APPLICATION INFRASTRUCTURE

MaxLinear Develops High Efficiency 5G Open RAN Radio Platform on Intel Agilex FPGA

MaxLinear | February 16, 2022

MaxLinear, Inc. announced its collaboration with Intel on a new reference design for next-generation 5G infrastructure Radio Units (RUs) that delivers industry-leading power amplifier (PA) performance and efficiency.

Combining Intel's high performance and low power Agilex FPGA family with MaxLinear RF transceivers and MaxLINTM ultra-wideband digital predistortion (DPD) solution allows developers to deliver new radio units rapidly and cost-effectively with pre-integrated and pre-tested cutting-edge linearization capabilities.

Radio Access Network (RAN) product developers are challenged to deliver quickly higher bandwidth, more compact, and lighter radios that support new 5G spectrum while simultaneously increasing energy efficiency. Implementing higher bandwidths at higher 5G RF frequencies also drives engineers to use new PA technologies such as Gallium Nitride (GaN) and Doherty architectures. To harness these exciting technologies in practical deployable radios, high-performance DPD signal processing is necessary to linearize the PA output. In addition, designers want a proven off-the-shelf solution for implementing an O-RAN Alliance compliant fronthaul interface. This joint solution addresses these challenges. Together MaxLinear and Intel are combining Intel® Agilex™ FPGAs and O-RAN interface IP with MaxLinear RF Transceivers and DPD IP to provide a complete programmable pre-verified software-defined radio (SDR) platform with the highest integration and lowest power consumption for macro, massive MIMO, and small cell applications.

“The Open RAN ecosystem will immediately benefit from our cooperation with industry leader Intel. Our RU reference platforms, featuring Intel Agilex FPGAs, will provide the most flexible programmable solution with the highest system performance available on the market for global Open RAN radio use cases,These high efficiency reference solutions will dramatically reduce the power consumed by 5G radio systems while simultaneously maximizing network performance."

Brendan Walsh, vice president of MaxLinear’s Wireless Infrastructure Group

“MaxLinear’s RF transceivers together with Intel FPGAs and IP deliver 2x performance per watt versus competing FPGAs and can offer a leading portfolio of radio solutions to our wireless systems customers allowing them to speed up time to market,” said Mike Fitton, vice president and general manager of the Intel Programmable Solutions Group's Network Business Division. “This is another strong example of industry collaboration – reducing time and complexity, while dramatically accelerating delivery of advanced 5G solutions.”

To allow developers to evaluate these solutions and speed up development, Intel and MaxLinear provide a range of development platforms and tools based upon the HiTek Systems Agilex eSOM module and carrier board. 
  • Intel's Agilex SoC FPGA devices use heterogeneous 3D system-in-package (SiP) technology to integrate Intel's first FPGA fabric built on 10 nm SuperFin Technology. Leveraging this advanced process technology and 2nd Gen Intel® Hyperflex™ FPGA Architecture enables these FPGAs to deliver ~2X better fabric performance per watt when compared to competitive 7 nm FPGA portfolios. Agilex SoC FPGAs offer an integrated quad-core Arm Cortex-A53 processor and a custom logic migration path from FPGA to structured eASIC for cost and power benefits in the high-volume production.
  • Intel's O-RAN, Fronthaul, and Low-PHY is a complete set of building blocks required to implement an O-RAN Alliance Split 7.2x compliant O-RU fronthaul interface in an Agilex FPGA. Intel also offers FlexRAN software stack, an O-RAN compliant Split 7.2x software covering High-PHY functionality for O-DU running on Intel Xeon processors.
  • MaxLinear's high-performance RF transceiver portfolio supports up to 400MHz instantaneous bandwidth (IBW) and includes the MxL16xx Quad-RF and MxL155x Octal-RF families. These transceivers are software compatible, creating a single platform solution that customers can leverage for radio applications, including macro, massive MIMO, and small cell.
  • MaxLinear's MaxLIN is the industry's leading DPD linearization solution. Its advanced machine learning algorithms exceed the 3rd Generation Partnership Project (3GPP) and Federal Communications Commission (FCC) unwanted emissions requirements with margin while delivering high PA efficiencies of >50%. This capability dramatically reduces power consumption for an 8-transceiver macro implementation by >10% compared with competitive DPD offerings.

About MaxLinear, Inc.
MaxLinear, Inc.is a leading provider of radio frequency (RF), analog, digital and mixed-signal integrated circuits for the connectivity and access, wired and wireless infrastructure, and industrial and multimarket applications. MaxLinear is headquartered in Carlsbad, California.

Spotlight

In this brief video you will learn how you can achieve 30% faster VDI on 30% less hardware with Turbonomic. This product demo walks through setting up dynamic VDI groups, virtual desktop SLA's, deployment reservations, planning, and more.

Spotlight

In this brief video you will learn how you can achieve 30% faster VDI on 30% less hardware with Turbonomic. This product demo walks through setting up dynamic VDI groups, virtual desktop SLA's, deployment reservations, planning, and more.

Related News

HYPER-CONVERGED INFRASTRUCTURE

TYAN showcases optimized HPC platforms at ISC 2022

Tyan Computer | May 31, 2022

TYAN, an industry-leading server platform design manufacturer and MiTAC Computing Technology Corporation subsidiary, will demonstrate its latest server platforms, which are powered by 3rd Gen Intel® Xeon® Scalable processors and optimized for HPC and AI markets, at ISC 2022 in Hamburg, Germany, from May 30th to June 1st, booth#D400. "The HPC market growth is heavily driven by new technologies and the increasing application of big data analytics. This is creating the need for better computing infrastructure to process large volumes of data available in the market. TYAN's HPC server platforms are based on 3rd Gen Intel Xeon Scalable processors which are foundational to help businesses and organizations accelerate the development of the next-generation HPC and AI system architectures." Danny Hsu, Vice President of MiTAC Computing Technology Corporation's Server Infrastructure Business Unit The Thunder HX FT83A-B7129 from TYAN is a 4U dual-socket supercomputer that can accommodate up to 10 high-performance GPU cards. The FT83A-B7129 platform provides great heterogeneous computing capacity for a variety of GPU-based scientific high-performance computing, AI training, inference, and deep learning applications, thanks to dual 3rd Gen Intel Xeon Scalable CPUs and 32 DDR4 DIMM slots. In addition, the system supports up to four U.2 NVMe devices in twelve 3.5-inch tool-less drive bays. The Thunder HX FT65T-B5642 is a 4U single-socket pedestal server designed for small HPC tasks that require a lot of processing power at the desk. A single 3rd Generation Intel Xeon Scalable processor, eight DDR4-3200 DIMM slots, eight 3.5-inch SATA, and two NVMe U.2 hot-swap, tool-less drive bays are included in the system. For GPUs to speed up HPC workloads, the FT65T-B5642 has two double-wide PCIe 4.0 x16 slots. The Thunder SX TS65-B7126 is a software-defined storage 2U dual-socket hybrid storage server. There are 16 DDR4-3200 DIMM slots, twelve 3.5-inch SATA hot-swap, tool-less drive bays with support for up to four NVMe U.2 devices, and two rear 2.5-inch hot-swap, tool-less SATA drive bays for boot drive deployment in this system. The Tempest CX S5560 is a Micro-ATX server motherboard with a single Intel Xeon E-2300 processor, four DDR4-3200 DIMM slots, three PCIe slots, up to eight SATA 6G connections, two NVMe M.2 slots, and dual 10GbE onboard LAN ports from TYAN. In 5G networks, the motherboard is excellent for multi-access edge computing servers and service access servers in a CSP environment.

Read More

HYPER-CONVERGED INFRASTRUCTURE

UbiStor Announces Infrastructure as a Service Partnership with Scale Computing

UbiStor | July 21, 2022

UbiStor, a globally-recognized Chicagoland-based Managed Services Provider, has announced its infrastructure partnership with Scale Computing. Scale Computing would be the newest provider to join UbiStor’s arsenal of Managed Backup and Disaster Recovery Solutions. Founded in 2001, UbiStor’s team of engineers have spent over two decades in the industry, specializing in Backup and Disaster Recovery as a Service. As data and threats to that data have become more complex, there has been an increasing need for expert MSPs familiar with integrating a variety of solutions depending on their customer’s environment, acting as a main point of contact during routine testing or in the event of disaster. UbiStor specializes in creating Backup and Disaster Recovery Roadmaps for customers with the ability to scale as they grow and weave new solutions into their evolving environment – all backed by their team of expert engineers on an ongoing basis. In 2021, UbiStor’s Managed SafeStor solution powered by Acronis quickly became one of their most-utilized services, but there was a clear desire from customers for an on-premises appliance with continuity across both platforms. The integration of the Acronis Cyber Protect Cloud and Scale Computing's Business Resilience System (BRS) Appliance gives UbiStor’s customers the agility of a hybrid cloud data protection service with the flexibility of instant local virtualization and recovery using the BRS hardware platform, completely changing how UbiStor brings Acronis services to the market. “The addition of Scale Computing to our Backup and DR-as-a-Service offerings enhances our ability to help customers meet their local restore and recoverability requirements, while still being able to leverage the advanced capabilities of the Acronis Cyber Protect Cloud.” Dan Hill, CEO, UbiStor About Scale Computing Scale Computing is a leader in edge computing, virtualization, and hyperconverged solutions. Using patented HyperCore™ technology, Scale Computing Platform automatically identifies, mitigates, and corrects infrastructure problems in real-time, enabling applications to achieve maximum uptime, even when local IT resources and staff are scarce. Edge Computing is the fastest growing area of IT infrastructure, and industry analysts have named Scale Computing an outperformer and leader in the space, including being named the #1 edge computing vendor by CRN. Scale Computing’s products are sold by thousands of value-added resellers, integrators, and service providers worldwide. When ease-of-use, high availability, and TCO matter, Scale Computing Platform is the ideal infrastructure platform.

Read More

IT SYSTEMS MANAGEMENT

Pure Storage redefines AI-ready infrastructure with AIRI/S built on NVIDIA DGX systems

Pure Storage | June 09, 2022

Pure Storage, the IT pioneer that provides the world's most sophisticated data storage technology and services, introduced AIRI/S, the next version of its comprehensive AI-ready infrastructure, created by Pure Storage and NVIDIA, at Pure/Accelerate techfest22. AIRI/S is a simple, on-demand platform that helps businesses accelerate AI initiatives at any scale. AI's capacity to drive innovation and gain a competitive edge with key business insights has been acknowledged by businesses across industries. However, the complexity and rigidity of legacy infrastructure technologies – both on-premises and in the cloud – is a hurdle for businesses hoping to realize the promise of AI-based data science. Moreover, the complexities of a "do-it-yourself" approach hinder too many AI ventures, resulting in considerable delays and idle time. Instead, in order to promote faster innovation and gather actionable, modern insights, AI necessitates an entirely new infrastructure. With the release of AIRI/S, powered by FlashBlade®/S, NVIDIA DGX A100 systems, and end-to-end networking based on NVIDIA Quantum and Spectrum networking platforms, Pure has taken the next step in its partnership with NVIDIA. AIRI/S provides enterprises with a proven, simple, and scalable architecture for all phases of the AI data pipeline. In addition, AIRI/S is the first solution to feature Pure's groundbreaking new FlashBlade/S, which offers significant improvements in non-disruptive performance, density, power economy, and scale, resulting in optimal alignment with DGX systems' industry-leading AI training performance and parallelism benefits. "Traditional approaches to AI infrastructure often result in silos of servers and storage that are either over-spent on capacity or starve AI workloads. With a focus on simplicity and scalability, AIRI//S enables global enterprises to achieve better time to insights and make the most out of their data with AI. We're incredibly excited to expand on our collaboration with NVIDIA to deliver the next-generation of AI-ready infrastructure." Amy Fowler, Vice President, Strategy and Solutions, FlashBlade, Pure Storage Customers can also take advantage of both always-modern infrastructure subscriptions and valid, consumption-based storage subscriptions by leveraging Evergreen/Forever (formerly Evergreen Gold) and Evergreen/One (formerly Pure as-a-Service) to manage their investments as they grow their AI infrastructure efficiently. "Enterprises are turning to AI to grow top-line results while enhancing customer experiences and reducing costs. These complex challenges require integrated, full-stack computing optimized for AI, and AIRI//S built on NVIDIA DGX systems provides customers with modular, high-performance enterprise infrastructure that scales easily as these demands continue to grow." – Charlie Boyle, Vice President, DGX systems, NVIDIA.

Read More