Author: Techtime

Why Anthropic May Pay $6 Billion for Decart

Posted on by Techtime

Above: Decart’s two flagship models: Oasis, which generates interactive worlds in real time, and Lucy, which enables real-time live video transformation. Both run on the company’s DOS optimization infrastructure. Source: Decart AI

AI company Anthropic is in talks to acquire Israeli startup Decart AI for approximately $6 billion, according to Bloomberg. Reuters separately confirmed through a source that discussions are underway, although no deal has been finalized and there is no certainty that one will be reached.

If completed, it would be a massive transaction for a company that only in May raised $300 million at a valuation of approximately $4 billion. Nvidia and SpaceX have also previously been reported as showing interest in Decart. But the more interesting question than the price tag is what exactly Anthropic sees in Decart that could be worth billions of dollars.

At first glance, Decart may look like a generative AI company specializing in video and “world models.” The company developed Lucy, which enables users to transform live video through text instructions — replacing characters or backgrounds and adding objects and effects, for example — while maintaining a frame rate of 30 frames per second.

It also developed Oasis, an interactive world model designed, among other applications, for simulations involving robots and autonomous vehicles. Unlike a conventional simulation engine, the model itself generates the environment and how it evolves in response to actions taken by a user or AI agent.

But these products tell only half the story. Underneath them lies what may be the technology of greatest interest to Anthropic: the Decart Optimization Stack, or DOS.

Decart’s premise is that a significant portion of the computing power available in AI accelerators is not being used optimally. DOS is a software layer that performs deep optimization across the stack — from hardware-aware model design to kernels, memory management, compilers and inference optimization.

The technology supports multiple architectures, including Nvidia GPUs, Google TPUs and AWS Trainium. The objective is straightforward: extract more work from the same chip, allowing AI models to run faster and at lower cost.

Decart already offers these capabilities as a service to chipmakers, hyperscalers and AI labs, allowing customers to bring their own workloads and have kernels and compilers optimized for them. It has also launched Cogito, an inference layer for open-source language models.

According to Decart, its technology can in some cases deliver speeds roughly five times faster than a typical inference provider and exceed 1,000 tokens per second in certain configurations. These are company-reported figures, but they illustrate the kind of asset Anthropic may be targeting.

Getting More From Every GPU

For Anthropic, the technology addresses one of the biggest challenges in the AI business: the enormous cost of compute.

As Claude evolves from a model that simply answers questions into a system of AI agents that perform lengthy tasks, write code, operate tools and repeatedly call the model to “think” about their next step, inference demand rises dramatically. Any improvement in throughput or accelerator utilization can therefore translate directly into substantial cost savings.

Reuters estimated that if Anthropic’s compute spending were to reach approximately $56 billion under a scenario in which the company generates $100 billion in revenue, even a 10% improvement in compute efficiency could theoretically be worth $5.6 billion annually. Viewed from that perspective, a $6 billion price tag for a company capable of materially improving infrastructure efficiency begins to look considerably less surprising.

There is another clue. According to reports about the negotiations, Decart employees are expected to join Anthropic’s Inference and Performance organization. If that happens, it would be a strong indication that the primary objective is not to turn Claude into a video-generation model, but to bring Decart’s expertise in AI model optimization directly into Anthropic’s infrastructure.

From Unit 8200 to Claude’s Infrastructure

Decart was founded in 2023 by Dr. Dean Leitersdorf and Moshe Shalev, who met while serving in Unit 8200, the Israeli military’s elite technology and intelligence unit.

Leitersdorf, the company’s CEO, holds three computer science degrees from the Technion – Israel Institute of Technology, including a PhD, while Shalev serves as chief product officer. The company has raised funding from prominent investors including Sequoia, Nvidia, Benchmark, Radical Ventures and Zeev Ventures.

In its latest funding round in May, Decart explicitly presented DOS as the infrastructure underlying its two main model families: Lucy for interactive experiences and Oasis for Physical AI.

That may be the best way to understand Decart: not as a video AI company that happens to have developed optimization technology, but as an AI optimization company that built real-time video and world models to demonstrate what its infrastructure can do.

The company describes its ambition simply: to make AI operate “at the speed of reality.”

For Anthropic, which is competing in a market where the cost of running a model is becoming almost as important as the model’s capabilities, that technology could be worth a great deal of money.

Stratasys Produces 12,000 Parts for Leading Drone Makers

Posted on by Techtime

Above: Stratasys’ J850 3D printing system. Photo: Stratasys

Stratasys is revealing the growing scale of its involvement in the defense drone industry. During the company’s investor call following its second-quarter results, CEO Yoav Zeif said that Stratasys Direct, the company’s on-demand manufacturing arm, produces more than 12,000 parts for the aerospace and defense sectors, primarily for drone manufacturers, and works with 10 leading drone companies.

Stratasys Direct effectively operates as a manufacturing-on-demand service. Instead of purchasing a printer and producing parts in-house, customers can send their designs to Stratasys and receive finished components. Its U.S. manufacturing operations span several sites and use a range of polymer additive manufacturing technologies, including FDM, SLA, SAF and SLS.

Stratasys did not disclose the identities of the 10 drone manufacturers, the types of components it produces for them, or the specific systems in which the 12,000 parts are used.

The investor call, however, indicated that the activity extends well beyond spare parts. The company described demand from defense-tech companies for “drone manufacturing, munitions manufacturing and manufacturing applications on next-generation platforms.” This activity helped Stratasys Direct increase quarterly revenue by 12.1% year over year.

According to Zeif, the division’s backlog is now at an all-time high, and the company plans to invest in additional production capacity.

“We are manufacturing more than 12,000 parts for aerospace and defense, mostly for drones. We work with 10 of the leading drone companies,” Zeif said. “We are going to invest in this capacity and further penetrate aerospace and defense.”

Defense Becomes a Major Growth Engine

The drone business is part of a much broader trend at Stratasys. According to Zeif, aerospace and defense is now “by far our largest vertical,” with revenue from the sector growing 17% year over year in the second quarter. Stratasys also believes it is currently the leading player in polymer additive manufacturing for the aerospace and defense market.

One prominent customer is the U.S. Air Force, which is expanding its use of Stratasys F900 printers for maintenance and spare-parts production. According to the company, the Air Force has already made repeat investments in multiple systems, with orders expanding to additional maintenance locations as part of long-term programs.

The trend recently received another boost when Stratasys was awarded a $7.8 million program by America Makes and the U.S. Department of Defense. Under the project, the company will develop real-time monitoring and quality-control capabilities for the F900 and a future solution based on the F3300 platform. The goal is to expand the use of additive manufacturing for critical parts across the U.S. defense sector.

The shift toward production-scale applications is also evident outside defense. Stratasys’ consumables revenue reached a record $66.3 million during the quarter, partly driven by growing consumption of high-performance materials. For the company, this is an indication that installed printers are increasingly being used for actual production rather than primarily for prototyping.

Markforged Adds Continuous Carbon Fiber

The trend also ties into Stratasys’ recently announced acquisition of Markforged. Stratasys will pay $42.5 million in cash for the company, which generated approximately $70 million in revenue in 2025. Stratasys expects the transaction to be accretive to EBITDA within the first year following completion.

One of the key technologies Stratasys is acquiring is Markforged’s continuous carbon fiber capability, which enables the production of strong, lightweight parts that can, in certain applications, replace metal components. According to Zeif, the advantage extends beyond weight savings to lower costs and significantly reduced post-processing requirements.

“Continuous carbon fiber can replace metal. It is lighter, cheaper and requires significantly less post-processing,” Zeif said.

He added that since the acquisition was announced, four major companies have already approached him about working with Stratasys to adopt the technology and develop standards for the use of continuous carbon fiber.

Markforged will also broaden Stratasys’ customer base. While Stratasys has a strong presence among large enterprises and in advanced industrial applications, Markforged brings a customer base that includes machine shops and midsize companies, along with a distribution network and software technology for simulation and distributed manufacturing management.

For Stratasys, combining the two companies is primarily intended to expand the range of applications it can offer defense and other industrial customers. As Zeif put it, the acquisition will allow the company to “say yes to more new business, faster — especially in aerospace, defense and automotive.”

IAI Drone to Detect Wildfires Within Minutes

Posted on by Techtime

[Above: IAI’s APUS 25 TactiQuad drone during wildfire detection system testing. Photo: Israel Aerospace Industries]

Israel Aerospace Industries (IAI) has completed the first flight test of a new national drone-based wildfire detection system and is now launching a broader series of operational trials designed to evaluate the platform under different field conditions and emergency scenarios.

The project is a joint initiative involving Israel’s Ministry of National Security, the Israel Innovation Authority, IAI, the Technion, and Israel Fire and Rescue Services. It combines a dedicated sensor suite with IAI’s APUS 25 TactiQuad tactical drone, a vertical takeoff and landing (VTOL) platform designed for high-altitude, long-endurance missions.

The system integrates day- and night-vision cameras, an infrared sensor capable of penetrating dense smoke, spectral image-analysis algorithms, and artificial intelligence that analyzes incoming data in real time to detect anomalies, identify fire outbreaks, and pinpoint their location with meter-level accuracy, even from high altitude. Alerts are transmitted immediately to the Fire and Rescue command center.

According to IAI, a single drone can monitor several hundred square kilometers and detect a fire within minutes of ignition, before it develops into a large-scale wildfire.

An Unconventional Propulsion System

The wildfire detection payload is mounted on the APUS 25 TactiQuad, a tactical UAV developed for long-endurance missions in harsh weather and demanding terrain.

Unlike conventional electric quadcopters powered by four independent electric motors, the APUS 25 uses a single liquid-cooled internal combustion engine mounted at the center of the aircraft. The engine operates at a constant RPM and drives all four rotors through IAI’s patented variable-pitch mechanism.

This architecture dramatically extends endurance, allowing flights of up to eight hours, compared with the tens of minutes typically achieved by battery-powered multirotor drones.

The aircraft is also designed to operate on heavy fuels commonly used in military logistics, including Jet-A1, JP5, JP8, as well as gasoline. This eliminates many of the logistical challenges associated with battery-powered drones, such as lengthy charging times and dedicated battery storage infrastructure.

Despite a maximum takeoff weight of just under 25 kilograms, the APUS 25 can carry payloads of up to 10 kilograms during extended missions. It is capable of operating in winds of up to 23 knots, reaching altitudes of 11,000 feet, and supplying 300 watts of onboard power to mission payloads.

Autonomous Wildfire Surveillance

The operational concept is designed to minimize operator workload. A firefighter simply defines the search area by drawing a polygon on a digital map. After takeoff, the drone autonomously navigates to the optimal observation point and continuously scans the designated area for four to five hours, providing emergency responders with a real-time operational picture.

Prof. Assaf Schuster of the Technion’s Faculty of Computer Science said the Fire Detection system demonstrates how advanced research can be translated into operational technology that saves lives and reduces damage. “The combination of the drone platform and its hardware with our algorithmic and AI capabilities will enable fire and rescue authorities to make decisions based on real-time information from the drone and respond to fires quickly and efficiently—at a very early stage.”

Former Mossad Chief David Barnea Joins Ondas

Posted on by Techtime

U.S.-listed Ondas Holdings has appointed former Mossad Director David Barnea as Global President and Chairman of Ondas Defense, strengthening the company’s ambition to become a fully integrated global defense technology provider.

Barnea joins Ondas’ senior leadership team and will support the company’s international expansion, strategic technology development, and relationships with governments and defense organizations. He will work closely with Ondas Chairman and CEO Eric Brock and Oshri Lugasi, Co-CEO of Ondas Autonomous Systems.

The appointment is part of Ondas’ strategy to build an integrated defense technology company centered on four complementary operational domains: Aerial Security, ISR & Persistent Intelligence, Precision Strike, and Autonomous Ground Systems. These capabilities are designed to operate through a unified software and command layer that integrates intelligence, communications, mission planning, decision-making, and coordinated operations.

Building an Integrated Defense Platform

As President of Ondas Defense, Barnea will help drive the company’s international strategy, initially focusing on the Middle East, Europe, and Asia—regions experiencing rapidly growing investment in autonomous defense systems, counter-UAS technologies, ISR platforms, precision strike capabilities, and autonomous ground systems.

He will also play a central role in Ondas’ acquisition strategy, helping identify operational capability gaps, evaluate technology investments and acquisitions, and prioritize the integration of new technologies into the company’s unified defense platform.

Barnea brings three decades of national security experience. He served as Director of the Mossad from 2021 to 2026, leading Israel’s intelligence agency through one of the most operationally intensive periods in its history, including conflicts involving Hamas, Hezbollah and Iran, complex hostage negotiations, and sensitive intelligence operations across multiple theaters. Before joining the Mossad, he served in the Israel Defense Forces’ elite Sayeret Matkal unit, later earned bachelor’s and master’s degrees in business administration, began his private-sector career in mergers and acquisitions at a leading Israeli investment bank, and joined Israel’s intelligence community in 1995.

Eric Brock said Barnea “led one of the world’s most sophisticated intelligence organizations during a period of active conflict, rapidly evolving threats, and highly complex operations in which artificial intelligence, advanced communications, cyber capabilities, and specialized technologies played an increasingly central role.”

Barnea added: “The nature of warfare is changing rapidly. Operational advantage now depends on the ability to integrate intelligence, artificial intelligence, resilient communications, autonomous platforms, and precision strike capabilities into a single coordinated environment.”

He continued: “Ondas has assembled a unique portfolio of technologies, companies, and experienced teams across some of the most critical areas of modern defense. The next step is to unify these capabilities, develop a focused operational and technology strategy, and build an integrated platform that can respond rapidly to the evolving requirements of governments and national security organizations.”

Growth Through Acquisitions

Ondas provides autonomous systems, robotics, and mission-critical technologies for defense, homeland security, public safety, critical infrastructure, and industrial markets.

Over the past three years, the company has built much of its defense business through acquisitions of Israeli companies, including Iron Drone, Airobotics, Sentrycs, Apeiro Motion, Roboteam, and 4M Defense. It also acquired the Israeli non-defense companies INDO Earth Moving and Omnisys.

This acquisition strategy has fueled rapid growth. In the first quarter of 2026, Ondas reported revenue of approximately $50 million, ten times higher than a year earlier. The company expects full-year 2026 revenue of approximately $390 million.

In January 2026, Ondas raised $1 billion in a public offering to finance its expansion. By the end of the first quarter, the company held approximately $1.48 billion in cash and cash equivalents.

Following the announcement of Barnea’s appointment, Ondas shares jumped about 9.7% on Nasdaq, giving the company a market capitalization of approximately $4.69 billion.

Intel Lays the Groundwork for a Dedicated Robotics Business

Posted on by Techtime

[Above: A humanoid robot powered by Intel Core Ultra Series 3 processors. Photo: Intel]

Intel is quietly building a comprehensive portfolio of robotics technologies that could eventually evolve into a dedicated robotics division or business unit. The scope of the initiative first became apparent at Computex 2026 in Taiwan, where Intel showcased a range of partners developing industrial, humanoid, and service robots based on the company’s hardware and software platforms.

Although the effort currently operates within Intel’s Client Computing and Physical AI group, the company took a more significant step about six weeks after Computex by formally introducing a new umbrella brand: Intel Robotics.

The new brand brings together several hardware and software initiatives, with software emerging as the centerpiece of Intel’s robotics strategy.

Today, the Intel Robotics portfolio includes the Core Ultra Series 3 processor family, including robotics-optimized versions designed for robotics and edge AI applications with integrated GPU and NPU accelerators, robotics development kits that enable customers to build new robotic platforms, and a unified software stack marketed as the Intel Robotics AI Suite.

Every Robot Is a Software Ecosystem

Intel also introduced several new software components designed to simplify robotics development and deployment.

One of the most significant additions is OpenVINO Physical AI, described by the company as the industry’s first open-source robotics library featuring a silicon-optimized inference runtime. The software provides developers with a standardized framework for transferring robot policies and multimodal AI models from the development environment to physical robots while maximizing inference performance.

Intel also announced Physical AI Studio, a software suite that supports data collection, model fine-tuning, optimization, and deployment of pre-validated Vision-Language-Action (VLA) models.

VLA models have become one of the fundamental building blocks of modern robotics because they enable robots to perceive their surroundings, understand context, and execute actions accordingly—a perception-action cycle that can occur dozens of times per second in real-world environments.

According to Dan Rodriguez, Corporate Vice President of Intel’s Edge Computing Group: “Deployment of Physical AI models has been slowed by fragmented software stacks and custom integrations required for every robot. With Core Ultra Series 3 and OpenVINO, we’re providing a unified path that accelerates the transition from AI experimentation to production-ready robots with accelerated inference performance.”

Reducing the Cost of Robotics

Physical AI Studio is already available to developers, while OpenVINO Physical AI is currently offered as a preview release on GitHub. General Availability is scheduled for the second half of 2026.

Intel argues that today’s robotics software landscape forces customers to rely on multiple overlapping software environments and excessive compute resources, increasing both deployment and maintenance costs.

According to the company, combining Core Ultra Series 3 processors with OpenVINO Physical AI enables customers to lower total cost of ownership (TCO), reuse software across different robot types, and streamline deployment in factories, warehouses, and retail environments.

Taken together, these announcements suggest Intel is moving beyond simply supplying processors for robots. Instead, the company is assembling an integrated robotics platform that combines silicon, development hardware, AI software, and deployment tools—potentially laying the foundation for a much broader push into the rapidly growing robotics market.

Is SolarEdge Headed for Another Inventory Crisis?

Posted on by Techtime

photo above: SolarEdge CEO, Shuki Nir

SolarEdge delivered one of its strongest quarterly performances in years in the second quarter of 2026. Revenue rose 19.6% year over year to $346.2 million, non-GAAP gross margin improved to 28.6%, the company reported its first non-GAAP operating profit since 2023, continued to reduce operating expenses, and generated both positive non-GAAP net income and positive free cash flow.

Investors, however, focused on the outlook rather than the quarter itself. SolarEdge guided for third-quarter revenue of $310-$340 million, along with a lower gross margin. The stock responded by falling more than 20%. The central question hanging over the earnings call was whether the company is entering the early stages of another inventory correction—the same process that triggered its dramatic downturn nearly three years ago.

The concern is not without basis. Several familiar warning signs have reappeared: the U.S. residential solar market remains weak, distributors are purchasing less equipment, installers are struggling to secure financing for new projects, and regulatory uncertainty is making the entire value chain more cautious. It is a combination that inevitably recalls the inventory crisis of 2023.

Not Excess Inventory, but Cautious Ordering

One of the main themes of the Q&A session was whether distributors are once again carrying excessive inventory and whether another wave of destocking is underway. New CFO Maoz Sigron addressed the issue directly: “Our channel inventory, to the best of our knowledge, is normalized.”

He added that inventory continues to move between distributors and product families, but the company does not currently see abnormal inventory accumulation. At the same time, he acknowledged that distributors have become significantly more cautious.

CEO Shuki Nir explained that the U.S. residential market continues to face a slower tax equity funding environment and ongoing uncertainty surrounding FEOC regulations. As a result, installers are finding it harder to launch new projects, their cash flows are under pressure, and distributors have “reduced the amount of inventory they carry.”

This may have been the most important takeaway from the call. In 2023, the problem was warehouses that were too full. Today, the concern is almost the opposite: distributors are reluctant to rebuild inventory. The key question is whether this caution will eventually turn into another prolonged order slowdown—or whether it simply reflects a temporary wait for greater regulatory clarity.

Europe Is Carrying the Business

The geographic breakdown highlights how much SolarEdge has changed since the previous downturn. Revenue from the United States totaled $154.9 million, down 2% sequentially and representing 44.7% of total revenue. Europe, meanwhile, grew 36% quarter over quarter to $154.4 million—almost matching the U.S. business.

According to Nir, European demand is benefiting from rising electricity prices and the gradual phase-out of net metering programs across several countries, both of which are accelerating demand for battery storage systems. Still, management is cautious about Europe. The company expects approximately $15 million of sequential revenue decline in Europe during the third quarter, primarily due to seasonality.

The picture in the United States is almost the reverse. Residential demand remains weak, but the commercial and industrial (C&I) business continues to gain momentum. SolarEdge said it now holds more than 50% of the U.S. commercial rooftop inverter market and that its systems are installed on the rooftops of more than 60% of Fortune 100 companies.

Nexis and Storage Are Reshaping the Business

Alongside its operational turnaround, SolarEdge continues to position the Nexis platform as the next growth engine for its core business. Nir said that shipments of the three-phase Nexis platform in Europe exceeded $60 million during the quarter, while the U.S. rollout is now beginning. At the same time, the company is seeing steady growth in battery storage demand. “We are expecting storage to become a bigger piece of our business.”

The same trends driving European solar adoption are also making battery retrofits increasingly attractive. SolarEdge disclosed that its retrofit business generated more than $20 million in revenue during the second quarter.

The Next Growth Bet Lies in AI Data Centers

Beyond solar, SolarEdge is attempting to establish an entirely new business in power infrastructure for AI data centers. The company is developing a Solid-State Transformer (SST) designed to convert medium-voltage AC directly into 800V DC power for AI servers.

As Nir explained: “Efficiency directly translates into greater compute capacity within a fixed power envelope… this additional compute leads to higher revenue, lower cost per token, and improved return on investment for the data center.”

The company has already demonstrated a working prototype to prospective customers, but management acknowledges that this is a long-term opportunity. Pilot deployments are planned for 2027, while meaningful revenue is not expected before 2028.

Valens Appointed Dean Martin to Lead the Automotive Business Unit

Posted on by Techtime

The provider of high-performance connectivity chipsets, Valens Semiconductor, announced the appointment of Dean Martin as SVP, Head of the Automotive Business Unit, effective September 1, 2026. Dean will succeed Adar Segal, who is stepping down from the role. Dean joins at a time of growing automotive momentum and will help drive commercialization and customer expansion.

“Dean is a veteran in the Automotive industry with a track record of turning innovative technologies into significant commercial successes,” said Yoram Salinger, Chief Executive Officer of Valens Semiconductor. “He has extensive experience in securing major design wins with leading global automakers and building the customer relationships needed to support long-term growth. I’m confident that he is the right person to execute on our existing design wins and further promote Valens chipsets across the global automotive market.”

Martin is a senior global sales leader with more than 25 years of experience scaling technology companies. He has held senior commercial roles at Harman Automotive, Red Bend Software, Agilent Technologies and Hewlett-Packard. He has secured breakthrough awards with global OEMs and built long-standing relationships with leading customers and partners across the industry.

Automotive is a major market for Valens

Based in Hod-hasharon, Israel, Valens provides high-performance connectivity chipsets for audio-video installations, videoconferencing and automotive. Its technology forms the basis for HDBaseT and MIPI A-PHY standards. 2025 revenues reached $70.6 million, compares to full year revenues of $57.9 million in 2024. Valens expect full year 2026 revenues of $75-$77 million, an increase of approximately 8% (midpoint) compared to the annual revenue of 2025. The company is listed on the NYSE with a market capitalization of $170 million.

Automotive revenues in 2025 accounted for 26.9% (equivalent to $19.0 million), compared to 37.3% in 2024. The decrease was due to gradual price erosion and a reduction in the number of units sold to Mercedes Benz. During Q1 2026, Automotive revenues accounted for approximately 35% of total revenues at $5.9 million, compared to $5.1 million in Q1 2025.