Marvell, Nokia upgrade ReefShark SoC chipset for 5G
Marvell, a semiconductor company, has extended its partnership with Nokia, a wireless connectivity solutions provider, to upgrade its co-developed ReefShark system-on-chip portfolio. The result of their collaboration is to deliver an advanced radio access and transport processing platform on Marvell’s 5nm Arm Neoverse N2 multi-core solution. This platform will help companies and mobile network operators innovate by leveraging new 5G services at high performance and low latency.
The Nokia ReefShark chipsets incorporate Marvell’s Octeon 10 data processing unit and are based on 3GPP 5G new radio specifications and are designed for 5G networks. The ReefShark chipset family consists of a digital front end for LTE and 5G radio systems, RFIC and transceivers, integrated transceivers and baseband processors.
“The extension of our strategic partnership with Marvell underscores Nokia’s commitment to support mobile network operators with the industry’s most advanced technology for their 5G radio access networks,” said Ari Kynaslahti, Nokia’s vice president and CTO of head of strategy and technology for mobile networks at Nokia.
Marvell’s Octeon 10 DPU is manufactured using 5nm process technology. The company claims the Arm Neoverse N2 cores deliver three times compute performance and 50 percent lower power. Incorporating Octeon 10 DPU allows Nokia to produce one of the most advanced 5G radio access and transport processing platforms, the company says.
Both companies recognize the essentiality of 5G wireless connectivity to meet ever-growing data traffic and performance needs while minimizing overall costs and energy consumption.
“We are proud of the accomplishments that our respective teams have cooperatively achieved and anticipate delivering further innovations based on the intersection of advanced technologies, expertise and market leadership this partnership brings together,” said Will Chu, the senior vice president for processors business group at Marvell.
Aetina unveils MXM module powered by Hailo-8 AI inference processor
Comments