Life sciences research is evolving through neuroscience, genomics, and proteomics, relying on bioinformatics to predict complex biological behaviors. This data-intensive era requires the raw power of HPC. With AI now a primary engine, powering multi-omics discovery and protein folding revolutions, the Kunpeng HPC Solution stands at the forefront. By fusing diverse computing resources, Kunpeng drastically boosts R&D productivity and empowers the next generation of AI for Science (AI4S) scenarios.
Industry Overview
Application Scenarios
Genome Sequencing
Genome sequencing is the process of determining a full DNA sequence from blood or saliva to predict individual risk of diseases, and has been widely applied to medical and non-medical fields, such as prenatal screening, genetic testing, and targeted therapy. The high-throughput but low-cost sequencing technologies make genome sequencing affordable for common people.
Depending on the research purpose, genomic data analysis presents varied demands for computing instances, ranging from I/O-intensive to CPU-intensive and memory-intensive workloads. Currently, more than 50 applications have been ported to the Kunpeng platform, such as GATK, ont-tombo, and HISAT2.
Cryogenic Electron Microscopy (Cryo-EM)
Cryo-EM is a biophysics technique that captures the 3D fine structure of biological macromolecules. This is achieved by rapidly freezing the sample in vitreous ice and using a transmission electron cryo-microscope (TEM) to collect 2D projections for subsequent 3D reconstruction. However, the workflow presents extreme computational challenges: single jobs, ranging from image correction to 3D reconstruction, often consume several days and demand high-speed I/O for the quick ingestion of massive data files (hundreds of GB). To meet these demands, critical cryo-EM software (RELION, EMAN2) has been ported to the Kunpeng platform, ensuring better performance.