Sadasivan's contributions have garnered global acclaim, with IEEE MICRO Top Picks recognizing his research as one of the best technical papers in computer architecture in 2021. The software, SquiggleFilter, earned MICRO artefact badges, further solidifying his position in the field. The Global Recognition Award 2023 and Haddard Fellowship received in 2018 stand testament to his outstanding achievements.

Current and future endeavours: As a member of the technical staff at AMD, Dr Sadasivan continues to optimize AI GPU kernels crucial for supporting leading LLMs. His contributions extend to the open-sourced development of genomics kernels, showcasing his commitment to advancing technology. Beyond his corporate role, he dedicates himself to educating the next generation, teaching GPU Programming at the University of Washington Seattle.

Educational journey and expertise: Dr Sadasivan's academic journey, including a Ph.D. and M.S. in Computer Science and Engineering from the University of Michigan Ann Arbor, reflects a pursuit of excellence. His expertise in GPU programming positions him as a standout figure in optimizing irregular workloads in genomics and AI on GPUs.

A journey of excellence and innovation: Dr Harisankar Sadasivan's journey is defined by a strong commitment to excellence and innovation. From early roles as a hardware engineer to creating transformative solutions like SquiggleFilter and DTWax, his passion and vision impact the future of technology. As he progresses in genomics and AI, the tech community anticipates his ongoing breakthroughs. Beyond work, his diverse interests in swimming, chess, and road trips showcase a well-rounded individual dedicated to pushing technological boundaries. Driven by curiosity, he remains an inspiring force in the tech community, illustrating the transformative power of collaboration and human ingenuity.

Shagun Sharma is an independent business journalist.



Next Story