Cerio has announced an industry breakthrough in composable infrastructure that moves PCIe systems beyond a single domain, delivering new ‘scale economics’ for AI and Cloud data centres.
A game-changer for availability and sustainability, this patent-pending technology provides an open, agile environment at the scale of the data centre for more efficient, dynamic access to any resource.
Using the new Cerio platform, data centres can programmatically control infrastructure, allowing unprecedented operational agility to manage system capacity, upgrades and repair. Cerio leverages deep roots in distributed architectures and multipath fabrics in the design of highly robust, heterogeneous systems.
“Every major inflexion point in computing is driven by the need for better economics, a better operational model and now, greater sustainability,” said Phil Harris, President and CEO of Cerio. “For the past couple of decades, we’ve been limited in how we build systems. No longer bound by a single PCIe domain, our customers can compose resources from anywhere in the data centre to any system.
“Pre-orders of the Cerio platform from hyperscaler, cloud service provider, enterprise and government organisations are a clear signal of the demand for a fundamentally new system architecture that is more commercially and environmentally sustainable,” added Harris.
Cerio is currently working with early-access customers in North America, Europe and Asia-Pacific on the implementation of critical use cases that include scalable GPU capacity and storage agility.
Using high-radix distributed switching, advanced multipathing and intelligent system adaptation of protocols across low-diameter topologies, the Cerio platform delivers PCIe Gen 5.0 agility, resiliency and heterogeneity across endpoints including GPUs, NVMe drives, DPUs, FPGAs and specialised accelerators. Full backwards compatibility offers support for new and older-generation PCIe resources.
“The Cerio platform is driving ground-breaking research into AI acceleration, to optimise the flow of data on a per-application basis,” said Dr Ryan Grant, Assistant Professor in the Department of Electrical and Computer Engineering at Queen’s University. “We’re using the unique multipathing capabilities of the Cerio fabric to optimise the precise calibrations of GPU selection, density and communications that will make traffic flows highly efficient and responsive in distributed, heterogeneous systems.”
Click below to share this article