Lightbits Targets 1 PB Storage With Just 256 GB RAM
New storage engine will dramatically reduce hardware costs, server requirements, and data center footprint amid mounting memory and flash supply constraints
Lightbits RAM Reduction circumvents this crisis by addressing a fundamental infrastructure bottleneck: the amount of system memory required to manage storage capacity.
As demand tightens global DRAM supply, component costs rise, and hardware lead times extend, coupled with NAND flash prices projected to climb over 30%, organizations running mission-critical workloads at scale face tremendous pressure. Lightbits RAM Reduction circumvents this crisis by addressing a fundamental infrastructure bottleneck: the amount of system memory required to manage storage capacity. By significantly reducing the DRAM-to-flash ratio, organizations can consolidate storage capacity into fewer, denser servers—reducing hardware CapEx, rack space, power consumption, and ongoing infrastructure costs.
Read more –Cloudtech Achieves the AWS Data Competency
Crucially, greater hardware efficiency doesn’t mean compromising performance. Lightbits SDS continues to deliver the high-performance, ultra-low-latency block storage that demanding cloud-native applications require, with latency <200µsec. Purpose-built for performance-intensive workloads running on Kubernetes, OpenShift Virtualization, OpenStack, Proxmox VE, and other cloud and virtualization platforms, Lightbits enables organizations to achieve the performance their applications demand within a significantly more cost-efficient infrastructure footprint.
RAM Reduction builds on Lightbits’ software-defined efficiency capabilities, including Intelligent Flash Management to extend SSD lifespans by up to 10x, Elastic Erasure Coding for advanced drive error handling, and support for mixed TLC and QLC SSDs across manufacturers.
Read more – Jamf Brings Apple Device Management and Security Apps to CrowdStrike Falcon Foundry
From the outside, Lightbits SDS remains unchanged: disaggregated, software-defined storage that pools NVMe SSDs across industry-standard servers and shares them over standard TCP/IP Ethernet using NVMe over TCP, delivering NVMe-oF’s low latency and high performance without specialized networking hardware. Storage scales independently of compute and provides enterprise data services, including volumes, snapshots, clones, cross-server replication, and compression. With RAM Reduction, the NVMe over TCP data path, data services, APIs, and tooling stay the same, and existing integrations require no changes.
Write to us editorial@demandmediaagency.com. to learn more about our exclusive editorial packages and programmes.