Qualcomm pushes on-chip AI with SenseTime deal

By Nick Flaherty at www.flaherty.co.uk

Qualcomm Technologies has teamed up with Chinese algorithm developer SenseTime on artificial intelligence (AI) and machine learning (ML) for future Snapdragon mobile and IoT chips.

The collaboration will add SenseTime’s ML models and algorithms to the premium and high-tier Snapdragon platforms, which offer advanced heterogeneous computing capabilities for client based AI. The companies expect to drive the popularity and development of on-device AI in areas such as innovative vision and camera-based image processing.

Devices such as smartphones and connected cameras are becoming more intelligent with the proliferation of AI. Implementing AI on the device provides a number of advantages over cloud-only implementations, enabling edge devices to provide reliable execution with or without a network connection. Additional benefits of on-device AI include real-time performance, privacy protection and enhanced reliability.

“To develop an AI ecosystem, it takes efforts from players in multiple industries,” said Dr Li Xu, co-founder and chief executive officer of SenseTime. “The strategic collaboration between SenseTime and Qualcomm Technologies will advance on-device intelligence by leveraging our algorithm and Qualcomm Technologies’ chipset. Together we’ll push the envelope and extend AI to places that are currently beyond reach. Our strategic collaboration will become a turning point for the whole AI ecosystem."

“Qualcomm has been conducting fundamental research in AI over a decade,” said Keith Kressin, senior vice president, product management, Qualcomm Technologies, Inc. “In fact, many devices shipping today using our Snapdragon mobile platforms already utilize on-device AI. We look forward to the results of our collaboration with SenseTime to further accelerate new and exciting capabilities of on-device AI for millions of customers using mobile devices.”

Currently, Qualcomm Technologies is focused on optimizing the Snapdragon mobile platform to accelerate myriad AI use cases in the areas of computer vision and natural language processing — for smartphones, IoT and automotive — and is researching broader executions in the areas of wireless connectivity, power management, and photography. SenseTime's proprietary deep learning platform called Parrots allows a variety of AI algorithms to be developed with low cost and quick turn-around in small footprints. It has more than 400 customers and strategic partners including Qualcomm, NVIDIA, China Mobile, UnionPay, Huawei, Xiaomi, OPPO, vivo and Weibo.

The first fruits of the collaboration are being demonstrated at the end of the month (October 2017).

Related stories:

• Another startup aims for embedded AI
• Qualcomm boosts its embedded AI development with Scyfer buy
• ARM takes aim at embedded AI
• NVIDIA pushes artificial intelligence into embedded designs with Jetson TX2
• Edge analytics vital for security says Greenwave
• Xilinx pushes machine learning and AI to the edge for embedded applications
• Startup aims to bring artificial intelligence to IoT nodes

New optical receivers open up gigabit broadband

By Nick Flaherty at www.flaherty.co.uk

A new receiver technology developed by UCL in London and Cambridge University enables dedicated data rates at more than 10Gbit/s for a truly super-fast, yet low-cost, broadband connection to every home.

"UK broadband speeds are woefully slow compared to many other countries, but this is not a technical limitation. Although 300 Mb/s may be available to some, average UK speeds are currently 36 Mb/s. By 2025, average speeds over 100 times faster will be required to meet increased demands for bandwidth-hungry applications such as ultra-high definition video, online gaming, and the Internet of Things," explained lead researcher Dr Sezer Erkılınç at UCL's Electronic & Electrical Engineering department.

"The future growth in the number of mobile devices, coupled with the promise of 5G to enable new services via smart devices, means we are likely to experience bandwidth restrictions; our new optical receiver technology will help combat this problem," he said.

Funded by the EPSRC UNLOC Programme and Huawei Technologies, scientists from the UCL Optical Networks Group and the University of Cambridge developed a simplified receiver that can cost effectively give each subscriber their own frequency band in a fibre.

"To maximise the capacity of optical fibre links, data is transmitted using different wavelengths, or colours, of light. Ideally, we'd dedicate a wavelength to each subscriber to avoid the bandwidth sharing between the users," said Erkılınç. Although this is already possible using highly sensitive hardware known as coherent receivers, they are costly and only financially viable in core networks that link countries and cities. 

"Their cost and complexity has so far prevented their introduction into the access networks and limits the support of multi Gb/s broadband rates available to subscribers," said co-author and Head of the Optical Networks Group, Professor Polina Bayvel at UCL. 

Simplification was achieved by adopting wavelength division multiplexing (WDM), a coding technique to fibre access networks that was originally designed to prevent signal fading in wireless communications and uses quadrature phase-shift keying (QPSK) and 16-quadrature amplitude modulation (QAM). This allowed for a low-complexity coherent transceiver that uses a polarisation-time block coding via digital signal processing (DSP) combined with heterodyne detection. 

This approach has the additional cost-saving benefit of using the same optical fibre for both upstream and downstream data. The transceiver has a 75% lower optical complexity, quantified by the number of balanced detectors, compared to conventional coherent receivers, which potentially eases the monolithic integration whilst preserving a high receiver sensitivity, robustness to chromatic dispersion, and colourless receiver operation due to its frequency/wavelength selectivity.

"This simple receiver offers users a dedicated wavelength, so user speeds stay constant no matter how many users are online at once. It can co-exist with the current network infrastructure, potentially quadrupling the number of users that can be supported and doubling the network's transmission distance/coverage," said Erkılınç. 

The receiver was tested on a dark fibre network installed between Telehouse (east London), UCL (central London) and Powergate (west London). The team successfully sent data over 37.6 km and 108 km to eight users who were able to download/upload at a speed of at least 10 Gb/s. 

www.ucl.ac.uk

Microsemi pushes RISC-V on FPGAs with eco-system of tools

By Nick Flaherty at www.flaherty.co.uk

Chip maker Microsemi has launched a series of IP, tools and partnerships to boost the use of the RISC-V processor core on its FPGAs.

The Mi-V RV32IMA development board and additional field programmable gate array (FPGA)-based soft CPUs are aimed at designs using RISC-V open instruction set architectures (ISAs), and can use other tools such as operating systems in an eco-system for developers.

“Microsemi is the first tier one vendor to build out a complete open RISC-V ecosystem, which not only supports our needs, but contributes to the entire development community,” said Jim Aralis, chief technology officer and vice president of advanced development at Microsemi. “Customers can now select RISC-V for their new designs knowing a tier one vendor committed to the success of this technology is providing all the necessary tools to confidently use RISC-V soft CPUs in their products.”

The open RISC-V ISA benefits from portability as well as enabling the open source community to test and improve cores at a faster pace than closed ISAs. As the RISC-V intellectual property (IP) core is not encrypted, it can be used to ensure trust and certifications not possible with closed architectures. 

Microsemi’s new Mi-V ecosystem brings together a number of industry leaders involved in the development of RISC-V to streamline RISC-V designs for customers.

Design tools include Microsemi’s SoftConsole Eclipse-based integrated development environment (IDE), the firmware catalog and Libero PolarFire system-on-chip (SoC). Operating systems include Express Logic’s ThreadX, Huawei LiteOS and Micrium µC/OS-II. Boards include the RTG4 development kit, IGLOO2 RISC-V board from Future Electronics, PolarFire Evaluation Kit and more. Debug dongles from Microsemi and Olimex, first-stage bootloaders and numerous soft peripherals are also included. Example projects, drivers and firmware are all available on GitHub, the world’s largest repository of open source software.

“Micrium is pleased to join Microsemi’s Mi-V ecosystem with our highly dependable uC/OS-II real-time kernel, a full-featured embedded operating system,” said Jean Labrosse, co-founder and chief architect at Micrium. “As RISC-V continues to grow in popularity, we look forward to working closely with Microsemi to support accelerated adoption of its RISC-V soft CPU product offerings as well as the entire ecosystem’s RISC-V advancements.”

Deployment of soft CPUs implemented with the R11C-V ISA is automatic and delivered to the user’s desktop via Microsemi’s IP Catalog. No end user license agreements are needed to gain access to the soft CPUs. Using RISC-V soft CPUs within the Mi-V ecosystem is simple, easy and free.

“Express Logic is pleased to be a foundational part of Microsemi’s Mi-V RISC-V ecosystem,” said William E. Lamie, President, Express Logic. “Our X-Ware Internet-of-Things (IoT) platform, including the industry-leading ThreadX RTOS with over 6.2 billion deployments, is the preferred embedded software platform for all designs requiring industrial-grade run-time solutions—making us an ideal fit for this new consortium.”

Microsemi is aiming the Mi-V soft CPU cores at IoT, secure communications and wireline bridging applications.

www.microsemi.com/design-support/accelerate-ecosystem-partners.

Related stories:

• SiFive adds linux support to RISC-V CPU cores
• UltraSoC raises £5m for embedded analytics
• Researchers find flaws in RISC-V core
• Imagination to sell off MIPS and ENSIGMA
• Open source hypervisor port to MIPS processors brings more security to the IoT

Samsung merges its Artik cloud with SmartThings in a bid to make IoT work

By Nick Flaherty at www.flaherty.co.uk



Samsung has launched new secure Artik modules and provided a way to integrate them with its SmartThings cloud service which is merging with the current Artik cloud service. This will enable interoperability with both Samsung Artik systems and third-party IoT devices and IoT cloud services.

• Samsung's IoT module taps SiLabs' Gecko technology

“Security in the age of IoT means new levels of complexity and risk. The next generation of IoT products and services will be more deeply integrated into our lives than ever before,” said James Stansberry, Senior Vice President and General Manager of ARTIK IoT, Samsung Electronics. “Unfortunately, most companies are not prepared to address the challenges of securing every link of the chain, from device to IoT cloud. With the ARTIK IoT Platform and our new security hardened system-on-modules, we make it easier and more affordable for companies to adopt best security practices and deliver trustworthy products that will shape the future of IoT.”

The new ARTIK secure IoT modules combine hardware-backed security with pre-integrated memory, processing, and connectivity for a broad range of IoT applications, from simple edge nodes like sensors and controllers, to home appliances, healthcare monitors, and gateways for smart factories. This helps protect data and prevent devices from being taken over, disabled, or used maliciously.

“The New ARTIK secure systems-on-modules provide the performance and security features we need for our new Samsung IoT appliances,” said Youngsoo Do, Senior Vice President of Digital Control Group at Samsung Electronics. “Together, Samsung ARTIK platform and SmartThings Cloud will create opportunities to help us to get to market faster and safely deliver new services that will enrich the lives of our customers.”

The modules provide a strong root of trust from device-to-cloud with a factory-injected unique ID and keys stored in tamper-resistant hardware. Samsung’s public key infrastructure (PKI) enables mutual authentication to the cloud to identify each device on the network and support whitelisting. Developers can use the new Secure Boot feature and code signing portal to validate software authenticity on start-up. In addition, the secure IoT modules provide a hardware-protected Trusted Execution Environment (TEE) with a secure operating system and security library to process, store, and manage sensitive resources, including keys and tokens on devices. Information is protected using FIPS 140-2 data encryption and secure data storage.

“Hardening the security of IoT devices once the product has been deployed is extremely difficult,” said Vikrant Gandhi, Digital Transformation Industry Director at Frost & Sullivan. “Manufacturers must use components that have built-in defences for both device and data integrity. ARTIK IoT platform’s new security measures provide this protection from the start, and helps companies operate more safely in the Internet of Things.”

The secure IoT modules will be available on November 30 through Samsung ARTIK partners.

Samsung’s new SmartThings Cloud will unite all of the company’s existing IoT cloud services – Samsung Connect Cloud and ARTIK Cloud – into one consolidated IoT cloud platform. This seamless, open ecosystem will allow better and easier connectivity across an expanded set of devices and services from home to industrial applications.

Samsung continues to try to reach profitability for the technology. This expansion includes the recent launch of the recent ARTIK service to monetize data from interoperable devices and enable an IoT data economy, as well as the addition of an Ubuntu Linux distribution on the high end ARTIK system-on-modules.

Related stories:

• Samsung takes on the IoT service providers
• Samsung starts volume shipments of its R4-based WiFi IoT chip
• Post quantum security becomes an issue for IoT
• South Korea gives away IoT modules as it cuts its monthly data plans by 90%
• Low cost crypto chip to secure the IoT

New LoRaWAN protocol simplifies device deployment and management for global roaming

By Nick Flaherty at www.flaherty.co.uk

The LoRa Alliance has tackled one of the main drawbacks with low power IoT networks in its latest protocol specification.

New features include roaming and separation of backend nodes, which will enable IoT devices to connect to and move between LPWANs around the world. This supports large-scale deployments and enables new global services such as cargo tracking. These are critical elements for effective IoT roll out:

• LoRaWAN 1.1 adds support for handover roaming, and Class B and security enhancements, and Regional Parameters rev. A describes region-specific radio parameters for LoRaWAN 1.1 end-devices 
• LoRaWAN Backend Interfaces 1.0 adds support for decomposing the network into interoperable nodes, as required for inter-vendor roaming 

The support for roaming will allow for large-scale deployments, since vendors will know that their LoRaWAN protocol-based products could potentially operate worldwide. At the same time, the new backend specification provides the protocols that interconnect servers with distinct roles––such as controlling the MAC layer, end-point authentication, or applications––behind the scenes in the core network. Separating these servers allows an open choice of vendors for each element of the value chain rather than requiring an end-to-end, global system.

The 1.1 specification also includes support for handover roaming, which allows transferring control of the end-device from one LoRaWAN network to another. Earlier versions of this specification can already be used for passive roaming, which is transparent to the end-device. 

Bidirectional end-devices with scheduled receive slots (Class B) are part of the specification enhancements and are now officially supported. 

With enhancements for additional security hardening, both LoRaWAN 1.1 end-devices and networks will support backward compatibility to interoperate with their LoRaWAN 1.0.x legacy peers. This will support heterogeneous deployments and not force a globally coordinated upgrade.

The Backend Interfaces 1.0 Specification includes the ability to break down the network into network server (NS), join server (JS) and application server (AS) and enables roaming for both LoRaWAN 1.0.x (passive roaming only) and LoRaWAN 1.1 networks (both passive and handover roaming). It allows the server to identify the entity that stores end-device credentials (including root keys) as JS and can be separated from networks and administered by an entity independent of the networks that the end-device may be using. This allows networks to offload the authentication procedure to a dedicated system, which can also be operated by a third party. This third-party JS also enables an end-device to be manufactured without having to be personalized for the networks it may eventually be connecting to. 

The alliance met in China this week, backed by ZTE. "Based on the LoRaWAN specification, ZTE CLAA leads the operation of the world's largest application alliance (850+ members, 100+ application types, 40+ commercial applications). With its strategic partners, ZTE CLAA tries to expand the city-level LoRaWAN network operation; with terminal and application partners, ZTE CLAA provides abundant, comprehensive IoT solutions,” said the company 

 “The importance of China in the global electronics supply chain is a key reason we are bringing the Alliance members together in Suzhou,” said Geoff Mulligan, chairman of the LoRa Alliance. “With today’s specifications extending the reach of the LoRaWAN protocol, now is an ideal time to engage with our colleagues in China to address their market-specific requirements. Only through the collaboration of all of our global ecosystem in a single standard can we advance members’ interests and drive the ongoing adoption of LoRaWAN technology.”

www.loraalliance.org

Related stories:

• Semtech delivers LoRaWAN IoT networking via ARM's mbed dev tool
• LPWAN technologies open up IoT opportunities
• Arkessa teams with telent for German LPWAN IoT network
• ST launches low cost LoRa dev kit for the Internet of Things
• LoRa module shrinks in size for the IoT

Synopsys buys fuse-based OTP memory supplier

By Nick Flaherty ay www.flaherty.co.uk

Synopsys has bought Sidense, a leading provider of one-time programmable (OTP) non-volatile memory (NVM) for automotive, mobile, industrial and Internet of Things (IoT) applications. 

The embedded one-transistor OTP technology is based on its patented split-channel 1T-Fuse bit-cell architecture that provides better yield, higher security, improved reliability and lower overall product cost. Sidense's antifuse-based OTP NVM technology offers ultra-low power, small area NVM IP with read access times as fast as 10 nanoseconds. A key advantage to Synopsys as a design tool adn IP supplier is that the IP can be manufactured in standard-logic CMOS fabrication processes and does not require any additional mask layers or process steps. The IP is in high-volume production in automotive, mobile and industrial applications.

The technology will be added to the DesignWare Multi-Time Programmable (MTP) NVM IP solution with OTP NVM IP in 16-bit to 1.28-Mbit configurations. This gives Synopsys access to proven, market-leading OTP NVM IP in process technologies from 180- to 16‑nm along with a team of highly experienced R&D engineers.

"Sidense's OTP NVM IP offers designers a secure, area-efficient solution for fuse replacement, secure key storage, device ID, analogue trim, and code storage," said Joachim Kunkel, general manager of the Solutions Group at Synopsys. "By adding OTP NVM from Sidense to our DesignWare IP portfolio, Synopsys provides designers with a broader NVM IP solution that offers small area, fast access times and high reliability for their SoC designs."

 www.synopsys.com/designware.

AMC processing module adds flexible RapidIO to increase bandwidth

By Nick Flaherty at www.flaherty.co.uk

CommAgility, now part of the Wireless Telecom Group, has launched a high performance DSP processing module in the compact Advanced Mezzanine Card (AMC) form factor. 

The AMC-4C6678-SRIO is an update of the previous AMC-4C6678, adding 20Gbps Gen2 RapidIO (SRIO) to an AMC.4 compliant backplane, to increase bandwidth and flexibility. RapidIO complements the module’s full Gigabit Ethernet infrastructure.

“By adding RapidIO, the AMC-4C6678-SRIO module increases the options for fast, flexible, IO,” said Edward Young, Vice President and General Manager at CommAgility. “Our lead customer for the card is using it alongside other CommAgility RF and DSP AMCs in a high performance specialised LTE application.”

The AMC-4C6678-SRIO offers a wide range of connectivity to the backplane and front panel. The RapidIO and Gigabit Ethernet connectivity use on-board switches for maximum flexibility and access to all DSPs on the module.

The AMC-4C6678-SRIO uses four TMS320C6678 fixed- and floating-point octal core DSPs from Texas Instruments (TI) running at 1.25 GHz, giving a total performance of 640 GFLOPS and 1280 GMACS. It is ideal for a range of high performance DSP applications, including wireless telecoms, image sensor processing, transcoding and stepper control.

The software available for use with the module includes full board support libraries and an IP stack with Telnet/TFTP for Ethernet-based board control and upgrade. Additional debug support is provided via on-board serial port connectors, plus RS-232 and JTAG debug via the CommAgility AMC-BB Debug Breakout Board.

The compact full-size, single width PICMG AMC.0 R2.0 AMC module can be used in both ATCA carriers and MicroTCA chassis. It can also operate standalone, with suitable power and cooling. A range of build options is available, and further customisation is possible in volume, to enable the best technical and commercial fit to a customer application to be achieved.

www.commagility.com.

Power news this week

By Nick Flaherty at www.flaherty.co.uk

POWER NEWS from eeNews Europe


. LG Chem delays European battery megafactory


. EU warns on lack of battery manufacturing in Europe


. Heliatek raises €15m to take its solar films to China

TECHNOLOGIES TO WATCH

. Pepsi backs paper-based thermoelectric generator for wearables

. Intel ships first 17 qubit chip to Dutch researchers

. Machine learning gives ultrafast atomic measurements of perovskite solar cell

NEW POWER PRODUCTS
. 48V battery targets smaller cars


. Drop in replacement for TO-220 linear regulator in harsh environments


. First USB-C buck-boost regulator replaces two converters


TECHNICAL PAPERS
. Plastic optical fibre for battery management systems in 48 V powertrain architectures


. Protection against thermal runaway


. CUI: Medical Design Standards for Power Supplies to IEC-60601-1

ISARA and Utimaco team for post quantum hardware module

By Nick Flaherty at www.flaherty.co.uk

ISARA in Canada has teamed up with Utimaco in Germany to test out different encryption technologies that could be used to secure the Internet of Things in a world with quantum conputers.

The two have successfully run quantum-safe hash, lattice and multivariate-based digital signature algorithms for a Hardware Security Module (HSM) in initial laboratory testing.

Utimaco, based in Germany, is a leading manufacturer of Hardware Security Modules (HSMs) that provide the root of trust to all industries, from financial services and payment to the automotive industry, cloud services to the public sector. HSMs help keep cryptographic keys and digital identities safe in order to protect critical digital infrastructures and high value data assets. 

• Post quantum security becomes an issue for IoT

“With the threat of quantum computers being able to break currently used public-key cryptography growing nearer and standards organizations, like the National Institute of Standards and Technology (NIST) and European Telecommunications Standards Institute (ETSI), working on creating new cryptographic standards, work on how these new schemes fit into existing protocols and systems needs to be done in parallel in order to ensure we are ready to mitigate the quantum threat in time,’’ said Thorsten Groetker, Chief Technology Officer of Utimaco. “In addition, it takes time and effort to ensure that hardware systems are able to run the new algorithms efficiently and that is why we are preparing Utimaco’s CryptoServer line of products for the challenge ahead.”

With immense computing power driven by quantum mechanics, quantum computing offers the potential for huge advancements in areas such as artificial intelligence and material design. But a quantum computer also will be capable of breaking the public key encryption standards that protect all of the data, software updates and technology we now safely store, share and use.

“Quantum computing will deliver amazing benefits and, in the wrong hands, could present serious threats,’’ said Michael K. Brown, CTO of ISARA Corp. “All organizations with critical data need to begin preparing today by protecting information whose confidentiality must persist over time. We are excited to work with Utimaco on quantum-safe HSM solutions as it allows the IoT industry to ensure their product’s roots of trust remain secure and over-the-air updates authentic.”

At the core of secure devices is a root of trust, typically in the form of an embedded public key or digital certificate. These are embedded in all Internet-enabled devices, from computers to vehicles to industrial equipment. The software updates sent to these devices are digitally signed and these signatures can be verified by the receiving devices using the embedded roots of trust.

Updating roots of trust is an urgent problem. A part of solving that problem is getting the quantum-safe algorithms to work on a Hardware Security Module (HSM). This module keeps the private keys safe in a tamper-resistant hardware and sign data that is fed into it, without revealing the private key. While providing physical protection to the private key, these HSMs have limited memory and computing resources.

Related stories:

• Quantum sensor market to top $330m by 2025
• SK Telecom commercialises quantum random number generator chip for IoT security
• Boosting post-quantum IoT security

Sony plans its first USB3 industrial camera module

By Nick Flaherty at www.flaherty.co.uk

Sony Europe's Image Sensing Solutions division is planning its first USB3 camera module as part of its move to CMOS image sensors.

The first of the new modules is the XCG-CG160 which has a Gigabit Ethernet interface and a 1/3-type IMX273 sensor. Sony sees this as a low-disruption way to move from CCD to Global Shutter GSCMOS to replace cameras using the highly-regarded Sony ICX445 CCD sensor. The GSCMOS camera starred shipping in January 2016.

The first module runs at 75 frames per second in SXGA resolution with 8bit colour. Black/white modules are available immediately and colour modules will enter mass production in December 2017.

A version using the USB3.0 protocol is scheduled to enter mass production by Q1 2018.

The modules are aimed at embedded applications such as print, robotics and inspection to ITS, medical and logistics, as well as being suitable for general imaging.

The XCG-CG160 provides a simple migration path from CCD to GSCMOS without having to upgrade or change architecture. The IMX273 shares comparable sensor and pixel size characteristics with the ICX445, but offers improvements in sensitivity, dynamic range, noise reduction and frame rate. These include defect-pixel correction, shading correction with both peak and average detection and area gain to automatically adjust for the target object. The module also supports 2x2 multi-region of interest, flip and binning.

The series uses a 1/2.9-type Sony Pregius GS CMOS sensor, delivering 1440 x 1080 (1.6 MP) images. The b/w module has a minimum illumination of just 0.5lx; the colour module requires just 12lx and comes with a manual, auto and one-push white balance setting. Both modules have a sensitivity of F5.6, a gain of 0 to +18 dB, a shutter speed of 60 s to 1/10,000 s.

"These cameras extract the best possible image from the IMX273 with additional performance enhancing unique features. With it, we believe, we have set the industry standard for an SXGA module," said Matt Swinney, Senior Marketing Manager, Sony Image Sensing Solutions.

Sony's v1.1 firmware is certified by Cognex VisionPro and is GigE Version 2.0 compliant; allowing both hard and software triggering, with the module able to act as a both an IEEE1588 (PTP) master and slave. The firmware uniquely allows for up to 16 action cue commands, plus acquisition sequence scheduling and GPO control.

The firmware also supports burst trigger and new event modes, as well as a reduced jitter function with single frame acquisition. An asynchronous image transfer enables a memory shot and output upon user demand, and its broadcast register write function enables synchronised camera feature controls to be implemented.

The C-mount module uses a highly compatible GPIO with 1xISO IN, 1xISO OUT and 1xTTL IN/OUT. It measures 29 x 29 x 42 mm, with a weight of 65g and has an operating temperature of -5°C to +45°C.

www.image-sensing-solutions.eu/ 

NXP and Synaptics team with Arrow for Alexa smart home dev kit

By Nick Flaherty at www.flaherty.co.uk

Synaptics and NXP Semiconductor are using Arrow Electronics to distribute a variant of the Synaptics AudioSmart 2-Mic Development Kit for Amazon AVS that is supported by a commercially available NXP system-on-chip (SoC) application processor.

This variant enables device makers to quickly and cost-effectively integrate the Amazon Alexa Voice Service (AVS) into a fully functioning prototype, and then take that work into design for manufacturing and production.

“Voice as a human interface is rapidly expanding in smart homes and the broader IoT landscape due to the broad global expansion of voice services like Amazon Alexa,” said Saleel Awsare, vice president and general manager, Audio and Imaging Business, Synaptics. “When voice services are combined with Synaptics’ AudioSmart far-field voice DSPs, new voice-enabled human interface devices can accelerate into numerous IoT applications whether consumer, enterprise, or commercial.”

The Synaptics AudioSmart 2-Mic Dev Kit for Amazon AVS includes Synaptics’ CX20921 Voice Input Processor with a dual microphone board and a wake-word engine tuned to "Alexa." It is built on the pico-imx7d board, a purpose-built small-footprint hardware platform which features NXP’s i.MX7D dual-core applications processor optimized for the internet of things (IoT). The kit also includes stereo microphones with an integrated processor, a microphone holder board, a micro USB cable, Type-A to Type-B cable, cable assembly and a +5V power supply.

The i.MX7D has dual Arm Cortex-A7 cores at speeds of up to 1.2 GHz, as well as the Arm Cortex-M4 core. It supports multiple memory types including 16/32-bit DDR3L/LPDDR2/LPDDR3-1066, Quad SPI memory, NAND, eMMC, and NOR. Several high speed connectivity connections include Gigabit Ethernet with AVB, PCIe, and USB. Both parallel and serial Display and Camera interfaces are provided, as well as a way to directly connect to the Electrophoretic Displays (EPD).

The development kit features a reference client built using the AVS Device SDK (software development kit), enabling commercial device makers to get to market faster with Alexa-enabled products.

“This new Synaptics development kit with an NXP SoC for the Amazon Alexa Voice Service can add intelligent voice control to connected products with a microphone and speaker, enabling developers and engineers to rapidly create Alexa-enabled proof of concepts and prototypes,” said Murdoch Fitzgerald, vice president of semiconductor marketing at Arrow.

The Synaptics AudioSmart 2-Mic Dev Kit for Amazon AVS with NXP SoC is available exclusively from arrow.com for $375.

www.arrow.com/en/products/synapticsnxp2micavs/arrow-development-tools.

ST targets first Chinese Cloud-to-Node IoT platform

By Nick Flaherty at www.flaherty.co.uk

STMicroelectronics has teamed up with Alibaba Cloud, the cloud computing arm and business unit of e-commerce group Alibaba to provide a complete cloud-to-node IoT solution.

The cooperation will allow any designer to easily create IoT nodes and gateways using ST’s key semiconductor building blocks for the IoT, run Alibaba’s IoT operating system AliOS on the node, and seamlessly connect with the Alibaba Cloud for faster time-to-market of large-volume IoT applications.

With the official release of AliOS, ST has made the AliOS available to run on its ARM Coretex-M-based STM32 microcontrollers (MCU). Beyond processing, designers will get access to ST’s complete portfolio covering all the key building blocks for the IoT, including connectivity, security, sensors, power management, and signal conditioning. ST’s modular and interoperable IoT development platform spans state-of-the-art semiconductor components, ready-to-use development boards, free software tools, and common application examples.

The ST/Alibaba collaboration assures development efficiency and access to the first Chinese IoT turnkey platform, bringing greater convenience and productivity to developers when creating IoT nodes and to integrators when connecting to the Alibaba Cloud. Customers also get a head-start on time-to-market with AliOS’s pre-validation of all software components on the platform, easing security provisions and field device upgrades.

“By collaborating with Alibaba, we can provide developers in China with the STM32 platform to build IoT nodes running AliOS. This gives customers a head-start to using the combination of AliOS Node-to-Cloud complete vertical solution and the broad portfolio of ST’s IoT technologies that includes sensors, processing, security, connectivity, and power,” said Arnaud Julienne, Vice President, Microcontroller, Memory, and Secure MCU & IoT, Asia Pacific Region, STMicroelectronics.

“Alibaba, a rich and powerful platform for everyone to conduct business, will help different IoT market players, both large and medium/small businesses, to find their customers or vendors. Alibaba is cooperating with ST to build joint solutions from Cloud to Node, to include security in applications, and we definitely believe AliOS and ST chipsets can help customers achieve their targets in the IoT mass-market,” said Ku Wei, general manager of IOT BU of Ali Cloud BG.

www.stmicroelectronics.com

Related stories:

• TiTAN wants to create an OS for the smart home
• Foghorn raises $30m to boost IoT development
• Intel widens security for automated IoT rollout
• STMicroelectronics teams with Objenious
• First Sigfox cellular hybrid IoT chip

MIPI moves into automotive

By Nick Flaherty at www.flaherty.co.uk

The MIPI Alliance,has formed an Automotive 'Birds of a Feather' (BoF) Group to get industry input from original equipment manufacturers (OEMs) and their suppliers to enhance existing or develop new interface specifications for automotive applications. 

The group is open to both MIPI Alliance member and non-member companies to represent the broader automotive ecosystem.

While MIPI stared out developing interface specifications for mobile and mobile-influenced industries, cars have become a new platform for innovation, and manufacturers are already using MIPI Alliance specifications as they develop and implement applications for passive and active safety, infotainment and advanced driver assistance systems (ADAS). 

MIPI interfaces such as Camera Serial Interface 2 (MIPI CSI-2), Display Serial Interface (MIPI DSI) and Display Serial Interface 2 (MIPI DSI-2) are used for a variety of low- and high-bandwidth applications that integrate components such as cameras, displays, biometric readers, microphones and accelerometers. MIPI I3C helps automotive systems designers use multiple sensors in a space-constrained form factor. Highly sensitive, mission-critical automotive applications also benefit from MIPI interfaces' low electromagnetic interference (EMI).

However interfaces such as MIPI CSI are designed for short distance links in cell phones up to 30cm, The Automotive BoF Group's initial focus will be to examine how MIPI specifications can potentially be extended to support communication link distances up to 15m, and at the same time support the high data rates associated with cameras and radar sensors for autonomous driving systems.

"Automakers already rely on MIPI Alliance's industry-standard interfaces to enable a wide variety of applications, including collision mitigation and avoidance, infotainment and navigation," said Matt Ronning, chair of the MIPI Alliance Automotive Subgroup and the Automotive BoF. "This call for participation helps ensure we cast a wide net to capture expertise to aid with extending existing and shape future MIPI specifications and collectively help realise the vision of how connected cars and automotive applications will evolve over the next decade. Just as mobile handset manufacturers benefited from the standardisation that MIPI Alliance has provided, automotive OEMs would similarly benefit."

"Active participation of automotive OEMs, tier-one and tier-two suppliers is greatly appreciated and necessary to, for example, work out the data link requirements between surround sensors, electronic control units, actors and displays for driver assistance and autonomous driving projects beyond 2020 and incorporate them into MIPI interface specifications," said Uwe Beutnagel-Buchner, vice chair of the MIPI Alliance Automotive Subgroup and the Automotive BoF.

The MIPI Automotive BoF is seeking additional qualified experts from OEMs, tier-one suppliers, component suppliers and related companies to provide key input into current and future MIPI interface specifications. The Automotive BoF is expected to convene via teleconference on a biweekly basis, with face-to-face meetings planned as necessary.

Companies already participating in MIPI Alliance’s Automotive BoF Group include: Analog Devices; Analogix Semiconductor; BitSim; BMW; Cadence Design Systems; Continental; Etron Technology; Ford; Microchip Technology; Mobileye, an Intel Company; NVIDIA; NXP Semiconductors; ON Semiconductor; Parade Technologies; Qualcomm; Robert Bosch; Sony; STMicroelectronics; Synopsys; TE Connectivity; Tektronix; Teledyne LeCroy; Texas Instruments; Toshiba; Western Digital and others.

To learn more and submit an interest form, visit: http://bit.ly/2wOqycY

Analog Devices taps imec research for the IoT

By Nick Flaherty at www.flaherty.co.uk

Analog Devices (ADI) has signed a strategic research partnership with imec in Belgium to develop the next generation of Internet of Things (IoT) devices. The two already have two programmes underway for low power systems with improved—or completely new—sensing capabilities. 

The IoT continues to grow and mature, and by 2020, it is expected to consist of billions of connected, “smart” objects that rely on unobtrusive sensors to constantly monitor the environment, provide status reports and receive instructions. By intelligently processing the data gathered, these devices then drive some type of short-term or long-term action. Today, however, the underlying sensors, and the chips upon which they are built, are often too big, too expensive and not accurate enough to be practical.

One joint research initiative focuses on localization technology.

“Building on imec’s world-leading position in innovative ultra-low power implementations, ADI and imec will pursue the development of a low-power sensor for highly accurate indoor localization in the context of smart building or smart industry solutions,” said Kathleen Philips, imec program director. “Concretely, we want this sensor to localize objects with a superior accuracy in a robust manner and achieve up to five times better accuracy than today’s best-performing solutions.”

A second initiative includes the creation, and ultimately the commercialization, of a highly-integrated liquid sensor that can be used in a variety of application domains, such as the analysis of water, blood or urine.

“Our single-chip sensor comprises multiple electrodes and excels in terms of cost and size, while demonstrating industry leading sensitivity and accuracy,” she added.

“In ADI we have found a partner that brings to market professional-grade, high-value-add sensor systems. Thanks to ADI’s commercial insights, in combination with its innovative mindset, we can build differentiating technology that meets IoT market requirements today and tomorrow,” added Rudi Cartuyvels, Executive Vice President at imec.



Prototype low-power integrated liquid sensor



Prototype low-power higlhy-accurate indoor localization sensor


The strategic partnership between ADI and imec is part of a broader, open research initiative to which other research partners are also invited to contribute.
www.imec-int.com.

Europe's first 5G network goes live with 2Gbit/s, 3ms links

By Nick Flaherty at www.flaherty.co.uk

For transmitters in Berlin from Deutsche Telekom have gone live with 5G New Radio links.

While these are highlighting connections at 2Gbit/s and carrying live transmissions of ultra-high-definition video, the key factor for embedded network designers is the latency of just 3ms.

To accelerate the development of new applications for 5G, Deutsche Telekom recently launched the Low Latency Prototyping Programs at its hub:raum sites in Berlin and Krakow. The programs are aimed at innovative developers who want to develop products using edge computing, as well as 5G network performance as it emerges. 

The four radio cells in the Deutsche Telekom network are the first in Europe with a live 5G connection in a real world setting. Using pre-standard 5G New Radio (5G NR), the network is achieving throughput of more than two gigabits per second to a single customer device, as well as a latency of three milliseconds on commercial sites in Berlin's Schöneberg district.

Deutsche Telekom is using the 3.7GHz band with Huawei equipment for its implementation in Berlin. The four Huawei antennas are located at three different sites: Winterfeldtstraße (two cells), Martin-Luther-Straße and Pohlstraße.

Arqiva and Samsung are planning to launch a 5G fixed network in London at 28GHz shortly

“5G is getting close to prime time. The successful live transmission over 5G is an important technical contribution to the 5G community for the two companies. Huawei will continue investing in Research & Development and enhance its cooperation with partners to make progress in 5G technology and the development of an industry ecosystem,” said Deng Tai Hua, president of the Huawei Wireless Network Product Line.

"We are demonstrating 5G live here, in the middle of Berlin, rather than in a lab. This is a very decisive developmental step on the way to the global launch of 5G, which is planned for 2020," said Claudia Nemat, Member of the Management Board responsible for Technology & Innovation at Deutsche Telekom. "If everything is connected to everything else, customers need a high-performance, reliable and secure network. Industry in particular will benefit from 5G as a powerful enabler for a wide range of applications. We are ready for 5G."

Deutsche Telekom demonstrated the network's high performance by showcasing some early application ideas that are enhanced by 5G. The augmented reality (AR) and virtual reality (VR) applications benefit from 5G’s high data rates and real-time responses.

The new network architecture and design will also play a significant role as they will be controlled by software, will be programmable and will be managed end-to-end. This creates the enabling platform for the provision of a large number of different services.

5G New Radio, the new 5G air interface, will work in combination with the evolution of 4G/LTE technology using massive MIMO (multiple input/multiple output) antenna elements deployed at both the base station and in the subscriber device, which can increase capacity through a much more efficient use of the frequency spectrum.

"5G New Radio in Berlin is the first major step towards 5G for everybody," said Deutsche Telekom CTO Bruno Jacobfeuerborn. "As soon as the standard is defined and is available, we will proceed in 2018 to lay the foundation for large-scale build-out. This is how we are driving the technology development in Europe and demonstrate our innovation power.”

The specifications in the Huawei equipment used are following the progress of the global 3GPP specifications for the 5G New Radio standard. It is anticipated that the specifications of the system used will be completed by 3GPP in December 2017.

Deutsche Telekom is piloting the use of 5G for industry at the Port of Hamburg. The aim is to tailor the 5G network to meet the diverse requirements of the port. Traffic control data, reliable traffic light steering and environmental monitoring are the first application scenarios for 5G. This enables the digital transformation of the port to support its future growth.

For more information, see Deutsche Telekom's 5G Special.

Related stories:

• Qualcomm teams with Nokia for 5G radio deployment
• Arqiva and Samsung to test out early fixed 5G technology at 28GHz
• 5G NR massive MIMO network goes live in Singapore
• ZTE makes move to dominate IoT and 5G
• Intel launches 5G development platform ahead of December standard

Power news this week - Samsung under fire

By Nick Flaherty www.flaherty.co.uk

POWER NEWS from eeNews Europe Power Management
. Dialog buys Silego for IoT bid
. Tessera sues Samsung and chip distributors in Europe


. UK sets up £65m battery research institute

. Asphalt holds promise for high capacity, fast charging lithium batteries


. Toshiba doubles driving range of fast charge lithium batteries


. Boosting battery pack production with laser welding and 3D printing

. Thermal gels help cool batteries

. GaN DC-DC converters reach 98% efficiency
. Chip-style DC-DC power with 1% regulation

. Protection against thermal runaway

. Primary protection against overvoltage events
. CUI: Improved Longevity and Performance in DC Fans

SiFive adds linux support to RISC-V CPU cores

By Nick Flaherty at www.flaherty.co.uk

Open source CPU core designers SiFive have launched the first multi-core for embedded designs that uses the RISC-V architecture and supports the linux operating system.

The quad core U54-MC Coreplex IP is based around the MIPS-like RISC-V open source architecture, which is supported by an ecosystem comprising more than 70 companies and has seen significant growth in the embedded segment. Part of this has come from the uncertainty around the future of the MIPS architecture.

• Reinventing MIPS as an open source chip venture
• Imagination/MIPS broken up and sold to two venture capital firms

The release of the U54-MC Coreplex marks the architecture's expansion into the application processor space and wider IoT applications such as gateways.

The U54-MC contains four U54 CPUs along with a single E51 CPU, and is the first commercial RISC-V core to include multicore support and cache coherence. Each U54 CPU has a highly efficient five-stage in-order pipeline supporting the RV64GC ISA, which is expected to be the standard for Linux-based RISC-V devices.

The 64-bit E51 CPU serves as a management core and is fully coherent with the main U54 cores. This makes it suited to applications which need full operating system support such as AI, machine learning, networking, gateways and smart IoT devices.

"The ability for RISC-V developers to develop Linux and other Unix-based operating systems on commercial grade silicon will enable the RISC-V software ecosystem to quickly expand beyond embedded systems, and bring new solutions to market," said Rick O'Connor, executive director of the non-profit RISC-V Foundation.

The Coreplex has been taped out as part of SiFive's Freedom Unleashed family of high-performance, customisable RISC-V SoCs. This has the U54 and E51 CPUs running at 1.5 GHz in 28nm technology. Each of the U54 CPUs implement a 32KB Instruction Cache and 32KB Data Cache, and all of the cores share a coherent, 2MB L2 Cache. Customers can license the U54-MC Coreplex in a variety of configurations besides the 4+1 default configuration.

"The U54-MC Coreplex is the first fully Linux-compatible CPU based on RISC-V. It takes the industry one step closer to making custom silicon available to everyone," said Andrew Waterman, co-founder of SiFive. "We continue to be amazed by the support SiFive has received since we launched the industry's first open-source RISC-V SoC last year, and look forward to additional milestones in the coming months."

The U54-MC Coreplex can be accessed at https://www.sifive.com/products/coreplex-risc-v-ip/u54-mc/. A development board based on U54-MC Coreplex IP will be available in Q1 2018. 

www.sifive.com/products/coreplex-risc-v-ip/.

Related stories:

• UltraSoC raises £5m for embedded analytics
• Researchers find flaws in RISC-V core
• MIPS wins Mediatek LTE modem chip slot
• Imagination to sell off MIPS and ENSIGMA
• Sierraware brings its embedded security to MIPS
• Open source hypervisor port to MIPS processors brings more security to the IoT
• Barco Silex and Imagination collaborate on IoT chip security
• prpl Foundation teams with CABA for MIPS-based security in the home

HPC moves into embedded with rugged server designs

By Nick Flaherty at www.flaherty.co.uk


The Crystal Group is launching a new range of high end rugged liquid-cooled servers for embedded applications.

The Fully Optimized Rugged Computer Equipment (FORCE) combine the latest Intel Xeon Scalable Processors with high-speed networking, enhanced I/O, platform security, and thermal management in a modular, customisable high-performance computing (HPC) system.

Those with long memories may remember Force Computers, bought by Motorola in 2004 from Solectron and then becoming part of Emerson Network Power, now owned by a venture capital firm and renamed Vertiv. 

Crystal's FORCE rugged servers have single or dual Intel Xeon Scalable Processors (Purley/Skylake), each supporting up to 24 cores and 48 PCIe lanes. The Intel Advanced Vector Extensions 512 (AVX-512) ultra-wide vector processing simplifies repetitive tasks. The Platform Controller Hub (PCH) offers 10 Gigabit Ethernet (10GbE) ports for high-speed communication, up to 14 SATA 3 ports for expanded storage capacity, and QuickAssist Technology (QAT) for enhanced security, authentication, and compression.

The RS1104, RS2608, and RS3712 Rugged Servers are available now. The newly designed, lightweight chassis come in 1U, 2U, and 3U sizes with 19-, 22-, and 25-inch depths, with a choice of ATX or eATX motherboard, and air- or liquid-cooled thermal management. Front-end configuration options include serial and USB ports, 2.5-inch drives, CMOS battery, pump assembly for liquid cooling, circuit breaker, card reader, USB sound card, sanitize button for security, DVD, and LCD. 

"The Crystal Group FORCE RS1104L22 can dissipate an unprecedented 280 watts of CPU power with its state of the art liquid cooling. It is the world's first and only Xeon Skylake powered server capable of such a feat," said Jim Shaw, Crystal Group Executive Vice President of Engineering.

"The FORCE line comes in direct response to customer requests for faster processing speeds and greater storage capacities in a rugged enclosure with advanced thermal management – optimized for size, weight, power, and cost (SWaP-C)," said Shaw. "We continue to expand the Crystal Group portfolio of trusted rugged solutions to provide precisely what's needed on today's multi-domain battlefield: robust, rugged, and reliable systems capable of withstanding harsh environments and tackling complex, compute-intensive applications, such as sensor fusion, machine learning and artificial intelligence, autonomy, electronic warfare, and so much more."

Crystal Group is demonstrating 1U, 2U, and 3U air- and liquid-cooled configurations of the RS1104, RS2608, and RS3712 in the US this week. 

The servers meet or exceed IEEE, IEC, and military standards (MIL-STD-810, 167-1, 461, MIL-S-901) and are backed by warranty (5+ year) with in-house support.

crystalrugged.com

Related stories:

• King Cobra rugged server design provides a “total rack” in only 2U
• IoT drives software defined power into the data centre
• Linaro launches 64bit ARM-based developer Cloud to go from server to node
• Updated: Amazon pushes the trend for vertical integration
• Starter set for COM Express Type 7 server-on-modules triggers modular micro server designs

TiTAN wants to create an OS for the smart home

By Nick Flaherty at www.flaherty.co.uk

Korean startup TiTAN has launched a smart platform for the home based around secure content in the US.

The 'TiTAN Core' home hub has unique smart controls that give users unlimited access to their smart devices, says the company, It wants to use this as the basis for an IoT operating system across the home with its TiTAN Content Identifier (TCI) Encryption.

"The time has come for content, platforms, networks and devices to be unified. We will continue to expand our business here in the US and around the world providing smart services that will help users be more connected to their content and smart home devices," said Jae Young Yoon, founder of TiTANplatform. "Backed by our TCI patented encryption technology, we have created a secure ecosystem through which consumers can safely integrate their entire home through one device, and content producers can reach viewers on a highly transactional platform that protects their intellectual property."

TiTAN Play allows content owners to securely upload their IP and have access to consumers through a secure content marketplace. TiTAN Core is a smart home hub device married with a lean-back living room entertainment experience, allowing consumers to manage their connected or smart Internet of Things (IoT) devices and appliances.

"With TiTAN Play and Core, we are creating a new 'OS for the Home' - a unified ecosystem that provides a smarter content experience and device connectivity. Consumers are increasingly buying smart IoT devices to improve their lives, yet few of these devices truly talk with each other," said Adrian Sexton, CEO of TiTAN Platform US. "TiTAN Core's open platform changes that and does it with security being paramount. With TiTAN Play, content creators and studios will be able to monetise their content with a major access point in China - a difficult entry market for YouTube, Netflix, Facebook and Amazon - thanks to our joint venture with China Unicom.."

Content creators will also be able to quickly launch SVOD (Subscription Video on Demand) channels by uploading content directly through either the TiTAN Core smart device in their homes or via TiTAN Play on their mobile devices.

The company is currently raising $100 million on top of $32 million raised to date out of Asia.

www.titanplatform.com

Surge in processors for tablets

By Nick Flaherty at www.flaherty.co.uk

After two consecutive years of decline, the global tablet Applications Processor (AP) market returned to growth and registered a 5 percent year-over-year growth to reach $984 million in the first half of this year, according to Strategy Analytics Handset Component Technologies service report.

• Tablet Apps Processor Market Share Q2 2017: Intel Improves as Revenues Jump 6 Percent

The report finds that Apple, Intel, Qualcomm, MediaTek and Samsung LSI captured the top-five revenue share rankings in the tablet AP market. Apple as market leader gained share and increased its  revenue share to 37 percent in 1H 2017. Intel and Qualcomm grabbed the number two and three spots with 18 percent revenue share and 16 percent share, respectively in 1H 2017.

“The global tablet AP market continued to show signs of stabilisation from past few quarters with slower year-on-year declines," said Sravan Kundojjala, Associate Director. "Despite this encouraging trend, tablet AP vendors have to contend with a significantly smaller overall market in 1H 2017 compared to two years ago. Strategy Analytics believes that the mature market pushed many of the tablet AP vendors to shift focus from high volume low-cost tablet APs to relatively lower volume but higher value tablet APs. This trend had boosted tablet AP average selling prices (ASPs) in 1H 2017 on a sequential basis.”

Stuart Robinson, Executive Director of the Strategy Analytics Handset Component Technologies service, added, “Strategy Analytics estimates that x86-based tablet AP shipments accounted for 14 percent of total tablet AP shipments 1H 2017, up from 10 percent in 1H 2016. Intel’s focus on Windows tablets paid off. The x86 chip maker Intel retrenched from the low-end Android tablet AP market but capitalised on Windows tablets and gain market share in 1H 2017. Intel is expected to maintain momentum through 2017." 

Aside from Intel, HiSilicon also performed well in the tablet AP market in 1H 2017. HiSilicon’s parent company Huawei increased its tablet shipments in recent quarters and this has helped HiSilicon to build a strong position in the tablet AP market says Robinson.

www.StrategyAnalytics.com

Related stories:

• UltraSoC raises £5m for embedded analytics
• Xilinx boosts Huawei 's cloud server

Foghorn raises $30m to boost IoT development

By Nick Flaherty at www.flaherty.co.uk

FogHorn Systems has raised $30m in its second round of funding to expand its engineering teams in California and Pune, India that are working on Industrial IoT (IIoT) edge computing.

FogHorn recently launched Lightning ML with machine learning at the network edge. This uses FogHorn’s pioneering CEP (complex event processing)-driven edge analytics software that was launched in 2016 to allow industrial customers to train and execute machine learning algorithms on streaming sensor data at the source of the data.

• See also: Moving IoT analytics to the network edge

Intel Capital and Saudi Aramco Energy Ventures led the round, with new investor Honeywell Ventures and all previous investors participating, including Series A investors March Capital Partners, GE Ventures, Dell Technologies Capital, Robert Bosch Venture Capital, Yokogawa Electric Corporation, Darling Ventures and seed investor The Hive. 

This new round brings FogHorn’s total funding to $47.5 million. “This major round of funding by many of the world’s largest and most innovative technology and industrial companies will enable FogHorn to continue its drive for industry-first innovation in the IIoT market segment,” said David King, CEO of FogHorn. “We have seen unprecedented interest from customers and partners in a huge variety of industries for advanced condition monitoring, predictive maintenance, asset performance management and process optimisation solutions.”

“Both the promise and challenge of IIoT lie in the ability to convert sensor data to actionable insights that improve customers’ operating efficiency and generate new sources of business value,” said Jonathan Ballon, Vice President and General Manager of Intel’s IoT Group. “As we’ve watched FogHorn’s progress with both leading IIoT solutions providers and major industrial end customers, we believe their edge analytics and machine learning technology will be a critical factor in enabling those operating efficiencies and delivering that business value.”

“For the oil and gas industry, harnessing IIoT insights will produce significant operating savings and spur major process improvements,” said Cory Steffek, Managing Director North America of Saudi Aramco Energy Ventures (SAEV). “We believe innovators like FogHorn will lead the way in redefining IIoT edge computing to deliver real-time analytics and machine learning value across our entire business — upstream, midstream and downstream.”

With more than 20 billion IoT-connected devices expected by 2020, according to Gartner’s latest projections, GE estimates that the industrial internet market segment will grow to $225 billion in that timeframe. This will require solving the fundamental computing challenge of generating actionable insights from all the real-time streaming data constantly being emitted by the profusion of physical, video/audio and other edge sensors deployed in any industrial environment — whether in manufacturing, energy, resources, transportation, smart cities or smart buildings — without the transport and hosting costs, cybersecurity risk and operational latency associated with sending all that operations technology data across a network to a public/private cloud or remote data centre.

www.foghorn-systems.com

Related stories:

• Intel widens security for automated IoT rollout
• 50 companies join EdgeX open framework for IoT edge computing
• Intel McAfee security deal is all about the IoT this time
• Bots move into industrial IoT
• Actility raises $75m for wide area IoT
• Cloud API for IoT gateways and IoT edge servers simplifies networks...
• Edge analytics vital for security says Greenwave
• First edge analytics integration for VxWorks