AMD Intuition GPUs Enhance DNA Sequencing with ROCm Acceleration

AMD has unveiled a major breakthrough in DNA alignment processing, leveraging its ROCm software program stack and Intuition MI300X GPUs. By accelerating key phases of the widely-used minimap2 genomics device, AMD claims to ship as much as 1.78x larger throughput in comparison with competing GPU options, significantly for long-read sequencing workloads.

DNA sequencing, as soon as a billion-dollar endeavor, is now a routine process because of advances in genomics and computational energy. Nonetheless, the bottleneck lies in processing the huge datasets generated. Trendy sequencing runs produce hundreds of thousands or billions of DNA fragments that should be aligned in opposition to a reference genome—a posh, computationally costly job. AMD’s newest implementation goals to sort out this scale by transferring probably the most intensive phases of minimap2—chaining and alignment—onto its GPUs, whereas sustaining compatibility with current CPU-based workflows.

Utilizing AMD’s HIP (Heterogeneous-Compute Interface for Portability) framework, the up to date minimap2 pipeline offloads these phases to the MI300X GPU. With reminiscence bandwidth of 5.3 TB/s and superior parallelism, the MI300X effectively processes massive genomic datasets. Early assessments present AMD’s GPU resolution outperforming rivals by 1.76x on PacBio CLR datasets and 1.6x on Oxford Nanopore workloads. These features convey DNA/RNA alignment nearer to real-time, a crucial step for scaling genomics in medication, agriculture, and analysis.

AMD’s ROCm platform, which underpins this growth, is gaining traction as a challenger to NVIDIA’s CUDA ecosystem. ROCm emphasizes open requirements and portability, attracting high-performance computing (HPC) and AI builders. With current releases like ROCm 7.14, AMD has expanded its ecosystem, integrating AI tooling and optimizing for broader workloads. The platform’s compatibility with AMD EPYC CPUs and Intuition GPUs positions it as a compelling selection for heterogeneous computing setups.

Past genomics, this announcement reinforces AMD’s broader technique to seize market share in GPU-accelerated HPC and AI workloads. Simply final week, AMD introduced a deal to provide Anthropic with Intuition MI450 GPUs, showcasing its deepening function in AI infrastructure. The ROCm ecosystem additionally continues to evolve, providing a CUDA-like programming mannequin by way of HIP, which simplifies porting current CUDA purposes to AMD {hardware}.

The genomics breakthrough displays AMD’s dedication to bridging real-world purposes with cutting-edge {hardware}. By accelerating minimap2, researchers can course of extra knowledge, sooner, enabling deeper insights into genetic mutations, pathogen monitoring, and inhabitants research. That is significantly crucial as sequencing prices plummet and knowledge volumes develop exponentially.

AMD is rolling out the GPU-accelerated minimap2 implementation as a part of a restricted early-access program. Researchers and builders can check the pipeline on their datasets, offering suggestions to fine-tune efficiency. The transfer additional solidifies AMD’s bid for dominance in GPU-accelerated genomics and different computationally demanding fields.

Picture supply: Shutterstock