Intel Corporation and Lightbits Laboratories have announced that they would work together to create innovative “disaggregated computing technologies” for data centers. This announcement came on Tuesday when the companies shared their objective of solving the problems faced by today’s data centre operators, who are eager to increase the total cost of ownership (TCO) due to stalled disc space and pandemic efficiency. The latter has created LightOS, a software-defined block storage network that pools NVMe storage across several storage locations into a high-performance virtual all-flash storage array.
In its blog post, Intel reported that the LightOSTM product provides high availability and read-and-write control designed to optimize the efficiency of flash-based storage. Unlike other NVMe storage systems, LightOS links NVMe media over industry-standard TCP networking, replacing the need for more advanced storage networks such as the Fiber Channel. Although completely designed for Intel hardware, LightOS provides consumers with dramatically enhanced storage density and eliminates under-use while retaining consistency with current networks without sacrificing capacity and convenience. Lightbits and Intel have shown 146μs of NVMe / TCP access latency using Intel 800 Series Ethernet network adapters with application system queues technologies. Also, the device can use Intel Optane Permanent Memory and Intel 3D NAND SSDs based on Intel QLC Technology, Intel Xeon Scalable Processors with special built-in artificial intelligence acceleration features other than Ethernet 800 network adapters. The problem with consumer-grade QLC is that although it provides high-performance readings, it slowly writes data and has problems when data is written to memory too many times.
LightOS finishes writing in a non-volatile buffer and then writes large chunks of data to QLC. Josh Goldenhar, Vice President of Product Marketing for Lightbits Labs, said, “The Company also expects to add data snapshot capabilities by the end of the year. Thus, through this collaboration, Intel aims to protect its business interests and maintain market dominance while benefiting from the massive increase in demand from GPUs to buy companies trying to accommodate their remote operations. According to Remi EL-Ouazzane, Vice President and Chief Executive Officer of Data Platforms Group Strategy and Business Development at Intel, the data center is being revamped, disaggregating, and composing the services required to fulfill the performance criteria expected to counter the data explosion.
