Infortrend expands edge computing portfolio for AI workloads

Infortrend Technology, Inc. said it has broadened its edge computing portfolio to address a wider range of edge AI deployments, from compact single-node systems to clustered configurations designed for resilience and scale. The updated lineup is organized by solution category and is aimed at organizations processing data where it is created, a model that supports edge AI inference, virtualization and container-based applications. The portfolio combines purpose-built hardware with fully integrated turnkey systems. Its bare-metal Edge Hardware is designed for organizations that want to run their own software stacks, while the turnkey offerings come preconfigured with an operating system and software for faster deployment. Those packaged systems include Standalone Edge for remote sites with limited IT staffing, HA Edge for high availability through a redundant two-node setup, and Advanced Edge for larger deployments using a 3- to 5-node cluster with automatic failover and dynamic resource scaling. Infortrend said the new offerings are built on two hardware series. The KS 3000U uses a single AMD EPYC 8004 processor and supports up to two GPUs for space-constrained environments. The KS 5000U supports dual Intel Xeon or AMD EPYC 9004/9005 processors, up to 320 CPU cores and as many as four GPUs for more compute-intensive workloads. Frank Lee, Senior Director of Product Planning at Infortrend Technology, said the expansion is intended to let businesses match infrastructure more closely to site-specific uptime and scale requirements.