https://devblogs.nvidia.com/building-accelerated-5g-cloudran-at-the-edge/

Mellanox ConnectX-6 Dx SmartNIC

exceeds stringent industry-standard timing specifications for eCPRI-based RANs by ensuring clock accuracy of 16ns or less

5T for 5G enables packet-based, ethernet RANs to provide precise time-stamping of packets for delivering highly accurate time references to 5G fronthaul and backhaul networks.

5T-for-5G, or time-triggered transmission technology for telco

https://news.developer.nvidia.com/new-real-time-smartnic-technology-5t-for-5g/

• Real-time transmission hardware acceleration: 5T-for-5G simplifies time synchronization and data transmission across servers, GPUs, radios, and baseband units in wireless network rollouts, making 5G rollouts easier and more efficient.

• Precise time stamping: It supports IEEE 1588v2 PTP for precise timing synchronization.

• Highest clock accuracy: It ensures extremely precise timing accuracy within 16 ns, adhering to the stringent ITU-T G8273.2 standard.

• eCPRI windowing: It supports precise transmission of eCPRI packets within the O-RAN specified timing window.

• ASAP2 time-based flow engine: It allows packet time stamping and time-bound packet steering using Mellanox ASAP2.

• Highly efficient: 5T-for-5G enables more accurate time synchronization than software solutions and lower costs and power consumption than traditional FPGA-based solutions, enabling efficient CloudRAN and edge rollouts.

• NVIDIA EGX–ready: Developers can create GPU-based CloudRAN solutions with the NVIDIA Aerial SDK that use precise timing without investing in expensive and proprietary FPGA solutions.

https://www.youtube.com/watch?v=rxLejojuMjM&feature=youtu.be

GPUDirect Remote Direct Memory Access (RDMA)

store data by eliminating unnecessary duplicates and decreasing latency. By reducing the back and forth with the CPU memory, GPUDirect saves computation cycles for memory access.

Header/data split functionality

With cuVNF, you have the header/data split functionality that allows agile packet filtering.

O-RAN format identification feature

does faster packet flow steering based on the MAC address

cuVNF

EGX A100 edge server platform

ConnectX-6 with 5T-for-5G

Links

• CUDA
• CUDA download and additional resources such as training, open source packages and etc
• https://news.developer.nvidia.com
• https://developer.nvidia.com/nvidia-aerial-sdk-early-access-program