NVIDIA Aerial SDK
Build and Deploy GPU-Accelerated 5G Virtual Radio Access Networks (vRAN)
NVIDIA Aerial™ is an application framework for building high-performance, software-defined, cloud-native 5G applications to address increasing consumer demand. Optimize your results with parallel processing on GPU for baseband signals and data flow.
Aerial SDK is in Early Access and available as a .zip package as well as a container image for easy deployment via NVIDIA NGC™.
Join Early Access Program Member Area
100% Software Defined
NVIDIA Aerial SDK is a highly-programmable PHY layer and has the capability to support L2+ functions seamlessly.
High Performance and AI Ready
With GPU-accelerated processing, complex computations run faster than existing L1 processing solutions. Faster implementation of L1 functions give improved performance results.
Scalable
100 percent commercial off-the-shelf (COTS) hardware support with Aerial SDK makes it easier to deploy cloud-native platforms such as NVIDIA EGX™. It’s Kubernetes based and provides container orchestration for ease of deployment and management.
NVIDIA Aerial SDK Stack
NVIDIA Aerial enables GPU-accelerated signal and data processing for 5G wireless RANs.
The NVIDIA Aerial SDK package simplifies building programmable and scalable software-defined 5G RAN. The Aerial SDK supports:
CUDA Baseband (cuBB)
The NVIDIA cuBB SDK provides a GPU-accelerated 5G signal processing pipeline, including cuPHY for Layer 1 5G PHY. It delivers unprecedented throughput and efficiency by keeping all physical layer processing within the high-performance GPU memory.
CUDA Virtual Network Functions (cuVNF)
The NVIDIA cuVNF SDK provides optimized input/output and packet processing, exchanging packets directly between GPU memory and GPUDirect®-capable NVIDIA ConnectX®-6 DX network interface cards.

Building on NVIDIA Aerial
There are two ways to get started with NVIDIA Aerial — the NVIDIA Aerial Developer Kit or building your own.
Designed to jumpstart performance evaluation and benchmarking for RAN development, the integrated reference design provides preinstalled test vectors under the NVIDIA Aerial SDK and high-compute and signal processing power with NVIDIA® GPUs and DPUs delivering an out-of-the-box guided experience. Key features include:
- Preinstalled Aerial SDK and software environment
- Test scripts that benchmark KPIs
- Logged results of developer kit configuration, including hardware, environment, SDK, and sample test execution
- Ability to exercise specific or all included test cases

The bring your own, bare metal installation provides flexibility to use your own hardware setup and installing Aerial SDK to implement 5G RAN functions.

Aerial vRAN partners


“SoftBank Corp. is committed to driving innovation in 5G networks, including by adopting softwarization and relying on centralized radio access networks, to improve and guarantee customers’ experiences. We believe that Nvidia’s GPUs and Aerial SDK will play an instrumental role in this effort and we welcome the results of our joint collaboration - the world’s first - that show outstanding performance in running vRAN networks using Nvidia’s GPU. Integrating vRANs with edge computing and the AI ecosystem is key to delivering high performance 5G experiences and creating innovative solutions for the telco industry.”
Ryuji Wakikawa, Vice President and Head of Advanced Technology Division, SoftBank Corp.
“The telco industry is eagerly adopting cloud-native architecture to meet the growing compute demands of 5G. We are learning firsthand how the remarkable compute performance of NVIDIA GPUs, together with NVIDIA’s Aerial SDKs, can address the challenges of building flexible, high-performance virtualized telecom networks. We look forward to Aerial’s continued development.”
Yasuyuki Nakajima, president and CEO, KDDI Research, Inc.
“The collaboration between OAI and NVIDIA enables us to provide an end-to-end 5G radio network gNB stack that’s a powerful and flexible tool for wireless innovation. This platform comprises the OAI 5G Layer-2/Layer-3 (L2/L3) radio stack interfacing with the NVIDIA Aerial GPU- and NIC-accelerated Layer-1 (L1) to deliver an end-to-end gNB solution complete with a 7.2 split radio unit. This will deliver the power of NVIDIA L1 and flexibility of the OAI open-source L2/L3 stack, allowing for showcasing exciting new 5G and xG use cases and technologies that are central to such future-generation networks.”
Irfan Ghauri, Director of Operations, OpenAirInterface (OAI) Software Alliance