Jan 23, 2024 4 min read

Multi-Access Edge Computing (MEC)

Multi-Access Edge Computing (MEC)

Beyond 5G (B5G) : Multi-Tenant Private Networks, Powered by AMAX’s MEC, HPC & AI Servers

As 5G Networks continue to evolve and commercial 6G developments loom on the horizon, expected to deploy between 2030-2035, telecom providers require increasingly robust systems to meet the growing demands of data. The 5G Network and Beyond 5G (B5G) are expected to meet the massive demand for data rates, especially for extended reality (XR), augmented reality (AR) and mixed reality (MR). In order to meet these various requirements, the future new radio not only considers the sub-6GHz but also takes millimeter-wave (mmWave) band and terahertz (THz) band into implementation.  Therefore, base stations require more effective mechanisms for real-time interaction, using deep reinforcement learning (DRL), and the system capacity becomes challenging.

The emergence of private networks guarantees the delivery of voice, text, data, video, and connection to IoT devices, as well as integration with computing systems and machines.  Among the technical challenges, the SDN (software defined networking) and NFVi (network function virtualization infrastructure) data-driven network management are validated through execution in several scenarios with self-learning machine learning (ML) algorithms.

AMAX’s 5G/B5G platforms draw inspiration from Open-RAN (O-RAN), IEEE, ETSI, 6G-IA, 5FPPP, and Next G Alliance. These platforms efficiently integrate a wide range of advanced technologies with highly efficient compute power and storage to support the current demand of upcoming 5G services.  The O-RAN platforms process independently at the Remote, the Distributed and Central Units (RUs, DUs, CUs) placed in multiple locations supporting the functional splits.  Our Multi-access Edge Computing (MEC) gateways and High-Performance Computing (HPC) servers play an important role to further guarantee the capability of the O-RAN solution to support the stringent requirements associated with end-to-end latency and transport network capacity.

  • Network Intelligence: AI/ML will bring a major disruption of data processing, systems management, and network infrastructure. The end-to-end network automation requires superior computing performance and resources.

  • New Devices and Interfaces: Future Networks will be connected to multitudes of devices and interfaces, enabling human-machine and machine-machine communications.

  • Network of Networks: The B5G/6G networks cover multiple scales of virtual and physical networks.  e.g. Campus networks expand processes with requirements on connectivity, edge processing for further automation and Quality of Service (QoS).  And, the tailored network functionality requires to support the New Controlled Value Chains which including various type of resources – AI, data and communication processing.

To future-proof IT architecture and address today’s most complex computing challenges, AMAX® benchmarks real-world performance improvements that the latest technology innovations bring to today’s fiercely competitive data-driven landscape. Our latest product lineup excels at accelerating computations for the platforms most valuable to our partners and customers.

  • Intel® Deep Learning Boost (Intel® DL Boost) is embedded into Intel® Xeon® Scalable Processors and designed to improve performance for common AI inferencing and training workloads. The Intel® DL Boost includes Intel® AVX-512 VNNI (Vector Neural Network Instructions) which is an extension to the Intel® AVX-512 instruction set. The combination of three instructions into a single execution can further unleash computing potential.

  • NVIDIA H100 AI Datacenter GPUs are the massively parallel computing engines behind the Large Language Model (LLM) and generative AI revolution. When used in a DGX platform, it is an AI powerhouse, featuring 6x more performance and 2x faster networking than the previous generation. High-speed scalability features include NVIDIA Base Command and NVIDIA AI Enterprise software suite plus expert deployment advice from AMAX solution engineers.

  • AMD delivers a comprehensive product portfolio to support AI deployments from cloud to edge. AMD EPYC™ processors and AMD Instinct™ GPUs supercharge & optimize large scale training workloads and real-time inferencing to unleash the power of AI.

5G Edge / AIoT Computing Platforms

AE-252
2U Expandable Short Depth Edge Computing Server

AE-211
2U Multi Node Server Network Rack Cabinet Compliant

AE-352
3U GPGPU Short Depth Edge Computing Server

AE-200
High Efficiency Edge AIoT Box

AMAX 5G Edge Network Switches

AMAX’s L3 switches are ideal for service provider edge aggregation, enterprise wiring closets, data center aggregation, and network core deployment. These switches bring a high level of security and traffic control to the edge of your network. They provide high performance, resilient stacking, wire speed Layer 2 switching and Layer 3 routing, comprehensive QoS, and advanced security to deliver the scalability and resiliency to increase your company’s productivity while reducing operating costs.

A4600-28X/52X Series
10G L3 ETHERNET AGGREGATION TOR SWITCH

A5800-54X Series
10G TOR DATA CENTER SWITCH Top of Rack Switch

A7712-32X
100G L3 SWITCH WITH SDN CAPABILITY

Disaggregated and Open Network Solutions / 5G and Legacy Aggregation

AMAX offers end-to-end 5G telecom solutions for telecommunication companies (telecoms) to implement a next generation network architecture. Our solutions make us an integral part of our customer’s journey to prepare their network infrastructure for the surging data demand of 5G centric applications and services. Whether it is aggregating 5G and legacy data for the backhaul, increasing capacity for transport network or disaggregating the core network for more flexibility, AMAX provides the best of breed solutions for the network infrastructure from access to core.

AS9600-72XC
72-Port, 25/100GE Open Aggregation Router

AS9700-23D
400G Disaggregated Core and Edge Router