Speak With An Expert

Mail icon to contact Liqid about our composable infrastructure technology

Independently Scale Static Hardware with Liqid

Posted on
February 19, 2020
Written By

Converged and hyperconverged architectures have offered IT departments simplicity of deployment at a cost: in order to scale in these environments, a new converged rack must be deployed. Require NVMe flash, but don’t need the CPU or networking? Too bad. What was once an efficient solution for addressing the huge data demand associated with cloud-connected devices has created its own version of sprawl, with siloed resources sitting idle and systems shifting out of balance.  

One of the most compelling benefits of composable infrastructure is its ability to take better advantage of the most data-intensive applications on existing hardware while preparing for future, disaggregated growth.

A key factor that enables Liqid composable infrastructure to increase resource utilization is its ability to deliver software-defined disaggregation. Liqid Command Center software pools compute, storage, and networking, GPU, FPGA, and storage memory accelerators that can also be provisioned as needed. Physical resources are collected by device type in individual trays (just a bunch of GPU/flash storage/etc, i.e. JBOG, JBOF, etc) and deployed through software to meet the demands of each individual application.

For example, if you are running a machine learning algorithm from the hours of 9:00-6:00 that requires six GPUs to run in tandem with other quantities of NVMe, CPU, and networking, a server can be defined through software to meet those needs, with additional pooled resources being used to address other workloads. Once those machine learning tasks are completed for the day, the resources can be redirected in the desired quantities to address other tasks. A software defined management layer is used to discover, assign, and manage the pools of disaggregated data center devices, ensuring that the right resources are in the right place at the right time.

Because Liqid’s Command Center software is vendor- and application-agnostic, the provider of any given data center device is irrelevant. GPUs from NVIDIA and AMD, for example, can be pooled and deployed in tandem, providing for greater resource utilization. In addition, device life cycles can be extended and IT users can grow their infrastructure as required, rather than in a fashion dictated by arbitrary sales cycles.

Find out how to grow your disaggregated data center with Liqid composable infrastructure solutions. Download our new white paper, “Why Should Enterprises Move to a True Composable Infrastructure Solution,” today and begin mapping out your data center’s transformation free from vendor lock that can be detrimental to innovation.

Learn more about Liqid’s technology and contact us to discuss your specific data center needs with a representative.

Written by
Posted on
February 19, 2020

Would you like to learn more?

Speak with one of our sales experts to learn more about how we aim to deliver complete composability. For other inquiries, you can drop us a line. We'll get back to you as soon as possible.