Cray this week is taking the wraps off a new supercomputing platform dubbed the Urika-GX.
From a hardware perspective, the Urika-GX is a 42U single rack system that is powered by up to 1,728 Intel Xeon cores and up to 22 TB of DRAM memory. Storage scales up to 192 TB of local storage, and there are also options for attached storage to expand the capacity even further.
“The Urika-GX combines the best features of the Cray Urika-GD and the Cray Urika-XA, which are the two predecessors to the Urika-GX,” Cray spokesperson Nick Davis told ServerWatch.
Davis added that the Urika-GX will replace the Urika-GD and the Cray Urika-XA system. He noted that the Urika-GD was a graph discovery-based appliance focused on graph solutions for big data analytics, and the Urika-XA was a Hadoop- and Spark-based analytics appliance.
“The Cray Urika-GX does come pre-integrated and pre-tested with Hadoop, Spark and the Cray Graph Engine,” Davis said. “The openness of the Urika-GX system also gives customers the flexibility to run existing and emerging analytics workloads.”
Davis emphasized there is no vendor lock-in with the Cray Urika-GX system.
From an operating system standpoint, the system is powered by the Urika-GX Operating System Environment, which is made up of the open-source CentOS 7.2 Linux operating system, with additional support for the Cray Aries interconnect networking technology.
“With the Cray Urika-GX, we had quality score recalibration results from our Genome Analysis Toolkit (GATK4) Apache Spark pipeline in nine minutes instead of forty minutes,” Adam Kiezun, GATK4 Project Lead at the Broad Institute, said in a statement. “This highlights the potential to accelerate delivery of genomic insights to researchers who are making breakthroughs in the fight against disease.”