en.Wedoany.com Reported - In the AI RAN space, Ericsson and Nokia have adopted different strategies: Nokia is deeply tied to Nvidia, while Ericsson leans more toward dedicated baseband solutions. Alok Shah, Vice President of Network Strategy, Business Development, Marketing, and Strategic Sales at Samsung Electronics America, explained that Samsung has taken a different, pragmatic approach. He noted that Samsung does not simply equate GPUs with AI RAN; the key distinction is that benefiting from AI RAN does not necessarily require GPUs, but if there is a need, an architecture capable of integrating GPUs on demand is essential. Shah cited an example: if a base station urgently needs performance improvements, deploying a GPU—despite its high cost and energy consumption—can still be a reasonable choice, as the performance gains from running more advanced algorithms offset the associated capital and operational expenditures. However, for many other base stations, this approach is currently not cost-effective. Therefore, Samsung aims to build an architecture that allows for flexible GPU integration when conditions are favorable.

Samsung is advancing its vRAN/Cloud RAN strategy with global operators. Verizon is its only major customer in North America, with deployments reaching tens of thousands of sites. Alok Shah pointed out that while Verizon's deployment scale is on an entirely different level, other operators such as Telus, KDDI, and Vodafone have also adopted Samsung's vRAN solution.

Samsung's collaboration with Nvidia spans over 20 years, initially starting with the supply of components such as memory. In the mobile network domain, the two companies have established a partnership in recent years, including benchmarking Samsung's vRAN software using Nvidia products. The testing covers the software's performance on CPUs, the performance of certain advanced algorithms on GPUs, and the evaluation of GPUs' ability to enhance both downlink and uplink throughput. Alok Shah stated that over the past five to six years, Samsung has maintained a leading position in software-driven architecture, vRAN, and disaggregated Open RAN, which has become a hallmark of the company. GPUs align very well with this architecture.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com