GigaIO and SourceCode today announced the launch of Gryf, the industry’s most powerful mobile AI platform for tactical edge environments, at GEOINT 2024. Billed as “Gryf, co-designed by GigaIO and SourceCode,” the system is the result of a unique technology collaboration between the companies that is designed to engineer complex customer solutions. GigaIO provides award-winning open workload-defined infrastructure for AI and accelerated computing, and SourceCode is a global provider of co-designed, compliant IT systems for next generation intelligent infrastructure.
Gryf sets a new standard for on-demand configurability in the field, and is unique in its ability to scale to the performance of a field supercomputer, all in the TSA-friendly form factor of carry-on luggage. This scaling is made possible by GigaIO’s groundbreaking FabreXTM AI memory fabric. FabreX can disaggregate and re-aggregate GPUs and other system components in order to enable optimal hardware configurations for each workload. For example, a single Gryf system can support server-class GPUs to process over a petabyte of information. This combination of power and portability led to the Gryf product name, which is derived from the mythical gryphon that combined the body of a lion with the head and wings of an eagle.
Gryf allows users to quickly transform the vast amounts of sensor data collected at the edge into actionable solutions. Prior to Gryf, data either had to be sent to a datacenter for processing and analysis, or be limited to the storage capacity that edge computing devices without GPUs can handle, severely limiting the ability to produce relevant, on-site results.
Highlights of the unique capabilities Gryf offers include:
- A singular, composable solution for a diverse set of mission needs. Gryf offers a single, adaptable solution for a variety of applications, ensuring users can meet changing field requirements. Fully configurable on-demand through software or by interchanging sleds, Gryf allows for dynamic reconfiguration, resulting in higher resource utilization, reduced hardware costs, higher availability, and lower power consumption.
- Maximum compute density for unparalleled AI performance at the tactical edge. Gryf sets a new standard for compute density in a portable format. By overcoming traditional cooling challenges through disaggregating power-hungry GPUs and providing for its own ventilation, Gryf is able to combine high-performance GPUs with substantial storage capacity (up to a petabyte) to allow users to execute complex AI tasks more quickly and effectively, directly at the operational site. Each Gryf chassis can accommodate a mix of six compute, accelerator, storage, or network sleds.
- Easily scale to meet diverse mission demands and boost operational responsiveness. Up to five Gryf units can be seamlessly interconnected using GigaIO’s FabreX AI memory fabric for processing of petabyte-size data sets. This configuration allows each Gryf unit to access and share sled components across all five connected chassis in an all-to-all topology.
- Easily transportable for field-missions. Designed for true mobility, Gryf features a rugged, roll-on TSA-friendly form factor that fits into an overhead bin to allow for deployment at any location.
“With AI becoming more pervasive, there is a growing segment of the edge market clamoring for a meaningful amount of accelerated compute locally,” said Alan Benjamin, CEO of GigaIO. “This is true for our Department of Defense customers, who have emphasized the critical need for timely and actionable intelligence in the field. Gryf’s novel architecture, made possible by FabreX, our AI memory fabric, provides those customers the advanced compute, storage, and GPU capabilities they crave in today’s sensor-rich edge environments.”
“Engineering this level of power into such a compact form factor for mission-critical applications is a massive achievement,” said Parag Shah, CTO/COO of Edge and Embedded AI, SourceCode. “This is possible due to a deep understanding of customers’ needs, decades of mission-focused engineering experience, and the collaborative power of co-design.”
Complex Needs Demand a Co-Design Approach
The co-design approach redefines the solution creation process by bringing the customer together with technology experts and best-of-breed products to build and deploy the optimal solution. The Gryf co-design partnership marries GigaIO’s deep expertise in interconnects and composable infrastructure solutions with SourceCode’s ruggedized system design and co-design manufacturing.
Gryf is available now for purchase from both GigaIO and SourceCode, with delivery in Q3; early access units configured for intelligence community workloads are scheduled for June. Gryf is manufactured in SourceCode’s US-based ITAR compliant co-design manufacturing facility, part of its global network of manufacturing sites. Visit SourceCode’s booth (#109) at GEOINT 2024 in Kissimmee, May 6-9 for more.
About GigaIO
GigaIO provides workload-defined infrastructure through its AI memory fabric, FabreX, which seamlessly composes rack-scale resources and integrates natively into industry-standard tools. The SuperNODE is an “impossible server,” fully engineered to “Just Work” for AI and accelerated computing. Our solutions allow users to deploy systems in hours instead of months. Visit www.gigaio.com or follow X / LinkedIn.
About SourceCode
Founded in 1992, SourceCode is a global provider of co-designed, custom, certified IT systems for next generation intelligent infrastructure. The company’s edge devices, desktop, servers, storage, clusters, and reference architectures are available via its industry-leading ecommerce platform as well as its customer-centered engineering-led engagement model. Learn more at www.SourceCode.com.
View source version on businesswire.com: https://www.businesswire.com/news/home/20240506933173/en/