Tuesday, March 22, 2022

NVIDIA unveils 51.2T Spectrum-4 switch ASIC

NVIDIA unveiled its next generation, 400 Gbps Spectrum-4 Ethernet switch with adaptive routing and enhanced congestion control mechanisms.

The NVIDIA Spectrum-4 uses a custom ASIC with a 51.2 terabits per second switching capacity, supporting up 128 ports of 400G or 64 ports of 800G, and based on 100 Gbps PAM4 SerDes technology. It integrates a 12.8 Tb/s crypto engine wtih support for MACsec and VXLANsec. It also support secure boot as default via hardware root of trust .

NVIDIA Spectrum-4 features a fully-shared and monolithic packet buffer that’s dynamically available to all ports. This provides  microburst absorption with true, port-to-port, cut-through latency. Spectrum-4 also supports programmability of the pipeline and packet modifier/parser without impact tolatency or packet rate, enabling flexibility without compromising performance.

NVIDIA also notes that its Spectrum-4 switch enables accelerated RoCE-based data transport and load balancing through adaptive routing (including adaptive routing notifications for dynamic flow rebalance), as well as high precision congestion control (HPCC) facilitated through in-band network telemetry.

The chip is based on a 4nm process and is expected to sample later this year.

"A new era of massive-scale cloud technologies, such as Omniverse, requires a transformation of data center architecture,” said Kevin Deierling, vice president of Networking at NVIDIA. “The Spectrum-4 platform’s extreme performance and robust security will equip data centers to power breakthrough discoveries that push the boundaries of what’s possible for the benefit of society.”


NVIDIA opens its NVLink die-to-die and chip-to-chip

NVIDIA will expand the use of its NVLink chip-to-chip and die-to-die interconnect technology in its GPUs, CPUs, DPUs, NICs and SOCs. The company also plans to open the technolgy to other for for custom chip and chiplet integrations. 

NVIDIA NVLink-C2C is built on top of NVIDIA’s world-class SERDES and LINK design technology, and it is extensible from PCB-level integrations and multichip modules to silicon interposer and wafer-level connections, delivering extremely high bandwidth while optimizing for energy and die area efficiency.

In addition to NVLink-C2C, NVIDIA will also support the developing Universal Chiplet Interconnect Express (UCIe) standard. Custom silicon integration with NVIDIA chips can either use the UCIe standard or NVLink-C2C, which is optimized for lower latency, higher bandwidth and greater power efficiency.

Some of NVLink-C2C’s key features include:

● High Bandwidth – supporting high-bandwidth coherent data transfers between processors and accelerators

● Low Latency – supporting atomics between processors and accelerators to perform fast synchronization and high-frequency updates to shared data

● Low Power and High Density – using advanced packaging, it is 25x more energy efficient and 90x more area-efficient than PCIe Gen 5 on NVIDIA chips

● Industry-Standard Support – works with Arm’s AMBA CHI or CXL industry-standard protocols for interoperability between devices

“Chiplets and heterogeneous computing are necessary to counter the slowing of Moore’s law,” said Ian Buck, vice president of Hyperscale Computing at NVIDIA. “We’ve used our world-class expertise in high-speed interconnects to build uniform, open technology that will help our GPUs, DPUs, NICs, CPUs and SoCs create a new class of integrated products 

NVIDIA estimates that its NVLink-C2C interconnect could deliver up to 25x more energy efficiency and be 90x more area-efficient than PCIe Gen 5 on NVIDIA chips and enable coherent interconnect bandwidth of 900 Gbps or higher.

“As the future of CPU design is increasingly accelerated and multichip, it is critical to support chiplet-based SoCs across the ecosystem,” said Chris Bergey, senior vice president and general manager of the Infrastructure Line of Business at Arm. “Arm is supporting a broad set of connectivity standards and designing our AMBA CHI protocol to support these future technologies, including collaborating with NVIDIA on NVLink-C2C to address use cases like coherent connectivity between CPUs, GPUs and DPUs.”


Ericsson intros Local Packet Gateway for enterprises

Ericsson introduced an all-in-one appliance – including hardware, Container-as-a-Service (CaaS), software and lifecycle management for enterprise edge applications.

The Ericsson Local Packet Gateway aims to make edge applications easier to deploy and manage. It brings dual-mode 5G Core user plane and network slicing to the edge, enabling both mobile broadband and enterprise use cases, from small user plane deployment in a CSP core network, to routing to local applications. Examples of CSP enterprise use cases include sport events applications, virtual reality (VR) gaming and augmented reality (AR) for inspecting factory quality.

Ericsson says optimized cloud-native user plane and service functions will enable new 5G edge use cases. 

Laurent Leboucher, Orange group CTO & Senior Vice President Orange Innovation Networks  says: “At Orange, we continue developing meaningful networks to adapt to the needs and uses all over the world. According to enterprise customers’ needs, a first option is to create a virtual private network on Orange’s public network. A second solution is to create a hybrid private network relying on shared equipment in the Orange network and dedicated equipment at the customer's site for critical flows and data. Ericsson’s 5G Core for standalone, including the Local Packet Gateway help us meet both these enterprise customers’ needs.”   

Monica Zethzon, Head of Solution Area Packet Core, Ericsson, says: “With the launch of the Ericsson Local Packet Gateway we have taken yet another step in supporting our CSP customers to address the enterprise market, by providing a powerful, slimmed-down user plane that’s easy to deploy at the edge, using an integrated lifecycle management making it easy to scale.”


HPE expands its GreenLake edge-to-cloud portfolio

HPE is rolling out significant enhancements to its flagship GreenLake hybrid cloud platform.

HPE GreenLak, which supports on-premise, edge, colocation, and public cloud deployments, is gaining traction worldwide, according to the company. In Q1 2022, HPE reported annual recurring revenue of $798 million, and increased as-a-service orders 136 percent year-over-year. The HPE GreenLake platform provides the foundation for more than 50 cloud services, including electronic health records, ML Ops, payments, unified analytics, and SAP HANA, as well as a wide- array of cloud services from partners.

Platform updates include:

  • Convergence of Aruba Central, a cloud-native, AI-powered network management solution, with the HPE GreenLake platform. Now, more than 120,000 Aruba networking customers, which includes nearly 2 million devices to manage and 2 million API calls per day, can use the HPE GreenLake platform to order services on-demand and manage their assets.
  • A new, unified operational experience that provides a simplified view and access to all cloud services, spanning the entire HPE portfolio, with single sign-on access, security, compliance, elasticity, and data protection.
  • 12 new cloud services in networking, data services, high performance computing and compute operations management.
  • HPE GreenLake for Block Storage -- a block storage as-a-Service with a 100% data availability guarantee built-in on a cloud operational model. 
  • Enhanced HPE Backup and Recovery Service for hybrid cloud.
  • New, purpose-built HPC capabilities for data-intensive workloads

 “HPE was among the first to deliver a cloud platform that enables customers to manage and extract insights from their data from edge to cloud, and our continued innovation is driving growth and furthering our market leadership,” said Antonio Neri, president and CEO, HPE. “In the hybrid cloud market, HPE GreenLake is unique in its simplicity, unification, depth of cloud services, and partner network. Today, we are furthering our differentiation, boldly setting HPE GreenLake even further apart as the ideal platform for customers to drive data-first modernization.”

Kioxia intros data center SSDs with PCIe 5.0 -- 14% performance boost

Kioxia is the first to introduce a data center SSD with a PCIe 5.0 interface, which doubles the bandwidth over PCIe 4.0 from 16 gigatransfers per second (GT/s) to 32GT/s. The CD8 Series is now available for customer evaluation.

Based on Kioxia’s 5th generation BiCS FLASH 3D flash memory technology, the CD8 Series utilizes a proprietary Kioxia controller and firmware, which can be customized to customer needs, and is housed in a 2.5-inch[2], 15mm Z-height form factor. The new drives are designed to the PCIe 5.0, Open Compute Project (OCP) Datacenter NVMe SSD 2.0 and NVMe 1.4 specifications.

Additional features include:

  • Read-intensive 1DWPD (Drive Write Per Day) endurance models targeted for hyperscale and server-centric workloads, in capacities from 960GB to 15.36TB
  • Mixed-used 3DWPD endurance targeted models are available, in capacities from 800GB to 12.8TB
  • Delivers up to 1.25M random read IOPS and 7.2GB/s sequential read throughput, an improvement of approximately 14% over the previous generation version
  • Security options include sanitize instant erase (SIE) and self-encrypted drive (SED) 

Eutelsat signs with OneWeb

Eutelsat Communications signed a global, multi-year Distribution Partnership Agreement (DPA) for OneWeb capacity. The agreement paves the way for Eutelsat to commercialise OneWeb services across key verticals including Maritime, Aviation, Enterprise, Telcos and Government.

Eutelsat Chief Executive Officer Eva Berneke said: "As a shareholder in OneWeb, we are excited to play a role in the success of this new constellation by incorporating OneWeb's connectivity services into our portfolio of solutions. This deal showcases the scope for synergies between our two companies and opens up the potential of low orbit to complement our geostationary assets in the fast-growing markets of aero and maritime mobility, fixed data and government services, building on the development of 5G and cloud technologies that will generate low latency requirements.”

OneWeb Chief Executive Officer Neil Masterson said: "This expanded partnership with Eutelsat offers us a significant opportunity to combine our GEO and LEO connectivity solutions, expanding our global capacity and further enhancing our commercial service. Drawing on Eutelsat’s decades of experience in serving the satellite industry, coupled with OneWeb’s substantial business momentum, the agreement demonstrates our collaborative approach to scaling up our LEO connectivity services. We are thrilled to see our services come together to expand connectivity solutions to communities.”

  • In December 2021, Eutelsat became OneWeb’s second-largest shareholder. 

Telstra to deliver teleport services for OneWeb in Australia

Telstra announced a 10-year deal to deliver teleport services to OneWeb in Australia. 

Telstra is building three new dedicated teleports across Australia. The first of the new teleports, located in Darwin Tivendale, is scheduled to begin installation this month with go-live planned in July. Two further sites – Charlton Toowoomba and Wangara, Perth, WA – are planned for completion later in 2022. Each facility will provide turnkey ground station support for OneWeb’s growing fleet of low-earth orbit (LEO) satellites.

Telstra said its turnkey approach for OneWeb includes designing, building and activating the teleports with ground station capabilities to meet OneWeb’s requirements. Telstra will also provide 24/7 monitoring and quality assurance services at each location.

“OneWeb had exacting requirements from the outset, and we worked in close partnership with them from site selection through construction,” said Vish Vishwanathan, Vice President Wholesale & Satellite, Telstra Americas. “Teleports are complex sites involving access to secure and resilient infrastructure and on-the-ground expertise, which Telstra has provided to OneWeb throughout this project.”

“Low Earth Orbit satellite technology is transforming the global connectivity landscape, not only by creating new business opportunities, but also giving more businesses, communities and governments the internet access they need for progress,” said Michele Franci, Chief of Delivery and Operations at OneWeb. “More connectivity options benefit everyone and our approach in establishing strategic partnerships with experienced providers like Telstra is core to how we deliver the OneWeb mission.”


  • Earlier this month, OneWeb signed an MOU with Telstra to explore new connectivity solutions for Australia and the Asia pacific regions.