The Challenge
Plant Genomics Does Not Behave Like a Typical HPC Job
HudsonAlpha’s laboratory work generates large, complex genome data sets. New sequencing instruments increased the amount of data flowing into the computational pipeline, while plant genomes often require larger memory working sets than conventional simulation-oriented HPC nodes provide.
The team needed enough memory bandwidth and local capacity to keep large assemblies in RAM and enough processing resources to move each stage of the analysis pipeline efficiently.
WHY STANDARD NODES STRUGGLED
Large Working Sets + High Data Rates
The Solution
Intel Advanced Performance Servers, Staged and Validated by ADS
ADS recommended and staged a large-memory architecture so HudsonAlpha could test its real algorithms before committing to production systems.
COMPUTE
High Core Density
The deployed systems used Intel Xeon Platinum 9200-class processors, providing a dense CPU platform suitable for highly parallel genomic analysis.
MEMORY
High Memory Bandwidth
A wide memory-channel architecture helped keep large genome data resident and accessible to the analysis pipeline.
VALIDATION
Pre-Purchase Testing
ADS staged recommended systems in its integration facility so HudsonAlpha could verify performance and scientific correctness before deployment.
SUPPORT
Operational Continuity
Fast warranty handling and a single support path helped a research environment where keeping compute resources productive matters every day.
Impact
A Bigger Scientific Problem Space
The key outcome was not simply faster hardware. The new servers allowed HudsonAlpha to process jobs that previously had to be divided and reassembled, reducing very large assembly timelines and opening additional genomic work to practical analysis.
Fewer Splits
More of a genome can be processed as one large working set.
Faster Runs
Very large assembly workflows moved from month-scale toward day-scale execution.
Validated Results
Testing before purchase helped confirm both speed and reliability for production research.
Genomics Infrastructure
Need Compute Sized for Your Sequencing Pipeline?
ADS can model memory, CPU, storage, and data movement around your actual bioinformatics workflow.
Request a Consultation
Customer Story • Genomics
HudsonAlpha Genome Sequencing Center
Large-memory HPC for a 20× jump in sequencing data
Rapid growth in PacBio and Illumina sequencing output created a compute and memory challenge: HudsonAlpha’s plant-genomics workflows were becoming too large and complex for conventional HPC nodes to process efficiently.
ADS validated a large-memory Intel Advanced Performance server architecture against HudsonAlpha’s real computational work before purchase.
Request a Consultation
20×
Increase in sequencing analysis throughput demand
12
Intel Advanced Performance servers deployed
192
Total CPUs across the deployed servers
Months → Days
Reported reduction for very large assemblies
The Challenge
Plant Genomics Does Not Behave Like a Typical HPC Job
HudsonAlpha’s laboratory work generates large, complex genome data sets. New sequencing instruments increased the amount of data flowing into the computational pipeline, while plant genomes often require larger memory working sets than conventional simulation-oriented HPC nodes provide.
The team needed enough memory bandwidth and local capacity to keep large assemblies in RAM and enough processing resources to move each stage of the analysis pipeline efficiently.
WHY STANDARD NODES STRUGGLED
Large Working Sets + High Data Rates
The Solution
Intel Advanced Performance Servers, Staged and Validated by ADS
ADS recommended and staged a large-memory architecture so HudsonAlpha could test its real algorithms before committing to production systems.
COMPUTE
High Core Density
The deployed systems used Intel Xeon Platinum 9200-class processors, providing a dense CPU platform suitable for highly parallel genomic analysis.
MEMORY
High Memory Bandwidth
A wide memory-channel architecture helped keep large genome data resident and accessible to the analysis pipeline.
VALIDATION
Pre-Purchase Testing
ADS staged recommended systems in its integration facility so HudsonAlpha could verify performance and scientific correctness before deployment.
SUPPORT
Operational Continuity
Fast warranty handling and a single support path helped a research environment where keeping compute resources productive matters every day.
Impact
A Bigger Scientific Problem Space
The key outcome was not simply faster hardware. The new servers allowed HudsonAlpha to process jobs that previously had to be divided and reassembled, reducing very large assembly timelines and opening additional genomic work to practical analysis.
Fewer Splits
More of a genome can be processed as one large working set.
Faster Runs
Very large assembly workflows moved from month-scale toward day-scale execution.
Validated Results
Testing before purchase helped confirm both speed and reliability for production research.
Genomics Infrastructure
Need Compute Sized for Your Sequencing Pipeline?
ADS can model memory, CPU, storage, and data movement around your actual bioinformatics workflow.
Request a Consultation
The Challenge
Plant Genomics Does Not Behave Like a Typical HPC Job
HudsonAlpha’s laboratory work generates large, complex genome data sets. New sequencing instruments increased the amount of data flowing into the computational pipeline, while plant genomes often require larger memory working sets than conventional simulation-oriented HPC nodes provide.
The team needed enough memory bandwidth and local capacity to keep large assemblies in RAM and enough processing resources to move each stage of the analysis pipeline efficiently.
The Solution
Intel Advanced Performance Servers, Staged and Validated by ADS
ADS recommended and staged a large-memory architecture so HudsonAlpha could test its real algorithms before committing to production systems.
COMPUTE
High Core Density
The deployed systems used Intel Xeon Platinum 9200-class processors, providing a dense CPU platform suitable for highly parallel genomic analysis.
MEMORY
High Memory Bandwidth
A wide memory-channel architecture helped keep large genome data resident and accessible to the analysis pipeline.
VALIDATION
Pre-Purchase Testing
ADS staged recommended systems in its integration facility so HudsonAlpha could verify performance and scientific correctness before deployment.
SUPPORT
Operational Continuity
Fast warranty handling and a single support path helped a research environment where keeping compute resources productive matters every day.
Impact
A Bigger Scientific Problem Space
The key outcome was not simply faster hardware. The new servers allowed HudsonAlpha to process jobs that previously had to be divided and reassembled, reducing very large assembly timelines and opening additional genomic work to practical analysis.
Fewer Splits
More of a genome can be processed as one large working set.
Faster Runs
Very large assembly workflows moved from month-scale toward day-scale execution.
Validated Results
Testing before purchase helped confirm both speed and reliability for production research.
Genomics Infrastructure
Need Compute Sized for Your Sequencing Pipeline?
ADS can model memory, CPU, storage, and data movement around your actual bioinformatics workflow.
Request a Consultation
Applied Data Systems
©2026 Applied Data Systems
