Cloud-scale data centers require a different approach
Cloud-scale data centers require disaggregation of storage and compute, as evidenced by the top cloud giants’ transition from inefficient Direct-Attached SSD architecture to low-latency shared NVMe flash architecture.
The Lightbits team, who were key contributors to the NVMe standard and among the originators of NVMe over Fabrics (NVMe-oF), now bring you their latest innovation: the Lightbits NVMe/TCP solution.
Split storage and compute, network & clients untouched
In stark contrast to other NVMe-oF solutions, the Lightbits NVMe/TCP solution separates storage and compute without touching the network infrastructure or data center clients. With NVMe/TCP, Lightbits delivers the same IOPS as direct-attached NVMe SSDs and up to a 50% reduction in tail latency.
The application teams won’t even notice the change
The transition from direct-attached SSDs to the Lightbits NVMe/TCP solution is so smooth your applications teams won’t even notice the change. Once they switch, they are freed to scale storage and compute independently, with consistently better user experience, and with lower overall costs. As the trailblazers in this field, the Lightbits solution is already successfully tested in industry-leading cloud data centers.