TAIPEI, TAIWAN, December 7, 2021 – Leading provider of advanced network appliances and rugged industrial computers Lanner Inc., joined NVIDIA GTC 2021 event to demonstrate Edge AI Platforms for accelerating Open RAN programmable packet processing. During his speech, Sven Freundenfeld, Lanner’s CTO of Telecom Applications BU, broached the market trend concept of programmable centralized Open RAN and the Open Edge Server HTCA-E400.
With the aim of reducing CapEX and OpEX costs for service providers, Lanner developed a flexible and consolidated platform, combining GPU-accelerated multiple compute engines for fronthaul gateway, and true software-defined programmable switches for disaggregated cell site router. The inter-changeable, front access, and short-depth centralized unit has the ability to consolidate cell site router and fronthaul gateway functions to further accelerate 5G Open RAN deployment. Deploying a RAN system allows for the use of open COTS x86 infrastructure, which meets telco requirements, to broaden the choice of hardware at the micro site. Hardware acceleration can significantly enhance any Open RAN implementation, increasing raw performance, and reducing power consumption.
The pre-launched HTCA-E400, a carrier-grade Open Edge Server, powered by 3rd Gen Intel® Xeon® Scalable Processors (codenamed Ice Lake) with built-in AI acceleration, for 5G Open RAN, Edge Cloud, and AI edge-focused applications with extensive configuration options. HTCA-E400 is a high-performance appliance designed to leverage edge computing for accelerating 5G deployment. HTCA-E400 supports 5x 1U compute sleds, or 2x 2U+1U compute sleds, designed to support FPGA, GPU or Time Sync cards, making it suitable as an Edge AI server for Multi-access Edge Computing (MEC) servers, 5G Edge Cloud, and Central Units (CU) Open RAN platforms.