Wednesday, May 31, 2023

Lightmatter lands $154 million for photonic interconnects / processors

Lightmatter, a start-up based in Boston, raised a $154 million Series C investment round for its photonic interconnects and processors.

Lightmatter’s photonics-enabled products encompass a full stack of hardware and software for AI and supercomputing applications.

“Rapid progress in artificial intelligence is forcing computing infrastructure to improve at an unprecedented rate. The energy costs of this growth are significant, even on a planetary scale,” said Lightmatter co-founder and CEO, Nick Harris. “Generative AI and supercomputing will be transformed by photonic technologies in the coming years, and our investors, partners, and customers are aligned with Lightmatter’s mission of enabling the future of computing infrastructure with photonics.”

The new funding came from SIP Global, Fidelity Management & Research Company, Viking Global Investors, GV (Google Ventures), HPE Pathfinder and existing investors. 

Lightmatter has now raised a total of over $270 million to date. 

  • In 2022, the company announced significant growth with the recruitment of top talent to its leadership team, including Richard Ho, who led Google’s Tensor Processing Unit program, and Ritesh Jain, who led datacenter chip packaging at Intel, as VPs of Hardware Engineering. Lightmatter also announced Jessie Zhang, who led corporate financial planning at Apple, as VP of Finance, and Steve Klinger, former VP at Innovium, as VP of Product.

CoreWeave raises $421 million for GPU cloud service

CoreWeave, a specialized cloud provider built for large-scale GPU-accelerated workloads, announced an additional $200 million in funding to expand its cloud GPU services.

The new funding , which was provided by  Magnetar Capital (“Magnetar”), a leading alternative asset manager, comes just one month after the company announced $221 million in Series B funding, also led by Magnetar, bringing the total for the round to $421 million.

CoreWeave said its  serverless orchestration layer makes it an attractive option for AI startups rushing to get a product to market, as teams can spin up workloads in as little as five seconds with increased portability, less overhead and less management complexity.

“In my 25-year career, I’ve never been a part of a company that’s growing like this,” said CoreWeave CEO and co-founder Michael Intrator. “It’s an incredible moment in time. From a demand standpoint, revenue and client scale, the rise has been exponential, and Magnetar has been an amazing partner from the beginning.”

CoreWeave was founded in 2017 and is based in Roseland, New Jersey.

Liquid Dataport connects Mombasa to Muanda via terrestrial fiber

Liquid Dataport launched its newest fibre route connecting Mombasa, Kenya, to Muanda on the west coast of the Democratic Republic of Congo (DR Congo). 

The new 3,800 km fiber route reduces data transmission latency by 20 milliseconds. 

The route connects Kenya and DRC, passing through Uganda and Rwanda and bringing more reliable and affordable broadband connectivity to over 40 million people living and working in all the major cities along the route. The latest route complements Liquid’s earlier achievement, already a first, in 2019, linking Dar Es Salaam to Muanda on the West Coast of  DR Congo via Zambia.

Hardy Pemhiwa, President & Group Chief Executive Officer of Liquid Intelligent Technologies, said, “The real challenge today is closing the access-usage gap in Africa so that more Africans can use the internet technologies available to them, now and in the future. This East-West route which compliments our existing Pan Africa fibre network, is significant because it is helping to solve that problem – it not only brings global traffic to the continent but also improves the cost economics of Africa’s broadband Internet access.”

“We have a significant number of wholesale, enterprise and hyperscale customers along this route, and we fully support them in operating their global networks. The availability of our latest and shortest East to West route brings many proven economic and social benefits – from providing access to online educational resources to creating more jobs and driving the adoption of new technologies,” said David Eurin, Chief Executive Officer of Liquid Dataport.

FCC updates National Broadband Map

The FCC released an updated version of its location-based national broadband map.

The FCC finds that more than 8.3 million U.S. homes and businesses lack access to high-speed broadband. 

On net, the improvements to the map since November helped to identify nearly 330,000 more unserved locations.

The FCC mapping team has reviewed challenges to availability data for more than 4 million locations. Over 75% of those challenges have already been resolved and the majority have led to updates in the data on the map showing where broadband is available. At the same time, the new map also reflects a net increase of more than one million new serviceable locations, as compared to the November 2022 pre-production draft.

MACOM opens European Semiconductor Center in France

MACOM is establishing a European Semiconductor Center to offer its customers higher frequency Gallium Arsenide (GaAs) and Gallium Nitride (GaN) monolithic microwave integrated circuits.

MACOM recently acquired the key manufacturing facilities, capabilities and technologies of OMMIC SAS. 

The OMMIC facility, which is located near Paris in Limeil-Brévannes, France, will become the foundation for MACOM’s newly established European Semiconductor Center. 

“We are excited to open MACOM’s European Semiconductor Center,” stated Stephen G. Daly, President and Chief Executive Officer, MACOM.

5G Open Innovation Lab aims for faster deployment of private mobile

The 5G Open Innovation Lab (5G OI Lab) is collaborating with F5, GXC and Spirent Communications to showcase an enterprise Private Mobile Network (PMN) capable of delivering service in hard-to-reach locations with greater security, control, and resilience.

As 5G infrastructure becomes more mainstream, enterprises and organizations are subsequently deploying more private networks and telco clouds into strategic locations to leverage newly available connectivity, lower latency, and agility. The consistent throughput and capacity on these PMNs bring the promise of 5G to revenue-driven use cases such as IoT and network slicing, while being mindful that security must expand accordingly beyond legacy policies and be thoroughly tested. Especially in edge use cases, one of the requirements most in demand is the need for local breakout to route traffic to as part of a Mobile Edge Compute (MEC) infrastructure or hosted within a data center.

F5, a hybrid and multi-cloud application services and security company, deployed its carrier-grade BIG-IP Virtual Editions. These security functions were consolidated with an N6/SGi-LAN solution adding another layer of speed and security at the network level. 

GXC, a NaaS communications early-stage tech company which is part of the 5G OI Lab ecosystem, contributed their unique distributed mesh configuration for easier and faster deployment of a private network while providing full coverage and capacity of a cellular network.

This collaboration was validated using Spirent’s Landslide test and emulation platform, which provided device and RAN emulation to create a diversity of traffic flows across multiple data networks through the 5G core which is monitored and managed by F5 behind the UPF. 

Finally, the entire solution was pressure-tested through technology made available through the 5G OI Lab and its enterprise and operator partners deployed with GXC’s ONYX Portal.

The 5G Open Innovation Lab notes that in the four years since its founding, its membership roster has grown to include Accenture, Amdocs, Avanade, Dell Technologies, Deloitte, Ericsson, F5, GAF (Standard Industries), Intel, Microsoft, Nokia, Palo Alto Networks, SK Telecom, Spirent, T-Mobile and VMware as well as 101 multi-stage enterprise startups who have collectively raised $1.641B of venture capital.

Tuesday, May 30, 2023

Amazon Security Lake consolidates data from cloud, on-prem and SaaS

Amazon Web Services launched a service that automatically centralizes an organization’s security data from across their AWS environments, leading SaaS providers, on-premises environments, and cloud sources into a purpose-built data lake.

Amazon Security Lake converts and conforms incoming security data to the Open Cybersecurity Schema Framework (OCSF) open standard, making it easier for security teams to automatically collect, combine, and analyze security data from more than 80 sources, including AWS, security partners, and analytics providers. 

The service builds the security data lake using Amazon Simple Storage Service (Amazon S3) and AWS Lake Formation to automatically set up security data lake infrastructure in a customer’s AWS account, providing full control and ownership over security data. 

The data lake collects security data from sources like Amazon VPC Flow Logs and AWS CloudTrail, third-party sources like Splunk, CrowdStrike, Datadog, and Cribl, and their own sources of data. 

Keysight and Nokia test 6G active sub-THz components

Keysight Technologies and Nokia Bell Labs will demonstrate component and radio frequency integrated circuit (RFIC) characterization for 6G subterahertz (sub-THz) radio frequency (RF) applications in the D- and E-bands at the International Microwave Symposium 2023 (IMS 2023) in June.

The companies have been collaborating to develop characterization techniques to test active sub-THz components, such as amplifiers, mixers, and frequency converters, to ensure they do not introduce additional signal distortion into communications systems.

The demonstration characterizes a Nokia power amplifier (PA) developed for use in 6G D-band wireless communications systems and designed to operate with a low error vector magnitude (EVM). Using the Keysight N5290A PNA Millimeter Wave System, Nokia Bell Labs can accurately detect the PA’s low EVM, which is a measure of the signal distortion introduced into the communications system by the component. With these measurements, Nokia Bell Labs can better understand the PA’s performance to optimize its 6G transmission system design.

Keysight and Nokia Bell Labs are also partnering to showcase an E-band test solution at IMS 2023. The demonstration features the Keysight N9042B UXA Signal Analyzer used in combination with the V3050A Signal Analyzer Frequency Extender to measure a Nokia transceiver RFIC designed for use in an E-band system for backhaul applications.

Shahriar Shahramian, Head of RFIC & Packaging Lab, Nokia Bell Labs, said: “At Nokia Bell Labs we continue to push the carrier frequency and complexity of our RFIC, packaging, and module integration. With these advancements we move the frontier of what is possible from 6G wireless and optical communication systems. Consequently, the test and measurement industry plays a vital role to enable the complex characterization demanded from our devices and to reduce our time to market.”

NVIDIA’s data center revenue rises 14% yoy to $4.28B

NVIDIA reported quarterly revenue of $7.19B, down 13% compared to a year ago but well above its outlook. More significantly, NVIDIA again affirmed that Cloud Service Providers (CSPs) and enterprises are racing to deploy its flagship NVIDIA Hopper- and Ampere-architecture GPUs to meet surge in interest in generative AI.

GAAP earnings per diluted share for the quarter were $0.82, up 28% from a year ago and up 44% from the previous quarter. Non-GAAP earnings per diluted share were $1.09, down 20% from a year ago and up 24% from the previous quarter.

“The computer industry is going through two simultaneous transitions — accelerated computing and generative AI,” said Jensen Huang, founder and CEO of NVIDIA. “A trillion dollars of installed global data center infrastructure will transition from general purpose to accelerated computing as companies race to apply generative AI into every product, service and business process. Our entire data center family of products — H100, Grace CPU, Grace Hopper Superchip, NVLink, Quantum 400 InfiniBand and BlueField-3 DPU — is in production. We are significantly increasing our supply to meet surging demand for them.” 

Some highlights

  • Data center first-quarter revenue was a record $4.28 billion, up 14% from a year ago and up 18% from the previous quarter.
  • Google Cloud is the first cloud provider offering the new NVIDIA L4 Tensor Core GPU to accelerate generative AI applications.
  • Gaming first-quarter revenue was $2.24 billion, down 38% from a year ago and up 22% from the previous quarter.
  • Professional Visualization first-quarter revenue was $295 million, down 53% from a year ago and up 31% from the previous quarter.
  • Automotive first-quarter revenue was a record $296 million, up 114% from a year ago and up 1% from the previous quarter.

HPE sees growth in GreenLake edge-to-cloud

Hewlett Packard Enterprise reported Q2 revenue of $7.0 billion, up 4% from the prior-year period and 9% in constant currency. GAAP margin amounted to 36.0%, an all-time record for the company.

“Building on a great start to the fiscal year, HPE grew revenue, increased the contribution of recurring revenue through the HPE GreenLake edge-to-cloud platform, and delivered exceptional profitability to generate a strong second quarter performance,” said Antonio Neri, president and CEO of Hewlett Packard Enterprise. “Our shift to a higher-margin portfolio mix led by the Intelligent Edge segment, and the strong demand for our AI offering, further strengthen the investment opportunity for our shareholders.”

Some highlights:

* Intelligent Edge revenue was $1.3 billion, up 50% from the prior-year period in actual dollars and 56% in constant currency, with 26.9% operating profit margin, compared to 12.6% in the prior-year period. 

* High Performance Computing & Artificial Intelligence revenue was $840 million, up 18% from the prior-year period in actual dollars and 22% in constant currency. 

* Compute revenue was $2.8 billion, down 8% from the prior-year period in actual dollars and 3% in constant currency.

* Storage revenue was $1.0 billion, down 3% from the prior-year period in actual dollars and up 2% in constant currency. HPE Alletra revenue grew triple digits from the prior-year period. 

* Financial Services revenue was $858 million, up 4% from the prior-year period in actual dollars and up 7% in constant currency.

IP Infusion cites surging demand for white box open networking

IP Infusion reported record-breaking growth in 1Q 2023 with sales at the highest ever level for a single quarter. 

The company, which is a wholly owned and independently operated subsidiary of ACCESS CO, said first quarter OcNOS sales orders are more than three times Year over Year compared to Q1 2022, led by robust demand for OcNOS’s disaggregated Data Center, Aggregation Router and Cell Site Router products.

In addition, 23 new customers in the first quarter chose OcNOS, the industry’s first and preferred full-featured modular extensible network OS designed for white box open networking solutions. 

“IP Infusion is proud to report yet another unequaled quarter of sales growth fueled by a powerful demand for our industry-acclaimed OcNOS products,” said Atsushi Ogata, President and CEO of IP Infusion. “This solid performance reflects our commitment to exceed expectations for delivering the most advanced disaggregated solutions to our expanding customer base and partners across the globe.”

Vector Capital to acquire Riverbed

Vector Capital, a leading private equity firm, agreed to acquire Riverbed Technology. Financial terms were not disclosed.

Riverbed, which was founded in 2002 and is based in San Francisco, specializes in network acceleration and telemetry.

Riverbed’s existing lenders will continue to support the Company by providing it with a new credit facility at closing. 

Stephen Goodman, a Principal at Vector Capital, added: “With a significantly enhanced balance sheet and greater financial flexibility, Riverbed will have the ability to accelerate innovative offerings like Aternity and Alluvio IQ, creating value for its blue-chip customer base.”

Semtech appoints new CEO

Semtech appointed Paul H. Pickle as its new president and chief executive officer, replacing Mohan Maheswaran, who announced his retirement in March.

Previously, Mr. Pickle has served as president and chief executive officer and member of the board of directors of Lantronix. Before that, he served as president and chief operating officer of Microsemi Corporation, a provider of semiconductor and system solutions, from November 2013 until Microsemi was acquired by Microchip Technology Inc. in May 2018 for $10.3 billion. 

Monday, May 29, 2023

NVIDIA positions Spectrum-X Networking for generative AI

NVIDIA introduced its Spectrum-X Networking Platform for Ethernet-based AI clouds.

The Spectrum-X solution consists of the company's existing 51 Tbps NVIDIA Spectrum-4 Ethernet switch and the NVIDIA BlueField-3 DPU, along with acceleration software and software development kits (SDKs).

Advanced RoCE extensions work across the Spectrum-4 switches, BlueField-3 DPUs and NVIDIA LinkX optics to create an end-to-end 400GbE network that is optimized for AI clouds. The Spectrum-4 Ethernet switch supports smart-leaf, spine, and super-spine designs. It offers 64 ports of 800GbE in a dense 2U form factor, and up to 256 200Gbs ports can be connected by a single switch, or 16,000 ports in a two-tier leaf-spine topology.

The software stack includes  Cumulus Linux, pure SONiC and NetQ — which together enable the networking platform’s extreme performance. The BlueField DPUs are powered by NVIDIA DOCA™ software framework.

The Spectrum-X solution set also includes a  selection of direct attach copper cables (DACs), active copper cables (ACCs), active optical cables (AOCs), and optics needed to connect OSFP and QSFP-based fabrics.

“Transformative technologies such as generative AI are forcing every enterprise to push the boundaries of data center performance in pursuit of competitive advantage,” said Gilad Shainer, senior vice president of networking at NVIDIA. “NVIDIA Spectrum-X is a new class of Ethernet networking that removes barriers for next-generation AI workloads that have the potential to transform entire industries.”

MediaTek’s auto chip integrates NVIDIA GPU chip via coherent interconnect

MediaTek will develop automotive SoCs and integrate the NVIDIA GPU chiplet, featuring NVIDIA AI and graphics intellectual property, into the design architecture. The chiplets are connected by an ultra-fast and coherent chiplet interconnect technology.

In addition, MediaTek will run the NVIDIA DRIVE OS, DRIVE IX, CUDA and TensorRT software technologies on these new automotive SoCs to enable connected infotainment and in-cabin convenience and safety functions. 

“AI and accelerated computing are fueling the transformation of the entire auto industry,” said JensenHuang, NVIDIA’s founder and CEO. “The combination of MediaTek’s industry-leading system-on-chip plus NVIDIA’s GPU and AI software technologies will enable new user experiences, enhanced safety and new connected services for all vehicle segments, from luxury to entry-level.”

Arm intros its next CPU + GPU platform for mobile computing

Arm introduced its Total Compute Solutions 2023 (TCS23) platform for mobile computing.

TCS23 is designed and optimized for specific workloads that need to run seamlessly together as a complete system. This includes a new Arm Immortalis GPU based on the company’s 5th Generation GPU architecture, a new cluster of Armv9 CPUs for next-gen artificial intelligence (AI) workloads, and new enhancements to deliver more accessible software for the millions of Arm developers. 

A vital part of the high-performance cluster for smartphones is the new Arm Cortex-X4, which is the fastest CPU that Arm has ever built, bringing 15 percent more performance compared to the Cortex-X3. Meanwhile, the new power efficient microarchitecture consumes 40 percent less power than Cortex-X3 on the same process.

Cadence and Arm target faster path to tapeout

Cadence Design Systems, in  collaboration with Arm , is targeting a faster path to tapeout through use of Cadence digital and verification tools and the new Arm Total Compute Solutions 2023 (TCS23), which includes the Arm Cortex-X4, Arm Cortex-A720 and Cortex-A520 CPUs and Immortalis-G720, Mali-G720 and Mali-G620 GPUs.

Through this latest collaboration, Cadence delivered comprehensive RTL-to-GDS digital flow Rapid Adoption Kits (RAKs) for 3nm and 5nm nodes to help customers achieve power and performance goals using the new Arm TCS23. 

Cadence said its Cerebrus tool provided Arm with AI-driven design optimization capabilities that resulted in 50% better timing (TNS), a 10% reduction in cell area and 27% improved leakage power on the Cortex-X4 CPU, empowering Arm to achieve power, performance and area (PPA) targets faster.

“With the delivery of Arm TCS23, we’re unlocking the creative potential for customers around the world who demand high performance and efficient compute for next-generation mobile experiences,” said Chris Bergey, senior vice president and general manager, Client Line of Business at Arm. “Through our ongoing collaboration with Cadence, we successfully leveraged the Cadence digital and verification flows to get our latest generation of CPUs and GPUs into our customers’ hands faster and accelerate time to market.”

“Arm is known for building foundational computing platforms that address performance and efficient compute requirements, and the new Arm TCS23 is aimed at empowering designers to create the world’s most advanced mobile designs that provide the best user experience,” said Dr. Chin-Chi Teng, senior vice president and general manager, Digital & Signoff Group at Cadence. “By collaborating with Arm on the development of TCS23 and by delivering innovative digital and verification flows, we’re jointly leading our customers to improve overall productivity and achieve faster time to tapeout.”

Dell'Oro: Improving component supply drives growth in optical transport equipment

Optical Transport equipment revenue grew 8 percent year-over-year in 1Q 2023, according to a new report from Dell’Oro Group.

"For a second consecutive quarter, the Optical Transport equipment market growth exceeded our expectations," said Jimmy Yu, Vice President at Dell'Oro Group. "Equipment manufacturers are benefiting from the record level of backlog built-up over the past two years due to a lack of component supply. As these shortages moderate and lead times shrink, vendors are in a position to deliver long-awaited DWDM systems to their customers. This was the first time in two years that we did not hear a single vendor mention component shortage as a reason for lower than expected revenue in the quarter," added Yu.

Additional highlights from the 1Q 2023 Optical Transport Quarterly Report:

  • The largest global suppliers in the trailing four quarter period were Huawei, Ciena, Nokia, ZTE, and Infinera. Among these vendors, Nokia and Infinera outperformed. For the trailing four quarter period, Nokia's optical revenue grew 13 percent and Infinera's grew 20 percent.
  • China market slightly contracted in the quarter. If we exclude China from the worldwide figures, the optical market grew at a much higher rate of 12 percent year-over-year.
  • Internet content provider (ICP) spending on WDM equipment increased year-over-year in the quarter, continuing the rebound in purchasing that started in 4Q 2022 after an unexpected drop in the third quarter.

Ofcom greenlights discounted wholesale fibre pricing from Openreach

Ofcom, the UK telecoms market regulator, will allow Openreach to offer discounted wholesale pricing for its full-fibre service.

The new pricing scheme, know as the Equinox 2 plan, gives lower prices to retail providers – such as BT, Sky, TalkTalk and Vodafone – if they agree to use mainly Openreach’s full-fibre products for new orders instead of its legacy copper products.

Thursday, May 25, 2023

Adtran launches FSP 3000 open line systems for the coherent optical edge

Adtran launched its FSP 3000 Edge OLS, an open line system designed for the new generation of coherent edge access, aggregation and metro networks. The platform is specifically engineered for the use of coherent transmission at the optical edge, including the recently introduced Coherent 100ZR, as well as 400ZR optics. 

Adtran says that by utilizing power-optimized wavelength broadcasting technology and a gridless architecture, its solution enables a simple and affordable optical layer with compact and flexible nodes that can transport any coherent signal and baud rate. 

The modular OLS supports multiple use cases, including linear, hub-and-spoke or tree topologies. The platform delivers single-degree functions in 1RU (add/drop, line amplification, OTDR, OSC). It features a carrier-grade and hardened design for outdoor deployments, enabling operation in extended temperatures range (-40°C to 65°C).

“At OFC, we showcased our world-first QSFP28 Coherent 100ZR, bringing robust, efficient and simple coherent technology to the edge of the network. Now we’re complementing this innovative transceiver with an OLS optimized for the cost-efficient transport of coherent technology at the optical edge. Providing maximum performance, it meets all critical demands of edge aggregation networks, including scalability, compact design, low power consumption, operational simplicity and extended temperature operation for outdoor deployment,” said Christoph Glingener, CTO at Adtran. “Our pioneering work at the network edge is creating whole new opportunities for service providers.”

“Our FSP 3000 Edge OLS offers a major boost to operators looking for an affordable way to tackle relentless growth in data traffic. Low-cost yet highly flexible, it can scale to transparently accommodate any line speed or baud rate. Now there’s a clear route to deploying an optical layer that leverages coherent transport technology, including 100ZR and 400ZR optics, and meets the stringent requirements of the new optical edge,” commented Henning Hinderthür, VP of product line management, Optical Networks at Adtran. “Our FSP 3000 Edge OLS harnesses advanced wavelength broadcast technology for new levels of power and cost efficiency. It’s also ready to be deployed anywhere from central offices to street furniture.

WSJ: Tipster in Ericsson's bribery case awarded $279 million by SEC

The tipster who reported Ericsson to the U.S. Securities and Exchange Commission for bribery was awared $279 million under the government's cash-for-tips program, according to The Wall Street Journal. A tipster can be awarded between 10% to 30% of fines collected, depending on circumstances.

Ericsson resolved the case with the SEC in 2019, agreeing to pay a $1.1 billion fine. A subsequent issue led to Ericsson being fined an additional $207 million.

The identity of the tipster has not been disclosed.


MaxLinear intros 2.5G Ethernet switches and eight port 2.5G Enterprise PHYs

MaxLinear introduced a new line of 2.5G Ethernet products, including new 2.5G Ethernet switches and eight port 2.5G Enterprise PHYs.

  • MXL862XX 2.5G Octal Switch/PHY Key Features
  • Integrated 16 port Switch with 10/100/1000/2.5GBASE-T PHYs (5 or 8 ports)
  • Two optional 10G uplink ports (MxL862x2)
  • Power: 6.5W/8 Port switch, 4.3W/5 Port switch, 5.5W/8 port PHY
  • 400MHz ARC CPU: Web smart
  • Unblocking wire speed switching
  • Feature rich switch with VLAN, QinQ, QoS, traffic shaping
  • BGA-277 12x12mm, 0.628mm pitch switch
  • BGA-256 13x13, 0.8mm pitch 8 port PHY

MaxLinear will demonstrate its new 2.5G solutions at Computex 2023 in the Taipei Nangang Exhibition Center, Nangang Hall 1- 4F, Booth M1413. The trade show takes place from Tuesday, May 30 – Friday, June 2.

"MaxLinear has a 20-year history of being an innovator and major supplier of Ethernet PHYs and switches," said James Lougheed, Vice President & GM, High Performance Analog at MaxLinear. "The 2.5G standard has continued to replace lower speed 1G solutions in the market with an exponentially increasing amount of end points adopting the technology. This new Octal Switch/PHY and Octal PHY family of products will now leverage all those upgraded end points by equipping small-medium business & Enterprise Switches with cost effective 2.5G PHY and Switch solutions.”

MaxLinear noted that it has shipped more than 1 billion ports of Ethernet physical layer (PHY) transceivers and switches.

Broadband Forum issues TR-459 Issue 2 spec

The Broadband Forum announced the second issue of Technical Report-459: Multi-Service Disaggregated BNG with CUPS.

"TR-459 Issue 2 includes an additional layer of flexibility to manage subscriber resiliency, simplifying operations and streamlining provisioning and maintenance," said Jonathan Newton, Co-Director of the Access and Transport Architecture Work Area at Broadband Forum from Vodafone.

Any faults that can cause network outages can be navigated around as the new specification enables the brain of the network (known as the Control Plane, responsible for managing interactions with the customer home router) to intelligently determine whether a switchover to a new User Plane is needed. The User Plane carries network traffic and enforces policies such as Quality of Service (QoS).

“The latest release of the TR-459 standard will continue to drive standards-based interoperability between vendors, with Broadband Forum simplifying network transformation and ensuring more agility and flexibility for broadband operators,” said Craig Thomas, Vice President Strategic Marketing and Business Development at Broadband Forum. “Our work on a Multi-Service Disaggregated Broadband Network Gateway (DBNG) continues to make it easier to deliver fixed broadband services quickly to new and existing customers as the industry continues to transition and offers a migration path from legacy infrastructures to a future agile cloud native approach.”

Standardization of the Control and User Plane Separation (CUPS) for a Multi-Service DBNG enables the Control Plane from one vendor to control and manage the User Plane from another vendor. This builds a more diverse supply chain with a greater choice of Control and User Planes from different vendors leveraging the TR-459 standard.

Broadband Forum members showcased the important capabilities enabled by TR-459 Issue 2 as part of the CloudCO Demonstration at Network X 2022. The demo highlighted Subscriber Group Resiliency and switchover, and DBNG-User Plane traffic steering. The network monitored the subscriber traffic SLAs and when the SLAs were not being met, switched the subscribers to the backup DBNG-User Plane. Subscriber Group Resiliency minimizes network outages during any network faults.

MaxLinear supplies PAM4 DSP for JPC's 800G active electrical cable

Taiwan-based JPC Connectivity has selected MaxLinear’s 5nm Keystone PAM4 DSP to produce 800G active electrical cables for hyperscale datacenters. 

JPC Connectivity will demonstrate the new cables at Computex 2023 in Taipei, Taiwan from May 30 – June 2.

MaxLinear said its Keystone can yield up to a 40% power saving over competitor solutions. The Keystone 5nm DSP family has been designed to address both 400G and 800G applications and is the first generation to provide 106.25Gbps host side electrical I/O to match the line side 106.25Gbps interface rate. Variants supporting single mode optics (EML and SiPh), multimode optics (VCSEL transceivers and AOCs), and Active Electrical Cables (AECs) are all available and can be paired with companion TIAs to provide complete solutions.

“We’re excited to partner with JPC Connectivity on these new active optical cables,” said Drew Guckenberger, Vice President of High Speed Interconnect at MaxLinear. “Our Keystone DSPs offer a significant power advantage in this type of application, which is increasingly becoming a critical factor with hyperscale data centers.”

“Our new 800G cables using the low power, high performance Keystone DSP is designed to meet the market’s need for next-gen interconnect solutions as datacenters transition from passive DAC cables,” said Jessica Chang CEO at JPC Connectivity. “We are happy to have MaxLinear as a partner for this innovative new product.”

Marvell posts revenue of $1.322 billion, seems AI as a growth driver

 Marvell Technology reported revenu of $1.322 billion for the first quarter of its fiscal year 2024, $22 million above the midpoint of the company's prior guidance and down 9% year-on-year. GAAP net loss for the first quarter of fiscal 2024 was $(168.9) million, or $(0.20) per diluted share. Non-GAAP net income for the first quarter of fiscal 2024 was $264 million, or $0.31 per diluted share. Cash flow from operations for the first quarter was $208.4 million.

"We delivered first quarter fiscal 2024 revenue of $1.322 billion, above the midpoint of guidance, and are forecasting sequential revenue growth in the second quarter. We are expecting revenue growth to accelerate in the second half of this fiscal year, accompanied by gross and operating margin expansion," said Matt Murphy, Marvell's President and CEO. "AI has emerged as a key growth driver for Marvell, which we are enabling with our leading network connectivity products and emerging cloud optimized silicon platform. While we are still in the early stages of our AI ramp, we are forecasting our AI revenue in fiscal 2024 to at least double from the prior year and continue to grow rapidly in the coming years."

Wednesday, May 24, 2023

Microsoft: China's Volt Typhoon targets US critical infrastructure

Microsoft uncovered stealthy and targeted malicious activity aimed at critical infrastructure in the United States and carried out by Volt Typhoon, a state-sponsored actor based in China.

Microsoft alledges that this Volt Typhoon campaign is pursuing development of capabilities that could disrupt critical communications infrastructure between the United States and Asia region during future crises.

The attack gains initial access to targeted organizations through Internet-facing Fortinet FortiGuard devices. From there, the attacker extracts credentials to an Active Directory account used by the device, and then attempts to authenticate to other devices on the network with those credentials. Volt Typhoon proxies all its network traffic to its targets through compromised SOHO network edge devices (including routers). Microsoft has confirmed that many of the devices, which include those manufactured by ASUS, Cisco, D-Link, NETGEAR, and Zyxel, allow the owner to expose HTTP or SSH management interfaces to the internet.

The National Security Agency (NSA) has also published a Cybersecurity Advisory [PDF] which contains a hunting guide for the tactics, techniques, and procedures (TTPs) of this attack.

Ayar Labs raises additional funding for optical I/O chiplets

Ayar Labs, a start-up based in Santa Clara, California, raised an additional $25 million in Series C1 funding, for its work in silicon photonics for chip-to-chip connectivity.

Ayar Labs uses industry-standard silicon processing techniques to deliver fast, high-density, low power optical I/O chiplets and multi-wavelength light sources. 

The new funding brings the company's total Series C raise to $155 million. The oversubscribed up round was led by new investor Capital TEN. VentureTech Alliance also entered the Series C expansion that included participation by previous investors Boardman Bay Capital Management, IAG Capital Partners, NVIDIA, and Tyche Partners. Existing strategic and financial investors include Applied Ventures, GlobalFoundries, Hewlett Packard Pathfinder, Intel Capital and Lockheed Martin Ventures.

“We’re extremely pleased with the ongoing interest and financing we’ve received from leaders in the semiconductor industry,” said Charles Wuischpard, CEO of Ayar Labs. “This C1 adds sophisticated investor partners that will allow us to accelerate our strategic roadmap, and is further validation of our technology and plan to bring silicon photonics-based interconnect solutions to market at scale.”

NVIDIA, which participated in Ayar Labs’ earlier Series C raise in April 2022, increased its investment in the company.

“NVIDIA is reimagining the data center with integrated hardware, software and networking for accelerated computing,” said Craig Thompson, Vice President of Business Development, Networking Business Unit at NVIDIA. “Generative AI models with trillions of parameters are accelerating demand for this platform, which is why we are increasing our investment in Ayar Labs.”

Tech Update video: Ayar Labs' Optical I/O Chiplets

How will optical I/O revolutionize data communications?

Charlie Wuischpard, CEO from Ayar Labs explains:

  • The benefits of optical communication brought closer to the source, replacing copper traces on motherboards with fiber optic communication
  • Ayar Labs' chiplet using silicon photonics technology and micro ring resonators for efficient data transmission
  • Applications in AI training, scale-out architectures, and data centers for improved performance and efficiency

Check out other Tech Updates on our YouTube Channel (subscribe today): and check out our latest reports at:

Have a tech update that you want to brief us on? Contact!

NVIDIA sees trillion dollar upgrade cycle for data center infrastructure

NVIDIA reported revenue for the first quarter ended April 30, 2023, of $7.19 billion, down 13% from a year ago and up 19% from the previous quarter.

GAAP earnings per diluted share for the quarter were $0.82, up 28% from a year ago and up 44% from the previous quarter. Non-GAAP earnings per diluted share were $1.09, down 20% from a year ago and up 24% from the previous quarter.

“The computer industry is going through two simultaneous transitions — accelerated computing and generative AI,” said Jensen Huang, founder and CEO of NVIDIA.

“A trillion dollars of installed global data center infrastructure will transition from general purpose to accelerated computing as companies race to apply generative AI into every product, service and business process. Our entire data center family of products — H100, Grace CPU, Grace Hopper Superchip, NVLink, Quantum 400 InfiniBand and BlueField-3 DPU — is in production. We are significantly increasing our supply to meet surging demand for them,” he said.

Dell'Oro: Open RAN and vRAN growth slowed in 1Q 2023

Revenue growth moderated in 1Q 2023 for both Open RAN and Virtualized RAN (vRAN), according to a recent report from Dell'Oro Group. 

“Since both Open RAN and vRAN are still driven by a few large service providers, the slower growth in the first quarter was for the most part in line with expectations, reflecting more so the state of the 5G rollouts with some of the advanced operators than a shift in the overall Open RAN/vRAN market sentiment,” said Stefan Pongratz, Vice President with the Dell’Oro Group. “There are also signs that activity is improving beyond the early adopters but it will of course take some time for the sum of these smaller early majority type deployments to move the broader market,” continued Pongratz.

Additional Open RAN and vRAN highlights from the 1Q 2023 RAN report:

  • After more than doubling in 2022, Open RAN revenue growth was in the 10 to 20 percent range in the first quarter while the vRAN market advanced 20 to 30 percent.
  • Positive developments in the Asia Pacific region were dragged down by more challenging comparisons in the North America region.
  • Short-term projections remain unchanged – Open RAN is still projected to account for 6 to 10 percent of the 2023 RAN market.
  • Top 5 Open RAN suppliers by revenue for the 2Q 2022 to 1Q 2023 period include Samsung, NEC, Fujitsu, Rakuten Symphony, and Mavenir.

Dell'Oro: MEC market fails to materialize

The Multi-access Edge Computing (MEC) market has failed to materialize as expected. Dell'Oro Group is is lowering expectations by more than 20 percent for 2023.

In addition, the China Mobile Core Network (MCN) market hit a steep decline in 1Q 2023 weighing down the total market.

“The MEC market is not reaching the potential market as expected initially. Except for China, the big promise for MEC was addressing the needs of Enterprises. Several factors are contributing to the slower uptake,” stated Dave Bolan, Research Director at Dell’Oro Group. “Some examples are: standalone private networks have garnered more attention and emphasis than MEC, which we define as extensions to Mobile Network Operators (MNOs) networks; lack of applications with well-defined return-on-investment (ROI); competition from Hyperscale Cloud Provider (HCP), and upcoming Wi-Fi 7. We believe Enterprises are evaluating all of these choices before moving forward with more aggressive MEC deployments. As a result, we are lowering the MEC forecast for 2023 by more than 20 percent from our previous forecast.

“The China market’s growth rate had a steep decline in 1Q 2023 as its build-out of 5G Standalone networks reached maturity and is projected to end 2023 with a negative growth rate, a first since we started tracking the China market revenues in 2018. However, the Worldwide market, excluding China, in 1Q 2023, got the year off to a good start with the highest growth rate since 4Q 2019. We expect positive growth rates throughout the year with continuing buildouts of 5G Standalone networks,” continued Bolan.

Additional highlights from the 1Q 2023 Mobile Core Network and Multi-Access Edge Computing Report include:

  • The top MCN vendors worldwide for 1Q 2023 were Huawei, Ericsson, Nokia, and ZTE.
  • The top MCN vendors worldwide, excluding China, for 1Q 2023 were Huawei, Ericsson, and Nokia.
  • The top 5G MCN vendors worldwide for 1Q 2023 were Huawei, Ericsson, ZTE, and Nokia.
  • The report counts 43 Mobile Network Operators (MNOs) that have launched commercial 5G SA eMBB networks as of 1Q 2023.

Tuesday, May 23, 2023

DriveNets designs network fabric for large-scale AI workloads

DriveNets introduced a networking solution designed to maximize the utilization of AI infrastructures and improve the performance and utilization of large-scale AI workloads. DriveNets Network Cloud-AI aims for 30% improvement in JCT (Job Completion Time) of large-scale AI workloads, substantially improving resource utilization, while also supporting a standard Ethernet which allows for vendor interoperability and choice.

"AI compute resources are extremely costly and must be fully utilized to avoid 'idle cycles' as they await networking tasks," said Ido Susan, DriveNets co-founder and CEO. "Leveraging our experience supporting the world's largest networks, we have developed DriveNets Network Cloud-AI. Network Cloud-AI has already achieved up to a 30% reduction in idle time in recent trials, enabling exponentially higher AI throughput compared to a standard Ethernet solution. This reduction also means the network effectively 'pays for itself' through more efficient use of AI resources." 

"Network Cloud-AI provides balanced fabric connectivity between all GPUs in a cluster just as InfiniBand does," said Susan. "The difference is that Network Cloud-AI interfaces with servers on standard Ethernet. InfiniBand uses proprietary equipment which creates vendor lock on the networking and GPU level." 

DriveNets Network Cloud-AI is based on OCP's Distributed Disaggregated Chassis (DDC) architecture which is built on a distributed leaf-and-spine model designed to support service provider high-scale networks. Highlights:

  • Scale - connects up to 32,000 GPUs at speeds ranging from 100G to 800G to a single AI cluster with perfect load balancing
  • Maximum utilization - equally distributes traffic across the AI network fabric, ensuring maximum network utilization and zero packet loss under the highest loads
  • Shortest JCT - supports congestion-free operations through end-to-end traffic scheduling, avoids flow collisions and jitter, and provides zero-impact failover with sub-10ms automatic path convergence
  • Openness - is an Ethernet-based solution that avoids proprietary approaches and supports vendor interoperability with a variety of white box manufacturers (ODMs), Network Interface Cards (NICs), and AI accelerator ASICs

DriveNets said early trials by leading hyperscalers using Network Cloud-AI over white boxes with Broadcom's Jericho family chipset achieved up to 30% improvement in JCT compared to other Ethernet solutions. 

"Ethernet has proven time and again to be the best choice for all networking needs by enabling an open, healthy, and competitive ecosystem," said Ram Velaga, senior vice president and general manager, Core Switching Group, Broadcom. "Large-scale training and inference of AI models will benefit from networks that can perform at 100% utilization like DriveNets Network Cloud-AI. Broadcom's Jericho3-AI delivers an Ethernet network with perfect load balancing and end-to-end congestion management, resulting in significant reduction in job completion time compared to any other alternative."

Farice lights Iceland-Ireland subsea cable with Ciena

Farice has deployed Ciena's GeoMesh Extreme submarine network solution to support its recently launched IRIS cable system connecting southwest Iceland to Galway, Ireland, with onward connectivity to the UK, Denmark, and the Nordics.

 The IRIS cable system, which went live on March 1, is a six-fiber pair trunk delivering 22Tb/s per fiber pair, for a total capacity of 132Tb/s, at 500Gb/s line rates. The 1,700km submarine cable is the third to connect Iceland to Europe, providing a diverse, high-capacity pathway for traffic flowing between the regions. 

For IRIS, Farice deployed Ciena’s 6500 Packet-Optical Platform, powered by WaveLogic 5 Extreme coherent optics, and Manage, Control and Plan (MCP) domain controller. Ciena Services provided consulting, systems integration, training, and optimization—and will perform maintenance services to ensure the network continues to operate at peak performance. Additionally, Ciena assisted Farice with upgrading terrestrial routes in Iceland for IRIS and DANICE to a protected ring system with a fast restoration using a ROADM-based GeoMesh Extreme architecture; DANICE was also updated to WaveLogic 5 Extreme coherent optics.

"By deploying a third submarine network cable in Iceland, Farice has dramatically increased the redundancy and resiliency of connectivity in the northwest Atlantic. Route diversity and green energy capabilities will help make Iceland highly attractive to data center operators, cloud providers, and other businesses that require high-performance networking," states Thomas Soerensen, VP, Global Submarine Solutions, Ciena.

Apple signs multiyear deal with Broadcom for 5G RF components

Apple signed a new multiyear, multibillion-dollar agreement with Broadcom for 5G radio frequency components — including FBAR filters — and cutting-edge wireless connectivity components. 

The FBAR filters will be designed and built in several key American manufacturing and technology hubs, including Fort Collins, Colorado, where Broadcom has a major facility woth more than 1,100 employees.

“We’re thrilled to make commitments that harness the ingenuity, creativity, and innovative spirit of American manufacturing,” said Tim Cook, Apple’s CEO. “All of Apple’s products depend on technology engineered and built here in the United States, and we’ll continue to deepen our investments in the U.S. economy because we have an unshakable belief in America’s future.”

Broadcom will continue to invest in critical automation projects and upskilling with technicians and engineers at the Fort Collins site.

CoreSite expands in Silicon Valley and Denver

CoreSite announced the continued expansion of its Silicon Valley and Denver data center campuses. Plans include:

  • Silicon Valley: CoreSite is adding its ninth data center in Santa Clara, California. With the addition of the new 240,000-square-foot, purpose-built data center – known as SV9 – the company’s newest Silicon Valley data center campus will have 1.179 million total square feet of data center space. Located at 2915 Stender Way, SV9 will offer 34 megawatts of IT load with a new utility substation onsite. CoreSite has also completed the design and permitting process to add 3,420 square feet and 500 kilowatts of IT load to its downtown Denver (DE1) data center at 910 15th Street, one of the most interconnected data centers in the Denver metro region. Construction of SV9 is expected to be complete in 2024, preleasing is available now.
  • Denver: Strategically located in downtown Denver, CoreSite’s DE1 and DE2 data center facilities deliver two network-rich data centers and the region’s largest peering exchange. The expansion at DE1 will increase the total square footage in CoreSite’s Denver campus to 37,420. CoreSite is pre-leasing the DE1 expansion, which will be available for customers in 2H 2023.

“With expansions in both our Silicon Valley and Denver markets, CoreSite is delivering on our long-term strategy to meet the ever-increasing data center capacity demands and interconnection requirements of our valued customers,” said Juan Font, President and CEO, CoreSite and SVP of U.S. Tower. “Our strong financial backing by Fortune 375 wireless infrastructure leader, American Tower, and private equity partner, Stonepeak, means we have the direct access to capital required to grow with our customers to anticipate and build solutions for their future needs.”

Nick Lippis ONUG perspective: Capex, Opex & Cloud Migration

Struggling to navigate the transition from hardware to software-based infrastructure? 

Nick Lippis, Co-Founder & Co-Chair of ONUG, explains:

– The shift from CapEx to OpEx and managing infrastructure you don’t own

– The rise of network and security teams in enterprise cloud infrastructure

– How ONUG provides tools, knowledge, and best practices for successful digital transformation 

See video:

Check out the rest of ONUG Spring 2023 showcase here:

Want to be in one of videos? Contact us at

ATX cites safety of its supercapacitors for broadband

ATX Networks recently completed testing of its hybrid supercapacitor-based energy storage modules that demonstrate the solution’s ability to withstand excessive physical and electrical stress without creating a safety hazard or dangerous environmental conditions.

 The testing was conducted by Georgia-based testing laboratory Energy Assurance, part of the Element Materials Technology (Element) Group.

ATX claims that its next-generation energy storage medium is safer than alternative backup power sources, such as lead-acid and some lithium-based batteries, and less susceptible to catching fire or leaking dangerous chemical substances because they are electrostatic, rather than electrochemical.

 “Broadband service providers want assurances that the equipment they use to provide standby power throughout their networks will not ignite, leak or explode if mishandled or subjected to extreme conditions,” said George Goncalves, ATX’s Director of Energy Storage Product Management. “The Areca units continued to perform as expected with no cell damage and no instances of thermal runaway during any of the eight tests run on our hybrid supercapacitor units.”

Vietnam's VNPT deploys 10G fiber broadband with Nokia

Vietnam Posts and Telecommunications (VNPT) has begun deployment of 10G fiber broadband with Nokia. The initial rollout will connect 10,000 homes and businesses in the 8 largest provinces of the country. 

Nokia is providing fiber access nodes for the operator’s exchanges and fiber modems for end users’ premises. The deployment includes Nokia's 7360-FX Lightspan series access node (OLT) which supports multiple fiber technologies including GPON, XGS-PON, 25GS-PON and Point-to-Point Ethernet to deliver a wide range of services.

In September 2021, the country’s Ministry of Information and Communications launched the National Digital Transformation Programme with the goal that fiber internet infrastructure should cover 80 percent of households and 100 percent of communes by 2025.

Mr. Dang Anh Son, CEO at VNPT Net Corporation, said: “This project is our first step towards high-class FTTx services which meet our residential and business customers’ demand for higher bandwidth and service quality using the latest generation XGS-PON. Next step, VNPT will work with Nokia to deploy cloud-based network management systems for the XGS-PON and enhance customer experiences through digital transformation solutions.”

Rubén Morón Flores, Head of Market Unit Vietnam at Nokia, said: “We are proud to have been selected by VNPT for this important milestone of bringing 10G broadband to the country for the first time. The APAC region is the largest in the world for fiber broadband. Our multi-gig and multi-PON solution provides for incredible flexibility and scale. Vietnam has ambitious plans and we’re delighted to be part of them.”

Nokia notes that its Quillion chipset now has more than 250 operator customers around the world. 

ADVA announces planned CEO transition

ADVA confirmed that Christoph Glingener will step down as CEO and transition to his previous role as CTO for ADVA effective June 30, 2023, and will continue as CTO for Adtran, to which he was appointed in January 2023. 

Tom Stanton will be appointed ADVA’s new CEO effective July 1, 2023, and will also continue as Adtran’s CEO. Tom Stanton was named Adtran Inc. CEO in 2005 and chairman of the board in 2007.

Monday, May 22, 2023

Blueprint: Using Carrier Geolocation Data for Real-Time Insights

by Heather Broughton, AVP Marketing, Service Provider Solutions at NETSCOUT

Many communication service providers (CSPs) investing in 5G technology are turning to geomarketing solutions to better reach their revenue goals. This follows reports that 9 in 10 companies worldwide use location data, and that use is expected to grow in significance over the next few years. 

However, geomarketing as an economic opportunity is underutilized because existing geolocation methods lack operational feasibility and precision, leading to many CSPs wondering how they can refine their approach and tap into additional geolocation revenue streams over the next decade. 

That is the topic we will explore in this article, specifically how carrier geolocation data can increase revenue opportunities for subscriber location-based services without the need for GPS. This approach allows CSPs to analyze their data better and discover intelligent and timely insights into their subscribers, devices, services, networks, and applications. 

Geolocation Solutions Rely on Precision

When we talk about precision in geolocation services, we first must address the fact that current legacy geomarketing methods in the market – such as Over the Top (OTT) GPS and passive geolocation – has fundamental flaws that restrict their capabilities for real-time use. Examples include errors in dense urban areas, limitations indoors where coverage is limited, and GPS systems that are easily turned off and run out of battery. Likewise, passive geolocation methods suffer from configuration errors rendering large locations, Net Energy Metering (NEM) and User Equipment (UE) Radio Frequency (RF) parameters that contain errors and bias, and higher latency due to ‘weighting’ past sessions to smooth movements.

By comparison, through a better understanding of network data, particularly by applying smart data instrumentation and analytics at the RAN level and beyond, mobile network operators can significantly improve accuracy compared with traditional methods. Machine learning (ML) models can be trained to learn on just 1% of 3GPP-based minimize drive test (MDT) calls to predict all other calls and provide a more precise, anonymized view of subscribers with fewer variations compared with traditional modeling. This type of data is extremely rich – giving insight into location-based activities and behavior – and, when combined with other sources within data lakes, can yield a range of revenue opportunities across sectors. 

Geolocation Data Monetization Opportunities by Sector 

In general, geolocation data is becoming harder and harder to come by. Smartphone manufacturers and ecosystem players like Apple and Google have introduced measures in recent years to limit the amount of information collected by applications to improve user privacy. This affords mobile network operators considerable opportunity to provide precision geolocation data with many important use cases across industries. Consider just a few examples:

  • Advertising and marketing: Geomarketing uses consumers' location to provide relevant messaging for products or services they may be interested in at that particular point in time. With accurate geolocation data, CSPs can help advertisers set geofencing to trigger select mobile ads – such as to promote a new store or regional offer – and verify ad conversions. 
  • Fraud detection and prevention: Credit card fraud in the U.S. continues to rise each year, amounting to billions of dollars in costs to financial institutions. However, even with sophisticated algorithms, spotting fraud in real-time remains a challenge. Accurate, real-time geolocation data offers an opportunity for card issuers to verify transactions and stop payment before any funds are lost.  
  • Business & government planning: Before building a new branch location or paving a new road, planners want to know traffic patterns and behaviors. Having accurate geolocation data is invaluable in verifying other data sources and can help organizations set or adjust strategies to reflect the real world.  

In conclusion, current geomarketing methods, including GPS and passive monitoring, have unique limitations, and emerging technologies such as smart data, RAN-based machine learning models can help ensure data accuracy to provide a range of unique services across industries. As 5G continues to advance in the next decade, CSPs will find new ways to improve existing geomarketing methods and provide a transformational experience for their subscribers.

About the Author

Heather Broughton is an area vice president overseeing communications service provider solutions at NETSCOUT. She has more than 20 years of experience in telecom, including positions held at Verizon/MCI focused on performance management and translations for both fixed and mobile networks.

Axelera AI raises $50 million for edge inference

Axelera AI, a start-up based in Eindhoven, Netherlands, raised over $50 million in Series A funding for its development of a hardware + software solution for AI interference at the edge.

The company says its AI edge solution enables computer vision applications to become more accessible, powerful and user friendly than ever before.

In addition to its headquarters in Eindhoven, Axelera AI has R&D offices in Belgium, Switzerland, Italy and the UK, with more than 120 employees in 14 countries. 

New investors include CDP Venture Capital, Verve Ventures and Fractionelera.

IBM offers Hybrid Cloud Mesh service

IBM introduced a SaaS offering to help enterprises to bring management to their hybrid multicloud infrastructure.

IBM Hybrid Cloud Mesh automates the process, management and observability of application connectivity in and between public and private clouds to help modern enterprises operate their infrastructure across hybrid multicloud and heterogeneous environments.

When generally available later this year, IBM Hybrid Cloud Mesh will also leverage the DNS traffic steering capabilities from NS1, an IBM Company, to find the best connection between clouds and end users and deliver applications that are optimized for performance, cost, and availability at every connection point.

IBM’s recent acquisition of NS1, a provider of network automation SaaS solutions that are designed to enable businesses to deliver content, services and applications to millions of users while helping optimize performance, reliability, security, and cost, will also play in a role in helping IBM Hybrid Cloud Mesh provide application-centric connectivity.

Applied Materials plans Semiconductor Innovation in Silicon Valley

Applied Materials plans to invest up to $4 billion over the next 7 years in Silicon Valley to build "the world’s largest and most advanced facility for collaborative semiconductor process technology and manufacturing equipment research and development (R&D)." 

The new Equipment and Process Innovation and Commercialization (EPIC) Center is planned as the heart of a high-velocity innovation platform designed to accelerate development and commercialization of the foundational technologies needed by the global semiconductor and computing industries.

The facility will include more than 180,000 square feet of state-of-the-art cleanroom for collaborative innovation with chipmakers, universities and ecosystem partners. 

“While semiconductors are more critical to the global economy than ever before, the technology challenges our industry faces are becoming more complex,” said Gary Dickerson, President and CEO of Applied Materials. “This investment presents a golden opportunity to re-engineer the way the global industry collaborates to deliver the foundational semiconductor process and manufacturing technologies needed to sustain rapid improvements in energy-efficient, high-performance computing.”

Applied’s new EPIC Center is designed to be a premier platform for leading logic and memory chipmakers to collaborate with the equipment ecosystem. For the first time, chipmakers can have their own dedicated space within an equipment supplier facility, extending their in-house pilot lines and providing early access to next-generation technologies and tools – months or even years before equivalent capabilities can be installed at their facilities.

Leading semiconductor and computing companies voicing support include AMD, IBM, Intel, Micron, Nvidia, Samsung, TSMC and Western Digital.

Google Cloud region opens in Qatar

Google opened a cloud region data center in Doha, Qatar. 

The new region launches with three zones, allowing users to distribute applications and storage to help protect against service disruptions. The Doha region also launches with a key set of Google Cloud products, including Compute Engine, Cloud Run, Cloud SQL, Google Kubernetes Engine, and Span.

The new Doha region is part of Google Cloud’s global network of 37 regions and 112 zones that bring cloud services to over 200 countries and territories worldwide.