Researchers from UNSW Sydney and the Garvan Institute of Medical Research have developed Slorado, the first open-source software/library solution for nanopore sequencing basecalling on AMD Graphic Processing Units (GPUs). Using Setonix, Australia’s most powerful supercomputer at the Pawsey Supercomputing Research Centre, this development enables researchers worldwide to process nanopore sequencing data on any mainstream GPU hardware for the first time, expanding computational options for the global bioinformatics community.
Oxford Nanopore Technologies (ONT) sequencers generate time-series raw signals that need to be converted into DNA bases through a process called basecalling. While ONT’s Dorado software supports NVIDIA GPUs for this process, AMD GPUs are currently not supported.
Led by Dr Hasindu Gamaarachchi and PhD candidate Bonson Wong, the team has created a streamlined version of the industry-standard Dorado basecaller that removes previous hardware limitations while maintaining and in some cases increasing performance. This innovation allows researchers to utilise AMD GPUs, including those in Pawsey’s Setonix, Australia’s fastest and most energy-efficient research supercomputer, reducing processing times and wait periods for critical genomic research.
Pawsey CEO Mark Stickells underscored the impact of this development:
“Dr Gamaarachchi’s work highlights our mission to accelerate scientific discovery. Fast, efficient bioinformatics is key to addressing challenges in medicine and biotechnology, from genomic analysis to drug discovery and personalised medicine. By enabling nanopore sequencing on Setonix’s AMD GPUs, we are breaking down computational barriers and empowering researchers with world-class tools,” Mark said.
“This collaboration demonstrates how state-of-the-art infrastructure can transform research workflows and broaden access to advanced computational tools.” He continued.
Dr Gamaarachchi and Bonson Wong shared their vision for Slorado:
“Our goal was to expand computational options for the bioinformatics community. By enabling basecalling on AMD GPUs and making the software fully open-source, we are empowering researchers to process their data more efficiently using both AMD and NVIDIA hardware.”
Slorado supports up to eight GPUs simultaneously on the same node, delivering unprecedented processing capabilities when combined with systems like Setonix. Researchers can now achieve faster results without enduring lengthy queue times, accelerating the pace of critical genomic discoveries.
The team access to Setonix has been supported via the Australian BioCommons Leadership Share (ABLeS).
To know more about Slorado, register and join us for an online event and discussion:
“Accelerating Nanopore Analysis: Introducing the new Open Source Slorado”
- Date and time: Tuesday, 18 March 2025
- Register here
The session will feature:
- Live demonstration of Slorado
- Q&A with the development team
- Real-world implementation cases
- Installation and optimisation guidelines
For technical details, read Dr Gamaarachchi ‘s Blog: https://pawsey.org.au/nanopore-basecalling-on-pawsey-supercomputer-using-amd-gpus/
About the Researchers
Hasindu Gamaarachchi
Hasindu Gamaarachchi is a lecturer at the School of Computer Science and Engineering, UNSW Sydney. He is also a visiting scientist in the Genomic Technologies Group at Garvan Institute of Medical Research. He completed his PhD in Computer Science and Engineering at UNSW Sydney in 2020. He completed his bachelor’s degree with first-class honours in Computer Engineering from the University of Peradeniya, Sri Lanka in 2015. His research focuses on the design, development and optimisation of bioinformatics software and hardware for real-time nanopore sequencing data analysis; and prototyping novel domain-specific computer systems for efficient genomics data analysis.
Bonson Wong
Bonson Wong is a first-year PhD candidate at the School of Computer Science and Engineering, UNSW Sydney. He is also a visiting student in the Genomic Technologies Group at Garvan Institute of Medical Research. During his bachelor’s degree, he worked as a research assistant at Garvan, focusing on developing and optimising software for the genomics field. After graduating in 2023, he immediately began his PhD at UNSW, applying his efforts to contribute hardware-accelerated solutions to long-read signal processing and analysis.
Dr Hasindu Gamaarachchi and Bonson Wong, the researchers behind Slorado, the new open source software for nanopore sequencing basecalling on AMD GPUs. Credits UNSW Sydney
Mark Stickells AM, Pawsey CEO
