Recently, Nvidia and Softbank announced that they have developed a dual-purpose, AI-powered 5G MEC and vRAN distributed platform based on Nvidia’s new GH200 Grace Hopper superchip. Both partners plan to roll out a network of regional data centers across Japan later this year to capitalize on demand for accelerated computing and generative AI services. The shared multi-tenant platform also offers a variety of 5G vRAN applications, with SoftBank developing his 5G applications for autonomous driving, AI factories, augmented and virtual reality, computer vision and digital twins.
Key Point 1: Platform Limits
To take advantage of the projected surge in demand for AI processing power, SoftBank offers a dual-purpose platform whose primary application is AI computing via a high-performance edge analytics platform. The company is also developing various 5G applications, but AI and 5G workloads will be delivered at the same time. Counterpoint Research believes this platform is unlikely to be viable for vRAN workloads alone, and understands that both partners are not targeting this market.
Key Takeaway 2: Leveraging AI and GPUs in RAN
Nvidia plans to leverage RAN’s GPU processing power in a number of ways, including increasing spectral efficiency. One way he does this is by applying AI to optimize channel estimation feedback data between user devices and base stations. mMIMO radios have a computationally intensive problem, and using AI to compress receiver feedback data reduces signaling overhead, which effectively increases uplink channel capacity. This can be especially effective at cell edges. Nvidia claims that using the GPU in this manner can boost the gain for cell edge users by 14-17 dB. Other applications include using AI to optimize mmWave beamforming management and accelerate layer 2 scheduling.
The full version of this Insights Report, including the complete set of key takeaways, is published in the following report and is available to Counterpoint Research clients. 5G network infrastructure services (5GNI).
Nvidia and Softbank Partner to Deploy AI-Based 5G MEC Communications Network Across Japan
table of contents:
snap shot
major highlights
-5G MEC communication network
– Grace Hopper Super Chip
– Utilization of software resources
– Performance details
– Use cases and deployment options
-Main partner
– Competitors
Analyst Perspective
– Platform restrictions
-Benefits of in-cloud RAN
– Elimination of RAN hardware dependency
– Intel vs ARM battle
– Leveraging GPUs in RAN
– Crowded Market
Related reports and blogs
Open RAN Networks – Layer 1 Acceleration Strategy Key to Carrier Success
Cloud RAN Platforms – Why Are Vendors Adopting Different Layer 1 Acceleration Strategies?
New Layer 1 Accelerator Cards Open Up RAN Market – Or Create More Lock-in?
