PCIe AI inference acceleration card M

compact

EdgeQ M Series PCIe In-Line Acceleration Card is the world’s first converged 4G, 5G and AI acceleration card for Open RAN applications. The M Series comes complete with in-line L1 acceleration function, virtualized RAN, and machine learning for the Distributed Unit (DU). The M Series offers telco-class, multi-carrier, maMIMO performance at 50% the power profile of O-RAN solutions today. EdgeQ M Series PCIe In-Line Acceleration Card is based on the world’s first programmable RISC-V based architecture, massively integrating hyperscale cloud functionalities into a compact SoC footprint. The M Series integrates complete 4G and 5G PHY processing with FEC acceleration, fronthaul and midhaul NIC with eCPRI, timing sync, and AI at an unprecedented low TCO. Multiple of these cards can be meshed together. The M Series supports non-massive MIMO, 32TRX and 64TRX massive MIMO for a variety of cell configurations. It also provides RAN virtualization, Multi-access Edge Computing, and Artificial Intelligence for cloud-native topologies, edge applications, and 5G RAN use cases Full Height Half Length, Single Slot PCIe Revision 4.0 EdgeQ TXU Processor (RISC-V Based) Compliant to O-RAN Split 6, 7.2x and 7.3 PCIe, Ethernet, eCPRI, Inter-Chip and More Multi Sectors with Up to 64T64R Massive MIMO Compatible with All Major Types of COTS Servers Concurrent Multi-RAT (4G, 5G), SA and NSA Modes Supported 5G Timing Sync Functions for all LLS Configurations Integrated Artificial Intelligence Multi-Access Edge Computing Network Function Virtualization Customer Programmable 4G and 5G PHY Processing