See how Liqid works in these environments
AI workloads represent a data-intensive, powerfully uneven series of compute processes. Infrastructure requirements significantly differ, for example, between the NIC-centric data ingest phase versus the GPU-oriented data training phase. Traditional static data center architectures have required that individual systems be built to handle each phase. The data is then moved from one system to the next for data ingest, cleaning/tagging, training, and inference.
This approach leaves resources sitting idle while simultaneously increasing the need for physical real estate, heating, cooling and maintenance. Rather than move heavy data from system to system, Liqid has devised an approach to infrastructure capable of adapting the same resources to each phase in the AI workflow as is required.
Liqid delivers software-defined, composable resource allocation across bare metal via Liqid’s ultra-low-latency PCIe fabric to ensure significant increases in data agility, capacity and bandwidth. Leveraging pools of disaggregated GPUs, FPGAs, CPUs, and Intel® Optane™ memory extension technologies, IT users can compose balanced systems for each AI phases while minimizing the data center footprint.
Many organizations have begun to realize that the cost associated with cloud migration and maintenance can be prohibitively expensive. Others, such as financial and medical providers, want the dynamic “point and click” resource management associated with cloud environments but remain limited due to regulatory requirements. Mid-market cloud and co-location providers cannot afford the capital expenditure associated with developing bespoke, software-defined dynamic cloud.
Liqid composable infrastructure solutions provide dynamic cloud capabilities for on-premises environments and cloud service providers alike. Liqid composable infrastructure solutions can be deployed to maximize the performance of existing hardware footprints through software, in all data center environments, on-premises, in co-locations, and for cloud providers.
Liqid software enables IT users to pool and deploy resources on demand, enabling the flexibility to direct resources toward the application that requires them. Through a simple GUI, API, or command line interface, organizations can manipulate resources with the same ease as the “easy button” associated with the cloud experience. Aggregated GPUs, NICs, NVMe, FPGAs, CPUs, and Intel® Optane™ memory technology and deploy it on demand, in the exact amounts required, then release back to the pool for use by other applications as required.
The next big movement in data centers is taking place at the edge, and increasingly incorporates the speeds and huge capacities of 5G wideband technologies. With sensor data is taken from a wide variety of devices in the field, calculations must be performed in remote installations away from core data center operations to take advantage of extremely low data latency. These smaller footprints are further constrained by power, floor space, cooling, and human access.
Liqid composable infrastructure solutions enable better hardware utilization for low latency data performance at the edge. Powerful, AI-driven applications can better make data-intensive operations such as inference at the point closest to where data is being ingested, often in small facilities beneath 5G towers. Tightly pack CPU and networking hardware and add accelerators such as GPUs, NVMe, FPGAs, and Intel® Optane™ as required without the necessity of purchasing additional equipment.
Because Liqid Command Center software enables remote resource management, edge facilities can be controlled from core data centers or other locations. Hardware resources are pooled portioned out to meet the needs of individual applications and then freed for additional use with other applications once the task is completed.
Hyperconverged data center architectures can no longer scale to keep pace with the uneven demands of virtualized compute environments. With limited ability to disaggregate data center resources, hyperconverged systems can quickly become uneven, with some resources sitting idle while others are taxed to their limits. Without the ability to pool these high-value resources through software and adapt for changes in unpredictable, AI-driven workloads, the benefits of virtualization can be significantly diminished.
To address the challenges associated with the evolving virtualized data center, Liqid delivers composable, software-defined infrastructure solutions and services that automate, orchestrate and compose resources at the bare-metal level. IT users can significantly increase data agility, capacity and bandwidth in all environments for which lightning-fast, adaptive data manipulation is necessary.
Liqid composable infrastructure is hypervisor agnostic and optimized for VMWare’s VSAN, Dell VXRail, and other leading providers. The ability to create multiple servers on demand with composable software reduces the number of licenses required for virtualization deployments. Liqid composable software also ensures complete fault tolerance, instantly transferring data from failed compute nodes to other available disaggregated hardware.
Disaggregated pools of GPUs, FPGAs, CPUs, NVMe SSDs, and Intel Optane™ memory can now be added at will and utilized as shared resources in quantities that correspond to the needs of a given workload or task, while hypervisors such as VSAN can dictate resource allocation via Liqid APIs, extending the life of existing hardware and allowing IT organizations to grow infrastructure as required.
High-resolution digital Media & Entertainment (M&E) content production requires highly data-intensive applications at all phases of the production process. Creative professionals drive against tight deadlines to capture, render, and produce content that relies must process massive amounts of visual data at every step, from on set to your TV set.
Liqid composable NVMe accelerator solutions are helping drive needed efficiencies, increased productivity and reducing time to market for 4K+ episodic television, feature films, independent productions, video game and virtual reality applications, and more. Deployed in a studio production environment, NVMe gives M&E professionals the ability to edit raw 4K and 8K media workloads in real time, with 8x improvements in data transfers, 6x reductions in video- ingestion times; and, overall reduction in idle time on set and in post production.
Liqid composable infrastructure solutions can be utilized in tandem with composable NVMe to aggregate and deploy powerful visual computing technologies like GPUs, FPGAs, and Intel® Optane™ memory technology can be pooled in capacities beyond traditional hardware configurations, deployed on demand, and released for use by other applications when a task is complete.
M&E professionals can utilize Liqid Command Center software to manage every aspect of state-of-the-art, end-to-end, digital production studio/theater facilities, capable of all aspects of ultra-high definition content creation, from pre-visualization intake to post-production to screening. Liqid’s composable solutions maximize resource utilization, minimize footprint, and accelerate overall time to market for content.