Intel® Select Solutions for Genomics Analytics v2 access performance, scale, and ease of deployment for genomics insight and discovery.

View the solution brief ›

Configuration for Intel® Select Solutions for Genomics Analytics v2

Ingredient

Intel Select Solutions For Genomics Analytics

1 x Application Node

Platform

1 x Intel® Server Board S2600WFT; recommended, not required

Processor

2 x Intel® Xeon® Gold 6252 processor (24 cores, 2.10 GHz), or a higher number Intel® Xeon® Scalable processor

Memory

12 x 32 GB DDR4 2933 MHz 1DC (total capacity 384 GB or higher) per node

4 x Compute Nodes
Platform 1 x Intel Server Board S2600WFT

Processor

2 x Intel Xeon Gold 6252 processor (24 cores, 2.10 GHz), or a higher number Intel Xeon Scalable processor (per node)

Memory

12 x 32 GB DDR4 2933 MHz 1DC (total capacity 384 GB or higher) per node

Local Storage

2 x 480 GB Intel SSD DC S4510 (mirrored)

Storage

14 x 41.6 TB Intel SSD DC P4610, PCIe HHHL (per node)

Host Adapters

Required: Integrated 10 GbE

Optional: 1 x 100HFA016LS Intel® Omni-Path Host Fabric Interface (Intel® OP HFI) Adapter, PCIe x16

Data Network Adapter

Intel® Ethernet Connection X722 with Intel Ethernet Converged Network Connection X527-DA2/DA4

or 10 gigabit (Gb) Intel Ethernet Converged Network Adapter X710

Host Network Adapter Integrated 1 gigabit Ethernet (GbE)

Network Infrastructure

Management Network 1 x 10 Gbps 24x port switch

Intel Omni-Path Architecture

Optional: 1 x Intel® Omni-Path Edge Switch (Intel® OP Edge Switch) 100 Series or better

Storage Infrastructure

File System

Recommended, not required:

  • Bandwidth—200 MB/s per client
  • Capacity—120 TB per compute node; allows for 30-day sample storage.

For systems larger than 4-8 nodes, a parallel file system (e.g., Lustre) is recommended.

Software
Required Software:
  • GATK, BWA, and GATK workflows optimized for Intel® technologies
  • Optimized Cromwell workflow
  • Intel GKL with optimized routines for accelerating developer codes
  • SJob scheduler (e.g., Slurm) for running clustered analytics jobs
Optional Software:
  • Docker for running multiple jobs in isolated containers across a cluster
  • Apache Spark for big data analytics processing
  • Lustre, the open source parallel file system, for high-performance storage
  • GenomicsDB, specializing in large-scale variant analysis

Firmware And Software Optimizations

Recommended, not required:
  • Intel® Advanced Vector Extensions 512 (Intel® AVX-512)
  • Linux frequency governor: performance mode
  • Intel® Hyper-Threading Technology (Intel® HT Technology): enabled
  • CPU power and performance policy: balanced performance
  • Workload configuration: balanced
  • Package C-state: C6(Retention) state

  • Processor C6: enabled
  • Hardware P-states: native mode
  • Hardware PM interrupt: disabled
  • Enhanced Intel SpeedStep® Technology: enabled
  • Intel® Turbo Boost Technology: enabled
  • Energy-efficient turbo: enabled