IDC predicts that worldwide data will grow 61% to 175 zettabytes by 2025, which will invariably put a strain on infrastructure as organizations look to sort, search, scan, and otherwise derive value from this data. That’s where NGD Systems comes in — it’s an Irvine, California-based company offering a computational storage drive tailor-made for on-device AI and machine workloads. To lay a runway for growth, NGD this week closed a $20 million series C financing round led by Western Digital Capital, with participation from existing investors including Orange Digital Ventures, Partech Ventures, BGV, and Plug-N-Play.
The round brings NGD’s total raised to date to $45 million following a $12.4 million series B in February 2018, and CEO Nader Salessi says the funding will enable NGD to execute its go-to-market strategy by supporting production as well as sales and marketing efforts.
“We are excited to have strong partners with the vision to see the value of computational storage,” Salessi said. Previously vice president of Western Digital’s SSD business unit, he cofounded NGD in 2013 with former Western Digital SSD system-on-chip director Vladimir Alves. “The continuous growth of massive data deployments, especially at the edge, requires low-power and large-capacity storage devices, coupled with the ability to drive more efficient in-place data analytics to take advantage of distributed and parallel processing.
NGD’s Newport Platform, which delivers up to 32TB of storage and packs a 14-nanometer multi-core Arm system-on-chip that consumes only 12 watts of power, optimizes performance with a particular focus on quality of service and data management. The storage units are designed to ensure only relevant data is handled at the host, with algorithms that manage power and flash memory while keeping latency to nine seconds or less.
Salessi claims that computational storage like the Newport Platform can offer a 20 times improvement in capability with a smaller footprint than competing solutions.
“Pioneering companies like NGD Systems are sending a clear message to the data storage industry — there is a better way to handle the petabytes of data that all organizations must contend with to process AI and ML workloads and that is Computational Storage,” said MIG Capital CEO Richard Merage, who plans to join the company’s board of directors. “This technology is a must for distributed and composable architectures like edge computing, and workloads to deliver on their potential, as it adds the processing power where it is needed.”
NGD’s competitors in the soon-to-be-$107.44 billion datacenter storage market include Eideticom, CNEX Labs, and Panasas, among others.