Genomics

Industry Challenge

Genomics sequencing is the mapping and correlation of genomes to gene reference sequences. It is fundamental to the diagnosis and cure of rare, inherited diseases as well as medical breakthroughs and personalized care. There are 3 billion human genomes, which translates to great volumes of data. The medical industry is progressing towards personalized medicine. This will require hardware with more processing power, storage capacity, and network bandwidth to keep up. With larger genomic data sets of roughly 90 gigabytes of data per patient, one can only imagine the computing and speed demands of genomic analysis.

Databases must expand beyond the current architectural limits for the efficient retrieval and processing of data. As the industry advances, we need scalable software and hardware resources to adapt to growing performance demands. Cost is also a concern, because traditionally, more computing power means more hardware components, translating to higher cost.

How can we harness technology to optimize performance to pioneer the path to precision medicine?

The Solution

The Acceleration Stack for Intel® Xeon® CPU with FPGAs on the Intel® PAC with Intel® Arria® 10 GX FPGA offers a programmable and transparent solution with low-latency, accelerated performance with simplified user experience across many industries. Specific to genomic workloads, Intel® FPGAs enable performance and latency optimization that involve large datasets.

Supporting the GATK, the Acceleration Stack for Intel® Xeon® CPU with FPGAs and Intel® PAC with Intel® Arria® 10 GX FPGA offers a powerful processing engine for people without deep hardware domain expertise. The performance, power, and reprogrammability make Intel® FPGAs a great solution to keep up with growing genomics sequencing demands. With our Acceleration Stack for Intel® Xeon® CPU with FPGAs and application programming interfaces (APIs), we enable hardware performance with familiar software frameworks.

Accelerate Genomics Research with Broad-Intel Genomics Stack* (BIGstack*) 2.0

Intel and the Broad Institute of MIT and Harvard have produced an integrated hardware and software solution to run some algorithms in Broad’s popular Genome Analysis Toolkit (GATK). The solution optimizes the most commonly used algorithm in the toolkit to compare sequences, PairHMM. It is the heaviest, most compute intensive workload in the pipeline, and therefore, typically a bottleneck in genomic research. The solution from Intel and the Broad Institute, BIGstack* 2.0, offers a PairHMM accelerated algorithm that can be loaded on the Intel® PAC with Intel® Arria® 10 GX FPGA to take advantage of high-performance computing power, memory, and storage with automatic performance scaling. The Intel® PAC with Intel® Arria® 10 GX FPGA can be customized to accelerate other algorithms in the GATK as well. Check out our resources below to learn more.