Author: Techtime

Israeli Exits Reached $14.48 Billion

Posted on by Techtime

In the first half of 2019, technology exit activity in Israel reached $14.48 billion in 66 deals, including the mega-deal in which Mellanox Technologies was acquired by Nvidia for $6.9 billion (subject to closing). Excluding Mellanox, total exit value reached $7.58 billion in the first half of 2019. Despite a slight decrease in the number of exits, from 73 exits in H1-2018 to 66 exits in H1-2019, the total value in 2019 increased significantly from the total of $6.49 billion in H1-2018.

According to IVC-Meitar Exit Report, the average exit value in H1 2019 set a five-year record, reaching $116.6 million. Almost double compared with $63 million in 2015 annually. Shira Azran from Meitar Liquornik Geva Leshem Tal Law Firm, said that there is increase in the number of growth companies raising large amounts of capital in recent years.“The increase in the value of exits is consistent with the expansion of the backlog of mature companies.”

As a result, the number of deals between $100 million and $1 billion climbed to a record of 23 deals compared with 18 deals in 2018. Analysis of exit value versus amount invested in H1 2019 shows the average ratio recovered to 3.9 compared with 2.91 in 2018 annually.

According to IVC-Meitar Exit Report, the average exit ratio of non-VC-backed companies increased to 13.65 while the VC-backed exit-ratio average increased to 3.7 compared with the annual benchmark since 2015. On average, the ratio in H1 2019 increased, representing higher efficiency of investments in the industry.

Delay in Orders From India Hit Ceragon

Posted on by Techtime

Further delays of orders from India caused a reduction of approximately $20 million in the expected revenues Ceragon Networks. The company revealed that the orders, which the company had expected to be able to recognize during the second quarter of 2019, are now expected in the third quarter. As a result of the delay, approximately $20 million of revenue was pushed out from the second quarter.

“The remainder of Ceragon’s global business was slightly better than management’s expectations during the quarter,” the company said in a press release. Ceragon is currently estimates that its second quarter 2019 results be $71 million to $74 million. “We received one of the smaller orders from India in the last few days, and we expect to receive the other, larger orders in time to recognize the related revenue during Q3 and Q4,” said Ira Palti, president and CEO of Ceragon (photo above).

“These orders are part of an ongoing 4G expansion and densification effort in India that we have been supporting for the past 5 years. The reasons for the delay relate to customers’ internal processes, and the aggregate value of these orders has remained the same since we first began working on the details. Since we don’t expect an acceleration of the deployment schedule, the delay is expected to roll through the remaining quarters of 2019 and push a total of approximately $20 million of revenue related to India from 2019 into 2020.”

Ceragon provides wireless backhaul solutions for mobile networks. Following the announcement, the company’s stock in Nasdaq dropped 9.8%, and it is valued at Market Cap of $220.4 million.

NeuroBlade Raised $27 Million for AI Processor

Posted on by Techtime

Photo above: NeuroBlade’s Founders, the CEO Elad Sity (right) and the CTO Eliad Hillel. Photo: Keren Gafni

NeuroBlade from Hod-Hashron, Israel has completed a $23 million financing round led by Marius Nacht, Co-founder of Check Point Software Technologies. Other investors in this round were Intel Capital, StageOne Ventures and Grove Ventures (headed by Dov Moran, the inventor of the USB flash drive and Co-founder of M-Systems). With the previous raising of $4.5M, the investments in the company totaled $27.5 million. The capital raised will be used to scale workforce and to bring the AI chip to the market.

The company was founded in 2017 by the CEO Elad Sity and the CTO Eliad Hillel, both are veteran of the technological unit of Israel’s Intelligence Corps and formerly key employees at SolarEdge. It designs a new architecture for Artificial Intelligence processors, aimed to provide a fast general purpose AI execution with no memory bottlenecks and no accuracy compromises. The new processor can handle multiple applications simultaneously,  and to to run several neural networks, of any AI topology, at the same time.

NeuroBlade joins a group of new companies that challenge Intel’s and Nvidia’s dominance in AI processing, with new architectures designed from scratch to meet the needs of modern AI tasks. Among them, LG, Qualcomm, and the younger Habana Labs and Hailo from Israel, as well as Wave Computing from California and Cambricon from China.

This market is on a rapid move: According to a recent report published by Allied Market Research (Artificial Intelligence Chip Market), the global artificial intelligence chip market size was valued at $6.64 billion in 2018, and is projected to reach $91.2 billion by 2025, growing at a CAGR of 45.2% from 2019 to 2025. Based on chip type, the GPU segment led the AI chip market in 2018. However, the ASIC segment, representing the new dedicated architectures,  is anticipated to overtake the GPU type in the near future, in terms of revenue.

TowerJazz and Lumotive Demonstrate Solid-State Beam Steering for LiDAR

Posted on by Techtime

TowerJazz and Lumotive from Bellevue, Washington, announced the successful demonstration of the first beam steering ICs for automotive LiDAR systems that are fully solid-state (without any moving parts). The steering concept is Based on Lumotive’s Liquid Crystal Metasurface (LCM) technology, to control the laser beam direction by applying electric fields. The idea was originally developed by  Dr. David Smith, Director for the Center for Metamaterial and Integrated Plasmonics in Duke University.

He had developed a concept called Holographic Beam Forming. By building miniature metal structure on a surfaces (called Metamaterials, or Metasurface), he could change the refractive index of this surface. When these structures are small enough to act like tiny array of antennas, they response to electric fields and allow the control of the refractive index by applying electric signals.  But to implement the idea, Lumotive needed a special IC.

Here TowerJazz came for the help: Lumotive’s beam-steering ICs uses TowerJazz’s 130 nm Cu back-end-of-Line technology, customized to meet specific optical performance requirements with optimized lithography and custom dielectrics. Lumotive’s complete LiDAR system based on this beam steering chip coupled with a custom SiPM (Silicon Photomultiplier) sensor, utilizing TowerJazz’s SPAD (Single Photon Avalanche Diode) technology, will be available for prototype testing in late 2019.

“Our partnership with TowerJazz enabled us to achieve this important milestone and will allow us to bring our revolutionary technology to production,” said Bill Colleran, President & CEO, Lumotive. Research firm Yole Développement estimates that the ADAS and autonomous vehicle LiDAR markets will grow dramatically in the coming years, increasing from $721 million in 2018 to $6.3 billion in 2024, with a CAGR of nearly 45% during that period.

BrightWay Raised $24 Million from Koito Group

Posted on by Techtime

Elbit Systems’ subsidiary, BrightWay Vision, has raised a $25 million from Koito Manufacturing Co., and Magenta Venture Partners. The lead investor was Koito who bought  36.92% of the company for $24 million. BrightWay is an Elbit’s spin off that was established in 2011 as an effort to find new markets for certain military technologies owned by Elbit. The company’s patented system, BrightEye, combines sensing and laser illumination technologies for the automobile industry, to generate a clear long-range image of the road ahead at night and in low visibility conditions.

It is based on Active Gated Imaging Technology developed in Elbit. AGIT is a fast gated-camera equipped with a unique Gated-CMOS sensor produced in TowerJazz, and a laser pulsed Illuminator, synchronized in the time domain. It record images of a certain range of interest which are then processed by computer vision algorithms. It is used for camera-based Advanced Driver Assistance Systems (ADAS) and provides high quality 3D imaged in low/high light-level and harsh weather conditions.

For Koito, this ia a strategic investment. The Japanese Automotive lighting solutions provider, announced plans to utilize BrightWay’s technology in its future products road map. “Koito has decided to invest in BWV and strengthen our relationship in order to promote our R&D of high-precision camera systems. BWV’s  Gated Imaging Technology is superior at night imaging, less susceptible to weather conditions such as rain and fog and capable of distance measurement even with mono cameras.

“By integrating Koito’s automotive lighting technology and BWV’s camera technology, we will work together to conduct research on next-generation automotive lighting equipment.” Tokio based Koito is a major supplier for the Automotive industry. Koito Group consists of 32 companies located in 12 countries worldwide. It employs approximately 24,600 employees, and had achieves annual sales of $7.7 billion in fiscal year 2018.

Habana Labs Announced Super-fast AI Processor

Posted on by Techtime

The semiconductor newcomer, Habana Labs from Israel, is determined to rewrite the rules of AI processors market. The company announced this week the Habana Gaudi AI Training Processor, that is supposed to deliver an increase in throughput of up to four times over systems built with equivalent number GPUs. Gaudi is based on an innovative architecture that enables near-linear scaling of training systems performance and high throughput even at small batch sizes. This allows the scaling from a single-device to very-large systems consist of hundreds of Gaudi processors.

Computing and Connectivity in a Single Chip

Gaudi brings another surprising industry first to AI training: The the AI processor included on-chip integration of RDMA over Converged Ethernet (RoCE v2) functionality, to enable the scaling of AI systems to any size, using standard Ethernet. It means that unlike GPU-based systems rely on proprietary system interfaces, Habana’s systems rely on standard interface. Ethernet switches are multi-sourced, offering virtually unlimited scalability in speeds and port-count, and are already used in data centers to scale compute and storage systems.

“Gaudi offers strong performance and industry-leading power efficiency among AI training accelerators,” commented Linley Gwennap, principal analyst of The Linley Group. “As the first AI processor to integrate 100G Ethernet links with RoCE support, it enables large clusters of accelerators built using industry-standard components.” The Gaudi processor includes 32GB of HBM-2 memory and is currently offered in two forms: HL-200 – a PCIe card supporting eight ports of 100Gb Ethernet; HL-205 – a mezzanine card compliant with the OCP-OAM specification, supporting 10 ports of 100Gb Ethernet or 20 ports of 50Gb Ethernet.

Habana Labs HLS-1 computer combines 8-Gaudi systems in a single rack
Habana Labs HLS-1 computer combines 8-Gaudi systems in a single rack

Habana is also introducing an 8-Gaudi system called HLS-1, which includes eight HL-205 Mezzanine cards, with PCIe connectors for external Host connectivity and 24 100Gbps Ethernet ports for connecting to off-the-shelf Ethernet switches, thus allowing scaling-up in a standard 19’’ rack by populating multiple HLS-1 systems. Gaudi is the second purpose-built AI processor to be launched by Habana Labs. Last year it announced the Habana Goya AI Inference Processor. Goya has been shipping since Q4, 2018, and has demonstrated industry-leading inference performance, with the industry’s highest throughput, highest power efficiency (images-per-second per Watt), and real-time latency.

Facebook is using Habana’s Inference Processor

“Training AI models require exponentially higher compute every year, so it’s essential to address the urgent needs of the datacenter and cloud for radically improved productivity and scalability. Gaudi’s innovative architecture delivers the highest performance while integrating standards-based Ethernet connectivity for unlimited scale,” said David Dahan, CEO and Co-founder of Habana Labs. “Gaudi will disrupt the status quo of the AI Training processor landscape.”

“Facebook is seeking to provide open platforms for innovation around which our industry can converge,” said Vijay Rao, Director of Technology, Strategy at Facebook. “We are pleased that the Habana Goya AI inference processor has implemented and open-sourced the back-end for the Glow machine learning compiler and that the Habana Gaudi AI training processor is supporting the OCP Accelerator Module (OAM) specification.”

Habana will be sampling the Gaudi to selected customers in the second half of 2019. Habana Labs was founded in 2016 in Tel Aviv by silicon experts. In November 2018, it has secured $75 million in series B funding led by Intel Capital. The round brought the total investments in the company to $120 million.

Intel to Open an Acceleration Program for Israeli Startups

Posted on by Techtime

Photo above (from Left): Tzahi Weisfeld, GM of Ignite; Avner Goren, Intel VP in the Architecture, Graphics and Software Group; Bob Swan, Intel CEO; and Yaniv Garty, GM of Intel Israel. (Credit: Ezra Levi)

Intel Corporation today announced a program to accelerate early-stage startup companies in Israel targeting key industry inflection points, including artificial intelligence (AI), autonomous systems and other data-centric technologies and business models. Based in Tel Aviv, the program called Ignite will leverage Intel’s global market access and technology to provide the startups with unique advantages.

Following a rigorous selection process, Intel will host 10-15 top pre-seed to seed startups through a 20-week program where they will receive hands-on mentorship from Intel and industry experts. “Israel has the deep skill base in AI, autonomous systems and the underlying technologies critical to these inflections that make it a natural choice to launch our Ignite program,” said Intel CEO Bob Swan.

The Ignite program will begin operations this year, with plans to scale to additional countries over time. Intel said it has no plans to seek equity in or rights to intellectual property from these companies. Tzahi Weisfeld, former global head of Microsoft for Startups, will serve as general manager and managing director of Ignite, reporting to Yaniv Garty, general manager of Intel Israel. Weisfeld will be assisted by Avner Goren, Intel vice president in the Architecture, Graphics and Software Group, who will connect the Ignite program to the technical community inside Intel.

“Intel is positioned to help companies charge forward,” said Yaniv Garty. “I’m confident that Intel’s expertise in hardware, software and manufacturing will help the startups grow and succeed. The fact that Intel is a large corporation with over 100,000 employees worldwide and very substantial Israeli presence (12,000 employees including subsidiary Mobileye) will greatly help the companies taking part in this program.”