Advanced Computing in the Age of AI | Tuesday, July 23, 2024

Scale-Out Platform Emerges With Foxconn Backing 

Chinese electronics manufacturing giant Foxconn is among the early investors in the enterprise infrastructure platform developer DriveScale Inc.

Emerging from stealth mode this week, DriveScale Inc. announced a $15 million Series A funding round while releasing a "rack-scale" platform targeting big data workloads. The architecture is designed to help datacenter operators run data analytics applications with 100 to 1,000-plus Hadoop workloads together on commodity servers and storage infrastructure.

The startup's focus is squarely on the big data market, stressing that its priority is "addressing the elephant in the room: Scaling Hadoop and big data workloads." Company founders said they are banking on data analytics and the Internet of Things applications driving the big data market, fueling the need to scale out existing datacenter infrastructure.

DriveScale executives said they see an opening for addressing the problem of standard rack servers one of the few options for scale-out infrastructure. "This equipment was not originally designed for big data processing platforms such as Hadoop and massively parallel computing environments," the company noted in a statement.

DriveScale's cofounders played key roles in developing Cisco's Unified Computing Services (UCS) line along with Sun's UltraSparc workstations and workgroup servers. They developed the new platform over the last three years, claiming it can provide enterprises with a scale-out architecture along with lines of hyper-scalers like Facebook (NASDAQ: FB) and Google (NASDAQ: GOOGL).

While DriveScale joins a growing list of datacenter infrastructure vendors seeking to bring web-scale performance to enterprise datacenters, the startup also emphasizes its approach helps datacenter operators leverage their current commodity server infrastructure. The startup said its software-defined approach disaggregates existing storage and computing resources and recombines standard servers and commodity storage. The scale-out approach is billed as transforming datacenter infrastructure from standard rack server deployments to a more flexible platform better suited to handling Hadoop deployments as well as other applications workloads.

The scale-out platform also is touted as allowing datacenter administrators to deploy "independent pools of commodity compute and storage resources." A set of on-premises and software-as-a-service tools allow available datacenter resources to be reconfigured as needed to run big data workloads

The founders of DriveScale, Sunnyvale, Calif., are Tom Lyon, Satya Nishtala and Duanne Northcutt. Lyon, who also serves as DriveScale's chief scientist, helped develop Cisco System's (NASDAQ: CSCO) UCS architecture. Lyon also worked for Sun Microsystems.

Co-founder Nishtala, DriveScale's CTO, also helped develop UCS servers as well as Cisco's Nexus switching technology. Northcutt, who serves as a DriveScale vice president, previously worked for a company that was acquired by Sun Microsystems.

Along with Foxconn's (TPE: 2317) Ingrasys network computing unit, early investors in DriveScale include funding round leader Pelion Venture Partners and Nautilus Venture Partners. Among DriveScale's board advisors are Scott McNealy, co-founder and former CEO of Sun Microsystems, and Java programming pioneer James Gosling.

Besides Foxconn Ingrasys, early DriveScale customers AppNexus, ClearSense and DST Systems. Foxconn Ingrasys also co-developed DriveScale's product line and serves as the startup's hardware manufacturer.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).