Cray Partners With Markley to Bring Supercomputing to the Cloud

The vendor teams with cloud provider Markley to offer its Urika-GX appliance in an IaaS environment.

Cray cloud

Supercomputer vendor Cray will make its systems and technologies available to high-performance computing organizations, enterprises and other end users via the cloud.

The company is partnering with Boston-based cloud provider Markley to offer the supercomputing technologies as a hosted service, starting with the life sciences industry, Cray officials said May 16. The companies will expand the reach of the supercomputing capabilities to other verticals and will develop industry-specific solutions in the future.

The supercomputer-as-a-service offering is part of Cray’s larger push to expand its reach into the commercial and enterprise spaces as the demand for greater data analytics capabilities increases. During a conference call earlier this month regarding the company’s most recent quarterly financial numbers, President and CEO Peter Ungaro spoke about a slowdown in the high-performance computing (HPC) market that Cray traditionally has played in and the need to move into adjacent markets.

Ungaro said he still expects the high-end supercomputing space to rebound from its current slowdown and that the company will refresh its four product lines in preparation. But he added that Cray’s “second main focus area is to continue to expand in the commercial and big data markets. As data continues to explode, our government and commercial customers are looking for new ways to leverage that data into actionable business decisions.”

The company also is pursuing other ways of delivering its technologies to the market, including partnering with solutions providers, Ungaro said, according to a transcript on Seeking Alpha.

Through the partnership with Markley, Cray is able to make its supercomputer technologies available to users—from researchers and scientists to engineers and data scientists—who in the past may not have had the chance to leverage them, Fred Kohout, the vendor’s senior vice president of products and chief marketing officer, said in a statement. The demand is coming from a broad array of industries, including pharmaceuticals, aerospace, financial service and government agencies, company officials said.

With Markley, Cray is partnering with a 26-year-old infrastructure-as-a-service (IaaS) provider based in Boston that runs more than 3 million square feet of data center space in the United States and Europe. Bringing supercomputing capabilities to its lineup of services makes sense at a time when the demand for cloud computing, analytics and machine learning is growing, according to Markley CTO Patrick Gilmore.

“The need for supercomputers has never been greater,” Gilmore said in a statement. “For the life sciences industry especially, speed to market is critical. By making supercomputing and big data analytics available in a hosted model, Markley and Cray are providing organizations with the opportunity to reap significant benefits, both economically and operationally.”

The companies will first focus on the life sciences industry, with the initial solution including Cray’s Urika-GX, an integrated offering for big data analytics that includes hardware and software. According to Cray officials, the appliance takes advantage of open technologies and can run multiple workloads at the same time, including enterprise analytics software like Hadoop and Apache Spark. The Cray Graph Engine, which provides pattern-matching capabilities and is optimized for the high scalability and parallelization of the Urika-GX, also is integrated. By partnering with Markley, Cray is making its Urika-GX platform available either on-premises or via the cloud.

Cray introduced the Urika-GX last year, offering a system that is preloaded with Hortonworks Data Platform, which includes Hadoop, Spark and the OpenStack management suite. The vendor sells other analytics platforms, like the Urika-XA and Urika-GD, which are aimed at the HPC space. The Urika-GX targets the enterprise and the workloads they run.

The companies plan to make more Cray infrastructure available through the cloud, officials for both companies said.