Pascal Menezes, CTO of MEF, talks about the 3 top technology themes inside of @MEF_Forum, for 2020, including (1) the overlay/underlay framework (2) how clouds are moving to the edge, enabled by SD-WAN, and (3) architecture to accelerate 5G.
Thursday, November 21, 2019
MEF's 3 major technical themes for 2020
Pascal Menezes, CTO of MEF, talks about the 3 top technology themes inside of @MEF_Forum, for 2020, including (1) the overlay/underlay framework (2) how clouds are moving to the edge, enabled by SD-WAN, and (3) architecture to accelerate 5G.
What's new at MEF19?
Kevin Vachon gives a quick overview of the big news from #MEF19 in Los Angeles, California.
There is a big spotlight here on SD-WAN, new certification programs, the Proof-of-Concept showcase and overall progress with carrier orchestration and MEF 3.0.
Spectra7 and Molex showcase 400G Active Copper Interconnects
Spectra7 Microsystems and Molex showed their 400G Active Copper Cables (ACCs) at this year’s SuperComputing 2019 (SC19) show in Denver.
Spectra7’s GaugeChanger technology is said to work equally well at 25 Gbps NRZ and 50 Gbps PAM-4 enabling new connector standards of 100, 200 and 400 Gbps.
Molex recently began shipping its Passive Copper QSFP-DD (Quad Small Form Factor–Double Density) Cable Assemblies in volume production and the same customers are asking for Active Copper versions for longer lengths up to 7-10 meters. The Molex system along with Spectra7’s GaugeChanger technology offers up to 12 times less power and significant cost savings over optical solutions for hyperscale customers.
"Spectra7 has created a compelling solution to address the dramatically increasing power and cost of data center interconnects," said Greg Walz, Advanced Technical Marketing Manager, Molex. “We are seeing significant interest from our largest customers for 400G ACCs with Spectra7 technology and are pleased to be showcasing this product at SC19.”
"Molex is a technology and market leader in the hyperscale interconnect market," said Spectra7 CEO Raouf Halim. "The adoption of our 400G GaugeChangerTM technology by Molex is a significant milestone in the coming volume deployment of GaugeChangerTM, our data center product line."
http://www.spectra7.com
Spectra7’s GaugeChanger technology is said to work equally well at 25 Gbps NRZ and 50 Gbps PAM-4 enabling new connector standards of 100, 200 and 400 Gbps.
Molex recently began shipping its Passive Copper QSFP-DD (Quad Small Form Factor–Double Density) Cable Assemblies in volume production and the same customers are asking for Active Copper versions for longer lengths up to 7-10 meters. The Molex system along with Spectra7’s GaugeChanger technology offers up to 12 times less power and significant cost savings over optical solutions for hyperscale customers.
"Spectra7 has created a compelling solution to address the dramatically increasing power and cost of data center interconnects," said Greg Walz, Advanced Technical Marketing Manager, Molex. “We are seeing significant interest from our largest customers for 400G ACCs with Spectra7 technology and are pleased to be showcasing this product at SC19.”"Molex is a technology and market leader in the hyperscale interconnect market," said Spectra7 CEO Raouf Halim. "The adoption of our 400G GaugeChangerTM technology by Molex is a significant milestone in the coming volume deployment of GaugeChangerTM, our data center product line."
http://www.spectra7.com
Ayar Labs selected for Intel’s DARPA PIPES Project
Ayar Labs has been selected as Intel’s optical I/O solution partner for their recently awarded DARPA PIPES (Photonics in Package for Extreme Scalability) project.
The PIPES project aims to develop integrated optical I/O solutions co-packaged with next generation FPGA/CPU/GPU and accelerators in Multi-Chip Packages (MCP) to provide extreme data rates (input/output) at ultra-low power over much longer distances than supported by current technology. In the first phase of the project, the Ayar Labs TeraPHY chiplet will be co-packaged with an Intel FPGA using the AIB (Advanced Interconnect Bus) interface and Intel’s EMIB silicon-bridge packaging. “We’re seeing an explosion of Datacenter workloads that have an insatiable demand for bandwidth and the need to connect devices at rack-scale distances,” said Vince Hu, VP of Strategy and Innovation for Intel’s FPGA products. “The best way to do that is with optical interconnect and by using an Ayar Labs chiplet(s), we can achieve very high bandwidth at low latency and low power consumption.”
“Bringing optical connectivity all the way into the CPU/SOC package has long been one of the ‘Holy Grail’ projects in High Performance and Hyperscale Computing, as it unleashes the performance of ever more powerful computing and network processors and removes a major bottleneck and set of constraints in systems architecture and design,” said Charles Wuischpard, CEO of Ayar Labs, “Moreover, the energy consumed in moving data through a system is now very significant and growing, and the best way to manage that is to move the data optically from end to end. We are pleased to be selected by Intel as the optical solution for their DARPA PIPES project and look forward to a multi-year collaboration.”
The TeraPHY chiplet is manufactured on GLOBALFOUNDRIES' 45nm platform, which enabled Ayar Labs to build a monolithic, single-die solution that integrates both electrical and optical photonic circuits and devices on a single chip.
“We have worked in close collaboration with Ayar Labs to deliver a new class of integrated electronic, photonics solutions,” said Anthony Yu, vice president of Computing and Wired Infrastructure at GF. “Going forward, we’re excited to work with the pioneers at Ayar Labs to continue disrupting the market by combining our next generation 45nm platform, targeted to future CMOS-based photonics solutions, with their differentiated technology that will push the limits of chip communication bandwidth for high-performance computing, cloud and AI applications.”
Ayar Labs also announced customer sampling of its fully integrated TeraPHY chiplet starting in early Q1 2020.
http://www.ayarlabs.com
The PIPES project aims to develop integrated optical I/O solutions co-packaged with next generation FPGA/CPU/GPU and accelerators in Multi-Chip Packages (MCP) to provide extreme data rates (input/output) at ultra-low power over much longer distances than supported by current technology. In the first phase of the project, the Ayar Labs TeraPHY chiplet will be co-packaged with an Intel FPGA using the AIB (Advanced Interconnect Bus) interface and Intel’s EMIB silicon-bridge packaging. “We’re seeing an explosion of Datacenter workloads that have an insatiable demand for bandwidth and the need to connect devices at rack-scale distances,” said Vince Hu, VP of Strategy and Innovation for Intel’s FPGA products. “The best way to do that is with optical interconnect and by using an Ayar Labs chiplet(s), we can achieve very high bandwidth at low latency and low power consumption.”
“Bringing optical connectivity all the way into the CPU/SOC package has long been one of the ‘Holy Grail’ projects in High Performance and Hyperscale Computing, as it unleashes the performance of ever more powerful computing and network processors and removes a major bottleneck and set of constraints in systems architecture and design,” said Charles Wuischpard, CEO of Ayar Labs, “Moreover, the energy consumed in moving data through a system is now very significant and growing, and the best way to manage that is to move the data optically from end to end. We are pleased to be selected by Intel as the optical solution for their DARPA PIPES project and look forward to a multi-year collaboration.”
The TeraPHY chiplet is manufactured on GLOBALFOUNDRIES' 45nm platform, which enabled Ayar Labs to build a monolithic, single-die solution that integrates both electrical and optical photonic circuits and devices on a single chip.
“We have worked in close collaboration with Ayar Labs to deliver a new class of integrated electronic, photonics solutions,” said Anthony Yu, vice president of Computing and Wired Infrastructure at GF. “Going forward, we’re excited to work with the pioneers at Ayar Labs to continue disrupting the market by combining our next generation 45nm platform, targeted to future CMOS-based photonics solutions, with their differentiated technology that will push the limits of chip communication bandwidth for high-performance computing, cloud and AI applications.”
Ayar Labs also announced customer sampling of its fully integrated TeraPHY chiplet starting in early Q1 2020.
http://www.ayarlabs.com
Credo Joins Open Compute Project (OCP) to Accelerate 400G
Credo, which develops Serializer-Deserializer (SerDes) technology for 100G, 400G, and 800G, has joined the Open Compute Project (OCP).
“As disaggregated chassis begin to displace tradition big-iron switching and routing chassis, low power, highly flexible and routable 400G interconnect becomes a core component of the architecture,” said Don Barnetson, Sr. Director of Product at Credo. “OCP provides a platform to standardize and proliferate 400G disaggregated chassis configurations at hyperscalers around the world.”
Credo is also a member of the HiWire Consortium, announced in September 2019 with 25 other founding members, nearly half of whom are also amongst the OCP ranks. Working in harmony with other standards bodies, the HiWire Consortium is dedicated to the standardization and certification of a new interconnect cable category, Active Electrical Cables (AECs), enabling its broad support and wide industry adoption.
Credo began the development of AECs two years ago to address the obstacles that were slowing the transition to 400G, specifically the performance and physical limitations of copper (DACs) and the power, cost, and availability hurdles of active optical cables (AOCs). AECs integrate retimer and PAM4 to NRZ speedshifting functions with gearbox in-cable making them a lower power, affordable plug and play interconnect for 100G and 400G systems.
Credo says the power, performance, and price of AECs are empowering system architects to rethink the design of next-generation data center configurations, deploy 400G disaggregated chassis and provide a clear path to 800G in the future.
http://www.credosemi.com
“As disaggregated chassis begin to displace tradition big-iron switching and routing chassis, low power, highly flexible and routable 400G interconnect becomes a core component of the architecture,” said Don Barnetson, Sr. Director of Product at Credo. “OCP provides a platform to standardize and proliferate 400G disaggregated chassis configurations at hyperscalers around the world.”
Credo is also a member of the HiWire Consortium, announced in September 2019 with 25 other founding members, nearly half of whom are also amongst the OCP ranks. Working in harmony with other standards bodies, the HiWire Consortium is dedicated to the standardization and certification of a new interconnect cable category, Active Electrical Cables (AECs), enabling its broad support and wide industry adoption.
Credo began the development of AECs two years ago to address the obstacles that were slowing the transition to 400G, specifically the performance and physical limitations of copper (DACs) and the power, cost, and availability hurdles of active optical cables (AOCs). AECs integrate retimer and PAM4 to NRZ speedshifting functions with gearbox in-cable making them a lower power, affordable plug and play interconnect for 100G and 400G systems.Credo says the power, performance, and price of AECs are empowering system architects to rethink the design of next-generation data center configurations, deploy 400G disaggregated chassis and provide a clear path to 800G in the future.
http://www.credosemi.com
Vcinity and Naval Research Lab demo fastest memory-to-memory
Vcinity and the U.S. Naval Research Laboratory (NRL) have achieved near line rate performance for memory to memory data transfer and remote data access across a 100G link from Washington DC to Denver.
The memory to memory data transfer was performed at 95Gbps across CenturyLink’s 100G circuit with ~37ms latency (between Washington DC and Denver). RAD X-1100 extended EDR InfiniBand (IB) fabric i.e., RDMA beyond the four walls of the data center and across the WAN with local performance. NRL also showcased client access to data with a double loop of the network adding the Chicago path to demonstrate RDMA extension and application execution at twice the latency i.e., ~74ms – equivalent to a cross-country link within the U.S.
The milestone was accomplished as part of NRL’s Resilient Distributed Processing demonstration at SC19 in Denver, Colorado.
Vcinity and NRL have a long-standing relationship since the days when Vcinity’s predecessor, Bay Microsystems, enabled the first high speed data solution for high latency environments as part of NRL’s Large Data Joint Capabilities Technology Demonstration (LD-JCTD).
Vcinity’s Radical X (RAD X) product family supports IB and RDMA over Converged Ethernet (ROCE) fabric extension and enables a seamless data fabric transparent to virtually any application. RAD X-1100 provides unmatched scalability of Vcinity’s global fabric extension capabilities with three pluggable interface modules of up to 120Gbps capacity each – supporting various interface configurations including EDR InfiniBand and up to 100G RoCE – in a small 1 RU form factor. The company’s RAD X products are integrated with its Ultimate X family of products to deliver a more turnkey solution for commercial and government enterprise.
“We are delighted to be part of the first ever memory to memory transfer and remote data access demonstration at 100G. This is an important milestone for us and is a testament to our long-term commitment to NRL. We continue to scale our solutions for wider government and commercial communities,” said Harry Carr, CEO and President at Vcinity Inc.
Basil Decina, acting director, Center for Computational Science of the Information Technology Division of NRL added, “NRL provided the original requirements, along with a prototype, for this type of technology almost two decades ago — we are proud to drive innovation within the industry and see it continue to advance. We have demonstrated a number of new capabilities at SC19 this year and this is one that we believe changes the way data will be remotely processed and moved across large distances.”
https://vcinity.io
The memory to memory data transfer was performed at 95Gbps across CenturyLink’s 100G circuit with ~37ms latency (between Washington DC and Denver). RAD X-1100 extended EDR InfiniBand (IB) fabric i.e., RDMA beyond the four walls of the data center and across the WAN with local performance. NRL also showcased client access to data with a double loop of the network adding the Chicago path to demonstrate RDMA extension and application execution at twice the latency i.e., ~74ms – equivalent to a cross-country link within the U.S.
The milestone was accomplished as part of NRL’s Resilient Distributed Processing demonstration at SC19 in Denver, Colorado.
Vcinity and NRL have a long-standing relationship since the days when Vcinity’s predecessor, Bay Microsystems, enabled the first high speed data solution for high latency environments as part of NRL’s Large Data Joint Capabilities Technology Demonstration (LD-JCTD).
Vcinity’s Radical X (RAD X) product family supports IB and RDMA over Converged Ethernet (ROCE) fabric extension and enables a seamless data fabric transparent to virtually any application. RAD X-1100 provides unmatched scalability of Vcinity’s global fabric extension capabilities with three pluggable interface modules of up to 120Gbps capacity each – supporting various interface configurations including EDR InfiniBand and up to 100G RoCE – in a small 1 RU form factor. The company’s RAD X products are integrated with its Ultimate X family of products to deliver a more turnkey solution for commercial and government enterprise.
“We are delighted to be part of the first ever memory to memory transfer and remote data access demonstration at 100G. This is an important milestone for us and is a testament to our long-term commitment to NRL. We continue to scale our solutions for wider government and commercial communities,” said Harry Carr, CEO and President at Vcinity Inc.
Basil Decina, acting director, Center for Computational Science of the Information Technology Division of NRL added, “NRL provided the original requirements, along with a prototype, for this type of technology almost two decades ago — we are proud to drive innovation within the industry and see it continue to advance. We have demonstrated a number of new capabilities at SC19 this year and this is one that we believe changes the way data will be remotely processed and moved across large distances.”
https://vcinity.io
Cisco shows switches powered by Innovium
At this week's SC19 in Denver, multiple network equipment vendors, including Cisco, showcased high-performance switches based on Innovium's TERALYNX silicon.
Innovium said its 12.8Tbps TERALYNX low-latency production switch silicon is being used by OEMs and cloud providers for hyperscale and enterprise deployments.
Cisco showcased two Innovium-based switches, the Nexus 3432D-S and Nexus 3408-S. The Nexus 3432D-S is a 1RU, QSFP-DD switch that supports up to 32 ports of 400G, with each port able to operate in 25/40/50/100/400G speed. The Nexus 3408-S is a 4RU, 8-slot chassis with flexibility to use either 100G or 400G Line-Card Expansion Modules (LEMs) offering up to 128 ports of 100G or 32 ports of 400G in a pay-as-you-grow fashion.
GIGABYTE showcased a 1RU, QSFP-28/56 switch that supports up to 32-ports of 100/200G speed.
“High performance computing customers running applications such as genomics, seismic imaging, financial modeling, data analytics, AI/ML require networking solutions with the highest performance, highest port densities and lowest latencies,” said Rajiv Khemani, Co-founder and CEO of Innovium. “We are excited to provide the best application performance to these HPC customers with TERALYNX-based switches, that deliver these critical requirements from multiple system partners.”
"As a leader in high performance computing hardware, GIGABYTE has released one of the most comprehensive lineups of PCIe Gen 4.0 capable server platforms onto the market. At SC19, we are excited to collaborate with Innovium, a leader in high performance networking, to showcase joint solutions including servers with PCIe Gen 4.0 based 200GbE RoCE NICs and Innovium’s TERALYNX based low-latency multi-terabit RoCE capable switches, so that customers can achieve even greater productivity gains in their HPC workloads,” said Etay Lee, General Manager at Networking and Communication Business Unit, GIGABYTE.
http://www.innovium.com
Innovium said its 12.8Tbps TERALYNX low-latency production switch silicon is being used by OEMs and cloud providers for hyperscale and enterprise deployments.
Cisco showcased two Innovium-based switches, the Nexus 3432D-S and Nexus 3408-S. The Nexus 3432D-S is a 1RU, QSFP-DD switch that supports up to 32 ports of 400G, with each port able to operate in 25/40/50/100/400G speed. The Nexus 3408-S is a 4RU, 8-slot chassis with flexibility to use either 100G or 400G Line-Card Expansion Modules (LEMs) offering up to 128 ports of 100G or 32 ports of 400G in a pay-as-you-grow fashion.
GIGABYTE showcased a 1RU, QSFP-28/56 switch that supports up to 32-ports of 100/200G speed.
“High performance computing customers running applications such as genomics, seismic imaging, financial modeling, data analytics, AI/ML require networking solutions with the highest performance, highest port densities and lowest latencies,” said Rajiv Khemani, Co-founder and CEO of Innovium. “We are excited to provide the best application performance to these HPC customers with TERALYNX-based switches, that deliver these critical requirements from multiple system partners.”
"As a leader in high performance computing hardware, GIGABYTE has released one of the most comprehensive lineups of PCIe Gen 4.0 capable server platforms onto the market. At SC19, we are excited to collaborate with Innovium, a leader in high performance networking, to showcase joint solutions including servers with PCIe Gen 4.0 based 200GbE RoCE NICs and Innovium’s TERALYNX based low-latency multi-terabit RoCE capable switches, so that customers can achieve even greater productivity gains in their HPC workloads,” said Etay Lee, General Manager at Networking and Communication Business Unit, GIGABYTE.http://www.innovium.com
Innovium debuts smaller Ethernet switching chip
Innovium introduced a smaller Ethernet switch chip family for designs ranging from 1.2 to 6.4Tbps.
The new TERALYNX 5, which is expected to begin sampling in Q4 2019, is aimed at ToR, enterprise, edge, and 5G applications. Key capabilities include up to 128x NRZ/PAM4 SERDES, 10GbE to 400GbE ports, the largest on-chip buffers, powerful analytics, and leading performance per $ and performance per watt.
“The Innovium team has amassed a breakthrough, innovative IP portfolio, designed from the ground up, enabling us to deliver a programmable, low-latency 12.8T switch in production a full technology node earlier,” said Rajiv Khemani, CEO and Co-Founder of Innovium Inc. “We are delighted to further optimize these technologies for ToR and edge applications with TERALYNX 5, allowing a single consistent architecture to power data center switching applications from top to bottom with unmatched performance and value.”
TERALYNX 5 Family Highlights:
The new TERALYNX 5, which is expected to begin sampling in Q4 2019, is aimed at ToR, enterprise, edge, and 5G applications. Key capabilities include up to 128x NRZ/PAM4 SERDES, 10GbE to 400GbE ports, the largest on-chip buffers, powerful analytics, and leading performance per $ and performance per watt.
“The Innovium team has amassed a breakthrough, innovative IP portfolio, designed from the ground up, enabling us to deliver a programmable, low-latency 12.8T switch in production a full technology node earlier,” said Rajiv Khemani, CEO and Co-Founder of Innovium Inc. “We are delighted to further optimize these technologies for ToR and edge applications with TERALYNX 5, allowing a single consistent architecture to power data center switching applications from top to bottom with unmatched performance and value.”
TERALYNX 5 Family Highlights:
- Wide range of pin-compatible SKU options: 1.2Tbps, 2.4Tbps, 3.2Tbps, 4.8Tbps, 6.4Tbps
- Up to 128 SERDES PAM4 or NRZ SERDES enabling port speeds from 10Gbps to 400Gbps
- Largest on-chip buffer for switch in this class (45MB+)
- Leading L2, L3 table sizes and ACL’s
- Robust RoCE and PFC support for lowest latency and rich QoS necessary for distributed storage and AI applications
- IEEE 1588 v2 timing synchronization required in 5G and other data center infrastructure
- FLASHLIGHT v2 line-rate HW analytics with unprecedented microburst detection features and application correlation
- Up to 2x performance per $ and performance per watt vs. alternatives
- SW programmability for support of new protocols, achieved without impact to throughput or latency suffered by alternatives
- Fully SW and architecture compatibility with TERALYNX 7, scaling solutions to 12.8Tbps
Carbon nanotube exceeds 100GHz in RF applications
Carbonics, a start-up based in Culver City, California, announced that carbon nanotube technology has for the first time achieved speeds exceeding 100GHz in-radio frequency (RF) applications. Potential applications include 5G mm-Wave technologies.
Carbonics employs a deposition technology called ZEBRA that enables carbon nanotubes to be densely aligned and deposited onto a variety of chip substrates including silicon, silicon-on-insulator, quartz and flexible materials. This allows the technology to be directly integrated with traditional CMOS digital logic circuits and overcomes the typical problem of heterogeneous integration.
“This milestone shows that carbon nanotubes, long thought to be a promising communications chip technology, can deliver,” said Dr. Joe Qiu, Program Manager at the Army Research Office, who reviewed the research. “The next step is scaling this technology, proving that it can work in high-volume manufacturing. Ultimately, this technology could help the Army meet its needs in communications, radar, electronic warfare and other sensing applications.”
“With this exciting accomplishment, the timing is ripe to leverage our CMOS-compatible technology for the 5G and mmWave defense communication markets. We are now engaged in licensing and technology transfer partnerships with industry participants, while we continue to advance this disruptive RF technology,” said Carbonics’ CEO Kos Galatsis. Carbonics believes that its ZEBRA technology will likely first be adopted in military applications before being used more broadly.
In 2014, Carbonics was spun-out from the joint center of UCLA-USC and King Abdulaziz City for Science and Technology (KACST) called the Center of Excellence for Green Nanotechnologies and academic funding support from SRC, DARPA and the U.S. Air Force. The work published in Nature Electronics was funded by the US Army contract No. W911NF19P002 and by King Abdulaziz City for Science and Technology Center (KACST).
http://carbonicsinc.com/
Carbonics employs a deposition technology called ZEBRA that enables carbon nanotubes to be densely aligned and deposited onto a variety of chip substrates including silicon, silicon-on-insulator, quartz and flexible materials. This allows the technology to be directly integrated with traditional CMOS digital logic circuits and overcomes the typical problem of heterogeneous integration.
“This milestone shows that carbon nanotubes, long thought to be a promising communications chip technology, can deliver,” said Dr. Joe Qiu, Program Manager at the Army Research Office, who reviewed the research. “The next step is scaling this technology, proving that it can work in high-volume manufacturing. Ultimately, this technology could help the Army meet its needs in communications, radar, electronic warfare and other sensing applications.”“With this exciting accomplishment, the timing is ripe to leverage our CMOS-compatible technology for the 5G and mmWave defense communication markets. We are now engaged in licensing and technology transfer partnerships with industry participants, while we continue to advance this disruptive RF technology,” said Carbonics’ CEO Kos Galatsis. Carbonics believes that its ZEBRA technology will likely first be adopted in military applications before being used more broadly.
In 2014, Carbonics was spun-out from the joint center of UCLA-USC and King Abdulaziz City for Science and Technology (KACST) called the Center of Excellence for Green Nanotechnologies and academic funding support from SRC, DARPA and the U.S. Air Force. The work published in Nature Electronics was funded by the US Army contract No. W911NF19P002 and by King Abdulaziz City for Science and Technology Center (KACST).
http://carbonicsinc.com/
U.S. Air Force Base to deploy AT&T 5G
AT&T will deploy 5G services on Tyndall Air Force Base (Tyndall), which is located in the Florida panhandle. Tyndall suffered catastrophic structural damage in 2018 caused by Hurricane Michael’s Category 5, 160-mph winds.
Under the agreement with the Air Force, AT&T is rebuilding and modernizing communications infrastructure and capabilities across Tyndall. It is providing future-forward networking capabilities powered by AT&T 5G to support augmented and virtual reality, IoT, and a broad array of innovative technologies. AT&T will also deliver and manage commercial and private enterprise information technology capabilities at Tyndall. Initial services include mobility, cloud access, unified communications, voice, broadband, Wi-Fi expansion and an array of connected devices. AT&T’s networking solution at Tyndall will support network compute and storage as well as network edge capabilities that can more quickly deliver data and applications to users.
“The Air Force and AT&T share a vision for the smart base of the future: one that uses modern, commercially available communications capabilities to help our military maintain its globally competitive edge in defending our freedoms,” said Xavier Williams, President, AT&T Global Public Sector. “We’re proud to work with the Air Force and help fulfill its vision for how technology can power improvements in mission delivery while helping it keep its technological advantage.”
The Air Force also plans to equip its first responders and eligible public safety users at Tyndall with FirstNet – the nationwide, dedicated communications platform purpose-built for public safety. FirstNet operates on a dedicated core network that supports enhanced security and provides always-on priority and preemption services for Tyndall’s contingency and disaster response operations.
http://www.att.com
Under the agreement with the Air Force, AT&T is rebuilding and modernizing communications infrastructure and capabilities across Tyndall. It is providing future-forward networking capabilities powered by AT&T 5G to support augmented and virtual reality, IoT, and a broad array of innovative technologies. AT&T will also deliver and manage commercial and private enterprise information technology capabilities at Tyndall. Initial services include mobility, cloud access, unified communications, voice, broadband, Wi-Fi expansion and an array of connected devices. AT&T’s networking solution at Tyndall will support network compute and storage as well as network edge capabilities that can more quickly deliver data and applications to users.“The Air Force and AT&T share a vision for the smart base of the future: one that uses modern, commercially available communications capabilities to help our military maintain its globally competitive edge in defending our freedoms,” said Xavier Williams, President, AT&T Global Public Sector. “We’re proud to work with the Air Force and help fulfill its vision for how technology can power improvements in mission delivery while helping it keep its technological advantage.”
The Air Force also plans to equip its first responders and eligible public safety users at Tyndall with FirstNet – the nationwide, dedicated communications platform purpose-built for public safety. FirstNet operates on a dedicated core network that supports enhanced security and provides always-on priority and preemption services for Tyndall’s contingency and disaster response operations.
http://www.att.com
SiTime surges 43% in IPO
SiTime completed its initial public offering (IPO) and shares are now trading on the Nasdaq Stock Market under the symbol SITM.
On the first day of trading, SITM surged 43% over the opening price to close at $18.65.
“SiTime’s IPO on Nasdaq begins an exciting new chapter for us,” said Rajesh Vashist, CEO of SiTime. “SiTime is a timing pure-play leader that is using the power of silicon to transform the timing industry. Our sole focus on timing, along with our industry-leading MEMS, analog and systems technologies, enables significant growth opportunities for us in the $7.7B timing market. On behalf of SiTime, I want to thank our customers, employees, partners and suppliers who have played a key role in our success, and will continue to do so in the future.”
On the first day of trading, SITM surged 43% over the opening price to close at $18.65.
“SiTime’s IPO on Nasdaq begins an exciting new chapter for us,” said Rajesh Vashist, CEO of SiTime. “SiTime is a timing pure-play leader that is using the power of silicon to transform the timing industry. Our sole focus on timing, along with our industry-leading MEMS, analog and systems technologies, enables significant growth opportunities for us in the $7.7B timing market. On behalf of SiTime, I want to thank our customers, employees, partners and suppliers who have played a key role in our success, and will continue to do so in the future.”
SiTime prepares for IPO -- MEMS timing solutions
SiTime, a supplier or MEMS timing, has launched a roadshow for the initial public offering of 4,300,000 shares of its common stock. The underwriters will also have a 30-day option to purchase up to 645,000 additional shares of common stock at the IPO price, less underwriting discounts and commissions. The IPO price is currently expected to be between $13.00 and $15.00 per share. SiTime expects to list its common stock on The Nasdaq Global Market under the ticker symbol “SITM.”
SiTime Corporation is a wholly-owned subsidiary of MegaChips Corporation. The company has shipped over 1.5 billion devices to date.
http://www.sitime.com
SiTime Corporation is a wholly-owned subsidiary of MegaChips Corporation. The company has shipped over 1.5 billion devices to date.
http://www.sitime.com
AT&T activates first 400G circuit on Dallas/Atlanta route
AT&T activated its first 400-gigabit optical connection on a Dallas-to-Atlanta fiber route -- an industry first for a 400G long haul commercial link, according to the company.
The 400G trunk is now carrying live Internet traffic for customers. This connection – between Dallas and Atlanta – is believed to be a first in the industry.
The 400G Dallas/Atlanta facility was deployed on an SDN-enabled Ultra Long Haul (ULH) system from Ciena. The 400G circuit was routed and configured completely through software using Ciena’s Manage, Control and Plan (MCP) domain controller. The 400G transport was terminated in the Dallas and Atlanta offices by a white box router from UfiSpace that is compliant with the Broadcom Jericho2 Distributed Disaggregated Chassis (DDC) design that AT&T recently submitted to the Open Compute Project (OCP). 400G pluggable transceivers from InnoLight were installed in the white box router and Ciena transponder to create the cross-office connectivity between the packet and optical technologies. Ciena’s MCP controller is integrated into AT&T’s ONAP management and control framework using an Application Programming Interface (API) specified by the Open ROADM Multi-Source Agreement (MSA).
The optical transport system consists of a new 400G transponder based on Ciena’s WaveLogic Ai coherent optical technology. The Ciena Reconfigurable Optical Add/Drop Multiplexors (ROADMs) have been upgraded with new software to support a sophisticated feature known as flex grid. This allows us, through software control, to optimize the allocation of spectrum on the long-haul fiber based on the required speed and reach of each wavelength. The Broadcom Jericho2 white box was supplied by UfiSpace and provides 10 x 400G interfaces on a 2RU “pizza box”. The white box uses the Broadcom Jericho2 packet processing chip, a state-of-the-art chip designed to meet the demanding needs of a service provider.
“The move from a 100G interface to 400G is a milestone for the communications industry, because it means we can continue to stay ahead of the tsunami of data demand we’ve seen over the last decade-plus,” said Andre Fuetsch, executive vice president and chief technology officer, AT&T. “This accomplishment also speaks to the tremendously collaborative ecosystem we’ve helped foster with key innovators in optical technology, white box hardware, and software-defined networking. Ciena, UfiSpace, Broadcom, and InnoLight have brought great ideas and technologies to bear to make 400G a reality.”
“We are very proud to collaborate with AT&T on the design of the white boxes based on Broadcom’s Jericho2,” said Vincent Ho, general manager of UfiSpace. “In spite of the complexity and difficulty with the design from an electrical, mechanical, and thermal perspective, we’ve used our breadth of understanding in networking to overcome the challenges rolling out the first 400G carrier-scale white box solution. UfiSpace is very excited to be amidst the first to participate and contribute to this new paradigm. We are thankful to AT&T for their leadership in building a robust and open white box ecosystem.”
The 400G trunk is now carrying live Internet traffic for customers. This connection – between Dallas and Atlanta – is believed to be a first in the industry.
The 400G Dallas/Atlanta facility was deployed on an SDN-enabled Ultra Long Haul (ULH) system from Ciena. The 400G circuit was routed and configured completely through software using Ciena’s Manage, Control and Plan (MCP) domain controller. The 400G transport was terminated in the Dallas and Atlanta offices by a white box router from UfiSpace that is compliant with the Broadcom Jericho2 Distributed Disaggregated Chassis (DDC) design that AT&T recently submitted to the Open Compute Project (OCP). 400G pluggable transceivers from InnoLight were installed in the white box router and Ciena transponder to create the cross-office connectivity between the packet and optical technologies. Ciena’s MCP controller is integrated into AT&T’s ONAP management and control framework using an Application Programming Interface (API) specified by the Open ROADM Multi-Source Agreement (MSA).
The optical transport system consists of a new 400G transponder based on Ciena’s WaveLogic Ai coherent optical technology. The Ciena Reconfigurable Optical Add/Drop Multiplexors (ROADMs) have been upgraded with new software to support a sophisticated feature known as flex grid. This allows us, through software control, to optimize the allocation of spectrum on the long-haul fiber based on the required speed and reach of each wavelength. The Broadcom Jericho2 white box was supplied by UfiSpace and provides 10 x 400G interfaces on a 2RU “pizza box”. The white box uses the Broadcom Jericho2 packet processing chip, a state-of-the-art chip designed to meet the demanding needs of a service provider.“The move from a 100G interface to 400G is a milestone for the communications industry, because it means we can continue to stay ahead of the tsunami of data demand we’ve seen over the last decade-plus,” said Andre Fuetsch, executive vice president and chief technology officer, AT&T. “This accomplishment also speaks to the tremendously collaborative ecosystem we’ve helped foster with key innovators in optical technology, white box hardware, and software-defined networking. Ciena, UfiSpace, Broadcom, and InnoLight have brought great ideas and technologies to bear to make 400G a reality.”
“We are very proud to collaborate with AT&T on the design of the white boxes based on Broadcom’s Jericho2,” said Vincent Ho, general manager of UfiSpace. “In spite of the complexity and difficulty with the design from an electrical, mechanical, and thermal perspective, we’ve used our breadth of understanding in networking to overcome the challenges rolling out the first 400G carrier-scale white box solution. UfiSpace is very excited to be amidst the first to participate and contribute to this new paradigm. We are thankful to AT&T for their leadership in building a robust and open white box ecosystem.”
MEF 3.0 SD-WAN Certification Program gets underway
MEF is launching a MEF 3.0 SD-WAN Certification Program, with Spirent Communications as the SD-WAN Authorized Certification and Test Partner (ACTP), enabling service and technology providers to validate that their SD-WAN services and products conform to the recently published SD-WAN Service Attributes and Services (MEF 70) standard.
The certification will test the service attributes and their behaviors defined in MEF 70 and described in detail in the upcoming MEF SD-WAN Certification Test Requirements (MEF 90) standard. Demonstration of conformity with the industry’s first SD-WAN services standard is a valuable reference baseline for users choosing a service supplier in the inevitable confusion of a fast-growing new market.
MEF and Spirent have already begun working with an initial group of 7 service providers and technology vendors selected to participate in the MEF 3.0 SD-WAN Certification Pilot Program. The first certified companies are expected to be announced in Q1 2020.
“There are expansive opportunities in the SD-WAN market, with tens of billions of dollars in revenue at stake during the next five years, and many service providers are looking to quickly develop and advance their SD-WAN offerings to ensure they remain competitive,” said Nan Chen, President, MEF. “MEF 3.0 SD-WAN standardization and certification will play a vital role in accelerating SD-WAN market growth and facilitating creation of powerful new hybrid networking solutions optimized for digital transformation. Spirent’s technology, market leadership, and global reach make it a great partner to help maximize SD-WAN’s full market potential through certification.”
Commenting on the significant new partnership, Abhitesh Kastuar, General Manager of Spirent’s Cloud and IP business, said, "Spirent has earned a global reputation as the advisor and vendor-neutral authority that guides complex deployments of new technologies. SD-WAN is a particularly important innovation that accelerates the value realization of virtualized infrastructures, and Spirent solutions assure that resulting SD-WAN services are ready to perform as expected. We are pleased to join forces with MEF on the first SD-WAN Certification Program and look forward to the important work ahead of us as MEF 3.0 SD-WAN rolls out around the world.”
“SD-WAN is a dynamic service delivered using software-defined networking, virtualization, and inherent security and therefore requires a significant level of testing, validation, and assurance to achieve certification,” said Marc Cohn, Spirent Head of Virtualization. “The transition from traditional to MEF 3.0 orchestrated and assured services is significant to Spirent and for the entire global communications industry. We are thrilled to participate as the official SD-WAN Authorized Certification and Test Partner.”
The certification will test the service attributes and their behaviors defined in MEF 70 and described in detail in the upcoming MEF SD-WAN Certification Test Requirements (MEF 90) standard. Demonstration of conformity with the industry’s first SD-WAN services standard is a valuable reference baseline for users choosing a service supplier in the inevitable confusion of a fast-growing new market.
“There are expansive opportunities in the SD-WAN market, with tens of billions of dollars in revenue at stake during the next five years, and many service providers are looking to quickly develop and advance their SD-WAN offerings to ensure they remain competitive,” said Nan Chen, President, MEF. “MEF 3.0 SD-WAN standardization and certification will play a vital role in accelerating SD-WAN market growth and facilitating creation of powerful new hybrid networking solutions optimized for digital transformation. Spirent’s technology, market leadership, and global reach make it a great partner to help maximize SD-WAN’s full market potential through certification.”
Commenting on the significant new partnership, Abhitesh Kastuar, General Manager of Spirent’s Cloud and IP business, said, "Spirent has earned a global reputation as the advisor and vendor-neutral authority that guides complex deployments of new technologies. SD-WAN is a particularly important innovation that accelerates the value realization of virtualized infrastructures, and Spirent solutions assure that resulting SD-WAN services are ready to perform as expected. We are pleased to join forces with MEF on the first SD-WAN Certification Program and look forward to the important work ahead of us as MEF 3.0 SD-WAN rolls out around the world.”
“SD-WAN is a dynamic service delivered using software-defined networking, virtualization, and inherent security and therefore requires a significant level of testing, validation, and assurance to achieve certification,” said Marc Cohn, Spirent Head of Virtualization. “The transition from traditional to MEF 3.0 orchestrated and assured services is significant to Spirent and for the entire global communications industry. We are thrilled to participate as the official SD-WAN Authorized Certification and Test Partner.”
