Wednesday, March 15, 2023

Broadcom begins volume shipment of 51.2 T Tomahawk 5

Broadcom has begun volume shipments of its Tomahawk 5 family of Ethernet switch/router chips.

Tomahawk 5 offers 51.2 Tbps of switching capacity, twice that of any other switch on the market. The high radix of Tomahawk 5 enables single-hop connectivity between 256 high-performance AI/ML accelerators, each having 200Gbps of network bandwidth. 

Broadcom says Tomahawk 5’s advanced Cognitive Routing ensures high network utilization even under the most demanding workloads. The Cognitive Routing suite includes innovative features such as Global Load Balancing which uses distributed inter-switch communication of congestion information to choose the best global path for packets through the network. Cognitive Routing also provides high network resiliency with extremely fast link-failover by detecting a failed link and re-directing traffic in less than 500ns.

“Taking Tomahawk 5 to production in a mere seven months after the product was initially sampled is a testament to Broadcom’s execution quality at scale,” said Ram Velaga, senior vice president and general manager, Core Switching Group, Broadcom. “Innovation and consistent execution enables us to be the trusted technology partner for our customers’ demanding networking needs.”

Broadcom delivers 51.2 Tbps Tomahawk 5

Broadcom has begun customer shipments of its StrataXG Tomahawk 5 switch series, providing 51.2 Tbps of Ethernet switching capacity in a single, monolithic chip.

“Delivering the world’s first 51.2 Tbps switch two years after we released Tomahawk 4, the industry’s first 25 Tbps switch, is a testament to the outstanding execution and innovation by the Broadcom team,” said Ram Velaga, senior vice president and general manager, Core Switching Group, Broadcom. “Since the introduction of Tomahawk 1 in 2014, Broadcom has consistently executed on doubling the bandwidth approximately every two years. With today’s introduction of the fifth generation Tomahawk family, we are proud to say that a single Tomahawk 5 replaces forty-eight Tomahawk 1 switches in the network, resulting in over 95 percent reduction in power requirements. We applaud our customers, partners, and engineers for making this possible.”

StrataXGS Tomahawk 5 highlights:

  • 64 ports of 800GbE switching and routing.
  • Virtualization of general compute and AI/ML workloads with single-pass VxLAN routing and bridging.
  • Unparalleled physical I/O options using 512 instances of 100G PAM4 SerDes.
  • High-precision PTP and SyncE time synchronization.
  • Six on-chip ARM processors for high-bandwidth, fully-programmable streaming telemetry, and sophisticated embedded applications such as on-chip statistics summarization.
  • Implemented as a monolithic 5nm die

StrataXGS Tomahawk 5 Series Key Features for AI/ML:

  • World’s highest radix of 200GbE ports: 256 ports supported on a single chip, enabling flat, low latency AI/ML clusters.
  • Advanced 51.2 Tbps shared-buffer architecture, providing the highest performance and lowest tail latency for RoCEv2 and other new RDMA protocols.
  • Advanced Broadcom Cognitive Routing, dynamic load balancing, and support for end-to-end congestion control capabilities specifically designed to handle the large, low entropy flows typical of AI/ML workloads.
  • Support for Clos and non-Clos topologies such as torus, Dragonfly, Dragonfly+, and Megafly.
  • Hardware-based link failover

Broadcom said it is targetting the Tomahawk 5 family at the massively shared infrastructure in large data centers, Tomahawk 5 provides AI/ML workload virtualization with features such as single-pass VxLAN routing and bridging. Critical to minimizing job completion time (JCT) for AI/ML workloads, Tomahawk 5 offers features such as Broadcom Cognitive Routing, advanced shared packet buffering, programmable inband telemetry, and hardware-based link failover.

Tomahawk 5’s Cognitive Routing improves network link utilization by automatically and dynamically selecting the most lightly loaded links in the system for every flow that transverses the switch. This is especially important for AI/ML workloads which often have a combination of short-lived mice flows and long-lived, high bandwidth elephant flows with low entropy. 

Tomahawk 5 includes real-time dynamic load balancing that tracks the utilization of all links, both at the switch and downstream in the network, to determine the optimal path for each flow. It also monitors the health of links in hardware and automatically steers traffic away from failed links. These features provide dramatically improved network utilization and reduced congestion, resulting in shortened JCT.

Tomahawk 5 also features extensive programmable inband telemetry on both live traffic and network probes. Real-time metadata can be inserted into traffic at line rate as it traverses the network to collect telemetry on queue size, packet latency, switch utilization, and a variety of other customer-selectable metrics. This metadata can be used for precise end-to-end network congestion control.

To enable the lowest power and lowest cost for physical connectivity, Tomahawk 5 enables a direct 100G PAM4 interface to direct attach copper (DAC), front panel pluggable optics, and co-packaged optics. The flexible, long reach Tomahawk 5 SerDes provides DAC connectivity to all devices within a rack, and even between racks, without the need for retimers or other active components. It can also interface directly to a broad ecosystem of standard front-panel pluggable optical modules.

Tomahawk 5 will be made available with co-packaged optics using Broadcom’s Silicon Photonics Chiplets in Package (SCIP) platform, providing more than 50 percent decrease in the power needed for optical connectivity. Since the same switch silicon provides all these options, customers can choose the optimal I/O for each part of their intra-cluster, inter-cluster, and inter-DC networks with no software porting required.

NTT Electronics releases 140 Gbaud coherent DSP

NTT Electronics announced commercial shipments of a 140 Gbaud high-performance coherent DSP (ExaSPEED GAIA)for enabling 1.2 Tbps per wavelength modules. The new DSP offers a programmable baud rate of up to 140 Gbaud powered by a flexible coded modulation scheme and advanced 5nm FinFET CMOS process.

The programmable baud rate can support very fine granularity that enables ExaSPEED GAIA to adapt to a range of links while maximizing spectral efficiency. These include short-reach data center interconnects (DCI), terrestrial metro links with cascaded ROADMs, and ultra-long-haul sub-sea links. Customers can use a wide range of 400G interfaces in 1.2T (3x400GbE) for DCI, 800G (2x400GbE) for metro and 400G (1x400GbE) for sub-sea applications. This capability helps simplify interfaces and optimize Capital/Operational expenditure. Moreover, ExaSPEED GAIA also offers fine capacity tuning in 50G increments to respond to various customer requirements.

Flexible coded modulation is the newly-developed coding scheme to efficiently implement advanced constellation shaping, forward error correction (FEC), and high-order modulation format with high performance and low power. This enables ExaSPEED GAIA to support thousands of operational modes that cover a multitude of customer's transmission links with high energy efficiency. To enhance the performance, ExaSPEED GAIA employs a number of novel compensation algorithms including nonlinear distortion compensation, adaptive fast state of polarization (SOP) tracking, fast device and transmission characteristics estimation, and ultra-long CD compensation for high-baud rate systems*. The performance of ExaSPEED GAIA can be further optimized by combining it with NTT Electronics’ industry-leading high speed coherent driver module (UHB-CDM).

“We are excited to add the ExaSPEED GAIA DSP to our product portfolio which cuts the power consumption per bit to half while doubling the baud rate over earlier product offering,” said Satoshi Shigematsu, NTT Electronics General Manager of Broadband System & Device Business Unit. “We believe this high-performance DSP will help our customers offer highly differentiated optical transmission equipment and systems”, he added.

“We are very impressed with the performance of NTT Electronics’ latest ExaSPEED GAIA DSP”, commented Hideki Matsui, Vice Head of Photonics System Business Unit, Fujitsu Limited. “This revolutionary DSP makes our products more competitive in the market and enables us to offer innovative solutions. It opens up a new ‘terabit era’, which will be a game changer in our industry”, he added.

Hawaiki achieves record spectral efficiency with Infinera

Hawaiki Submarine Cable achieved recrod subsea spectral efficiency of more than 4.5 bits per second per hertz (b/s/Hz) on an uncompensated cable spanning 15,000 km and connecting Australia and the United States in a trial of Infinera's ICE6 solution.

Conducted with Infinera’s GX Series Compact Modular Platform, featuring the ICE6 optical engine, the trial underscored the value of its unique features, including ultra-high baud rates, long-codeword probabilistic constellation shaping (LC-PCS), second-generation Nyquist subcarriers, and SD-FEC gain sharing.

Hawaiki also tested the integration of Infinera’s suite of software automation tools, including Intelligent Power Management (IPM) and Instant Bandwidth 2.0, which enable important capabilities such as smart cable automation, embedded spectrum sharing, and rapid service activation.

“As a carrier and vendor-neutral operator, we are consistently looking for the best technology to upgrade our network. This trial leveraging Infinera’s latest coherent optical solution proved that the Hawaiki cable system has been optimally designed and is capable of achieving record-breaking performance to deliver more capacity and open spectrum sharing capabilities, essential to meet our customers’ needs,” said AnaĆ«l Greiveldinger, Hawaiki’s Director of Engineering.

“Extending the capital life of important subsea assets for our customers remains a key goal of Infinera,” said Nick Walden, Senior Vice President, Worldwide Sales, Infinera. “We are excited about the results of this trial proving how ICE6 can significantly increase the number of services that can be delivered over long submarine infrastructure.”

O-RAN Alliance brings enhancements to RAN Slicing

The O-RAN Alliance and OpenAirInterface Software Alliance (OAI) agreed to co-operate on open Radio Access Network issues, open source software development, and 5G as well as next generation platforms for demonstrating O-RAN technology.

The O-RAN Alliance also announced two new Open Testing and Integration Centres (OTIC) in Asia & Pacific. OTICs are vendor-independent, open, and qualified labs approved by the O-RAN ALLIANCE that issue awards in the O-RAN Certification and Badging Program. These include the Asia & Pacific OTIC by ritt7layers and Asia & Pacific OTIC in Singapore.

O-RAN Release 003 includes enhancements to RAN Slicing and SMO features. It also introduces new features:

  • Security requirements and countermeasures for RAN Intelligent Controller (RIC) interfaces and apps, O-Cloud and O2 interface
  • Security technical reports with threat and risk analysis outputs for application lifecycle management, log management, service management and orchestration, shared O-RU
  • RIC-enabled massive MIMO optimization for non-Grid of Beams (GoB) beamforming method
  • RAN Analytics Information Exposure (RAIE) use cases and RIC architecture enhancements for extending RAN-aware optimization capabilities to external services and applications

Australia Japan Cable tests Infinera's advanced coherent algorithms

Australia Japan Cable (AJC) is the first to trial Infinera’s advanced coherent algorithms for next-generation optical engines.

The trial was conducted over a 7,289-km segment of the AJC cable system between Paddington, New South Wales, Australia and Tumon Bay, Guam. The trial delivered a 17% increase in total fiber pair capacity compared to the current generation of optical engine deployed in AJC’s network. This trial leveraged Infinera’s advanced development testbed used for comparative algorithm testing, which included a high-performance arbitrary waveform generator-based transmitter, a receiver based on high-performance photodiodes, and a real-time oscilloscope. The trial tested advanced modulation, forward error correction (FEC), and nonlinear compensation algorithms.

The AJC cable system was brought into service in 2001 with a design life of greater than 25 years. Its design predates the introduction of coherent transponders by almost a decade. Capacity growth in the Pacific region is around 35-40% per year, making the increase of capacity on submarine cables more critical than ever.

“With the enormous cost and the delay of several years to gain approval permits for new cable systems, it is important to find ways to extend the economic life of existing cables like ours,” said David Crofts, CEO, Australia Japan Cable. “Older submarine cables tend to exhibit higher optical penalties compared to newer cables, but by deploying advanced modulation and compensation techniques, it is still possible to squeeze additional capacity from these valuable assets, as this trial demonstrates.”

“Collaboration with key network operators, like AJC, helps us to bring our innovative concepts from the lab to real-world environments and is a critical part of Infinera’s continuous drive to increase the value delivered by optical transmission solutions,” said Robert Maher, Infinera CTO, Optical Modules & Coherent Solutions. “The more capacity we can squeeze out of deployed fiber optic cables using advanced transponder designs and innovative signal processing algorithms, the more we can help to extend the economic life of mature cables like AJC.”

Dell'Oro: Broadband equipment revenue up 17% in 2022

Total global revenue for the Broadband Access equipment market increased to $19 B in 2022, up 17 percent year-over-year (Y/Y), according to a new repor from Dell'Oro Group. Spending on PON equipment continues to fuel the overall market, with revenue for PON OLTs and ONTs reaching $11.7 B for the year.

"The focus on upgrading and expanding broadband access networks was a clear focus for operators around the world in 2022," said Jeff Heynen, Vice President with Dell'Oro Group. "We expect broadband to remain a critical area of investment in 2023, as operators expand their footprint and increase the speeds and services they offer subscribers," explained Heynen.

Additional highlights from the 4Q 2022 Broadband Access and Home Networking quarterly report:

  • Total XGS-PON OLT port shipments exceeded 2 M for the year, reaching a total of 2.3 M worldwide.
  • Total cable access concentrator revenue was up 1 percent Y/Y at $1 B. Remote PHY devices, Remote OLTs, and Virtual CMTS platforms all recorded significant revenue gains for the year, as cable operators continue to expand their DAA and fiber initiatives.

T-Mobile US to acquire Mint Mobile, an MVNO

T-Mobile US agreed to acquire Ka’ena Corporation and its subsidiaries and brands: Mint Mobile, a successful direct-to-consumer (D2C) prepaid wireless brand in the U.S.; Ultra Mobile, a unique wireless service offering international calling options to communities across the country; and wholesaler Plum. 

T-Mobile will pay up to a maximum of $1.35 billion in a combination of 39% cash and 61% stock to acquire Ka’ena. The final price depends on Ka'ena's performance.

Ka'ena has been using T-Mobile's 5G network exclusively.

“Mint has built an incredibly successful digital direct-to-consumer business that continues to deliver for customers on the Un-carrier’s leading 5G network and now we are excited to use our scale and owners' economics to help supercharge it - and Ultra Mobile - into the future,” said Mike Sievert, CEO of T-Mobile. “Over the long-term, we’ll also benefit from applying the marketing formula Mint has become famous for across more parts of T-Mobile. We think customers are really going to win with a more competitive and expansive Mint and Ultra.”

“Our brands have thrived on the T-Mobile network, and we are thrilled that this agreement will take them even further, bringing the many benefits of 5G to even more Americans,” said David Glickman, founder and CEO of Mint, Ultra and Plum. “This transaction validates our meteoric success and will unite two proven industry innovators committed to doing things differently in the wireless industry.”

“Mint Mobile is the best deal in wireless and today’s news only enhances our ability to deliver for our customers. We are so happy T-Mobile beat out an aggressive last-minute bid from my mom Tammy Reynolds as we believe the excellence of their 5G network will provide a better strategic fit than my mom’s slightly-above-average mahjong skills. I am so proud of the entire Mint team and so excited for what’s to come,” said Ryan Reynolds.