SoftBank and EdgeCortix Partner to Jointly Realize Low-latency and Highly Energy-efficient 5G Wireless Accelerators
March 22, 2023 -- SoftBank Corp. ("SoftBank") and EdgeCortix Inc. ("EdgeCortix") announced they collaborated for technology development to improve the performance efficiency of wireless accelerators*1 by using EdgeCortix IP (intellectual property) solutions, which is characterized by high-speed processing and low power consumption applied to AI.
A network infrastructure that is scalable, highly reliable and energy-efficient is essential not only for Beyond 5G and 6G mobile communication systems, but also for the environment and the sustainable development of the telecommunications industry. To reduce overall network infrastructure cost and power consumption as well as offer the next generation of services, there is a growing need for a wireless base station virtualization technology that offloads computationally demanding Layer 1 High-PHY*2 processing to dedicated accelerators.
SoftBank and EdgeCortix have adapted EdgeCortix's DNA (Dynamic Neural Accelerator®) IP technology*3 to Layer 1 High-PHY processing and implemented low-density parity-check (LDPC) encoding*4. This is a type of wireless error correction code that has been deployed on FPGA boards. As a result, SoftBank and EdgeCortix achieved throughput of approximately 36Gbps, which is equivalent to almost 10 times the throughput compared to publicly available data of commercial IP solutions. By expanding the application of this IP technology, SoftBank will realize more energy-efficient and low-latency wireless accelerators. In the future, SoftBank will determine the feasibility of applying this technology to the entire High-PHY process to realize a radio accelerator with higher performance.
*1 Hardware specialized for specific processing that is difficult to run at high speed on a general-purpose processor (CPU) and that supports high-speed processing.
*2 The part of the physical communication layer (PHY) in wireless communications that performs modulation and error correction of digital signals.
*3 EdgeCortix's proprietary patented technology that excels in efficiently performing operations specific to Opportunistic Learning applications, especially neural networks.
*4 One of methods of wireless error correction code and used to transmit message on a communication channel with noise.
Encoder throughput measurement results (COTS: Commercial off-the-shelf)
* Measured with parameters defined in 3GPP TS 38.212 (BG=2, CodeBlockSize=3840, Zc=384, E=19200)
Comment from each company
“Improving the performance and the energy efficiency of our network infrastructure is a major challenge for the future” said Ryuji Wakikawa, Head of SoftBank Corp.’s Research Institute of Advanced Technology. “Our expectation of EdgeCortix is to be a partner who can provide both the IP and expertise that is needed to tackle these challenges simultaneously. We will continue to work with the EdgeCortix team to research the best accelerators to further improve performance and conserve power to deliver a high-quality network."
“The Open RAN acceleration opportunity with SoftBank Corp. was tailor-made for our Edge AI Platform and the Dynamic Neural Accelerator IP we’ve developed over the last three years,” said Sakyasingha Dasgupta, CEO and Founder of EdgeCortix, “Our IP was designed to be runtime-reconfigurable, while being agnostic to AI models and the inherent complex mathematics in them. We have developed a very nimble architectural design delivered through our key assets: MERA compiler and software framework, scalable neural network engines and the SAKURA system-on-chip architecture; we are able to customize our Platform to the unique needs at the intersection of AI and the 5G space. The performance we’ve achieved on our prototype system while collaborating with SoftBank’s R&D team to date has even surprised us. In the future, we will pursue adaptation to end-to-end 5G wireless accelerator solutions that span MEC and other systems.”
SoftBank will hold an event called "ギジュツノチカラ Advanced Tech Show 2023" on March 22 and 23, 2023 to showcase various initiatives of SoftBank's Research Institute of Advanced Technology, including the content of this press release. For more details about the event, please visit the event website. (*The event will be held in Japanese only)
Furthermore, readers are invited to visit the Research Institute of Advanced Technology's redesigned website to learn more about the various technology areas and initiatives that the institute is pursuing.
|
Related News
- ShortLink AB and Dolphin Design partner to create a highly energy-efficient Sub-GHz ASIC design platform
- EdgeCortix Acquires Multiple Patents for Dynamic Neural Accelerator AI Processor Technology
- IcyHeart EU project launch: Highly integrated ultra-low-power SoC solution for unobtrusive and energy-efficient wireless cardiac monitoring
- Cadence Tensilica HiFi DSP Enables Highly Energy-Efficient Audio Playback for Dolby Atmos for Cars
- Dolphin Design wins an Embedded Award for Tiny Raptor, its Energy-Efficient Neural Network AI Accelerator
Breaking News
- VeriSilicon Obtained Bluetooth 5.3 Certification for Its Complete Bluetooth Low Energy Solution
- UMC Reports Sales for May 2023
- Actions Technology Partners with PUFsecurity to Secure Wireless Bluetooth and IoT Applications
- Marquee Semiconductor Acquires Semikunn Technology Services, Expanding Design Services Portfolio
- GlobalFoundries and STMicroelectronics Finalize Agreement for New 300mm Semiconductor Manufacturing Facility in France
Most Popular
- Are Chiplets Enough to Save Moore's Law?
- GlobalFoundries and STMicroelectronics Finalize Agreement for New 300mm Semiconductor Manufacturing Facility in France
- Consortium's Move Will Boost RISC-V Ecosystem, Thankfully
- Andes Technology Showcases Pioneering RISC-V CPU IP Solutions at RISC-V Summit Europe
- Microchip Slashes Time to Innovation with Industry's Most Power-Efficient Mid-Range FPGA Industrial Edge Stack, More Core Library IP and Conversion Tools
![]() |
E-mail This Article | ![]() |
![]() |
Printer-Friendly Page |