Connect with us

AI

Ceva CEO Amir Panush Named 2026 AI Company CEO of the Year

Ceva CEO Amir Panush took AI Company CEO of the Year at the 2026 AI Breakthrough Awards on June 25, backed by a dozen-plus NeuPro NPU design wins.

Published

on

The 2026 AI Breakthrough Awards named Amir Panush, the chief executive of Ceva, Inc. (NASDAQ: CEVA), as “Artificial Intelligence Company CEO of the Year,” the IP licensing company announced on June 25, 2026. The 9th annual program, run by AI Breakthrough, recognized Panush for reading a structural shift in how AI workloads actually run, away from cloud-only stacks and toward hybrid systems that keep inference local where power, latency, cost and privacy demand it. Ceva built its silicon and software portfolio around that split, and the award formalizes a CEO honor into a record of a calculated architectural bet.

Panush framed the recognition as a team honor and positioned AI inference as a distributed challenge rather than a cloud-only workload. He said the cloud would always play a role, but argued that billions of connected devices need to sense, reason and act locally, with Ceva’s portfolio built for exactly that need. The framing matters because the wider AI hardware conversation still defaults to data center accelerators, a category where Ceva does not compete. Its edge-first stance is a deliberate alternative, not a fallback.

A CEO Award That Points to a Bigger Architecture Bet

AI Breakthrough framed the recognition around a strategic call, not a product cycle. The program’s citation points to a move Ceva made earlier than most of the IP licensing market: aligning connectivity, sensing and inference technologies into one portfolio before customers had asked for them in one bundle.

That bundle now goes to market under the name AI Fabric, and it lets Ceva sell a full stack to chip designers rather than a single radio block or a single NPU. The release frames the bet as one about where inference actually has to happen. The press release uses the wording “hybrid model in which AI processing is distributed across cloud and local devices, with the right model running in the right place at the right time,” and that language shows up in both the program’s framing and Panush’s own remarks. Edge inference is the workload Ceva’s IP is built to run, and the AI Breakthrough judges called that out by name.

The framing matters because the wider AI hardware conversation still defaults to data center accelerators, a category where Ceva does not compete. Its edge-first stance is a deliberate alternative, not a fallback. Ceva’s pitch as a silicon and software IP licensor lets the company argue for distributed inference without having to ship accelerators of its own.

The awards cycle was the right vehicle for that story. AI Breakthrough organized its 2026 winners into an Industry Leadership category that places the CEO award next to vertical AI solutions and platform-level recognition, and Panush’s name appears there in the official AI Breakthrough 2026 winners announcement.

How Ceva Built Its Connect, Sense and Infer Stack

Ceva’s strategic response to the hybrid shift is a portfolio the company calls the AI Fabric, an integrated set of silicon and software IP spanning connectivity, sensing and inference. The portfolio is organized around three pillars, Connect, Sense and Infer, that Ceva describes as covering the essential capabilities for devices at the intelligent edge. Ceva is selling this as a stack, not as a parts catalog, which is the architectural claim the AI Breakthrough award effectively endorsed.

On the Connect side, Ceva licenses IP for 5G, cellular IoT, Bluetooth, Wi-Fi and UWB, the radios that move data on and off a device. Sense adds sensor fusion processors and AI DSPs that turn raw signals into something a model can act on.

Infer, the most-watched pillar in this announcement, delivers scalable edge AI NPUs that include the NeuPro family. Together, those three layers turn a chip designer’s building blocks into a blueprint for what Ceva calls Physical AI, systems that are always connected, contextually aware and capable of real-time decision-making. The press release describes the bundle as providing “the essential building blocks for accelerating the development of Physical AI systems,” a phrase worth quoting because it sets up the licensing numbers that follow.

The NeuPro Wins That Made the Story Concrete

The headline number behind the award is licensing traction. Ceva has secured more than a dozen NeuPro NPU IP licensing wins, with customers actively designing Ceva’s edge AI processors into the next generation of AI-enabled products. Those wins span consumer IoT, industrial, automotive, infrastructure and PC applications, five end markets the company identifies by name in its announcement.

The application list matters because it shows the strategy landing outside the IoT niches where edge AI is often assumed to live. Automotive and PC, in particular, carry volume and certification requirements that gatekeep smaller IP vendors.

  • 9th annual AI Breakthrough Awards program
  • More than a dozen NeuPro NPU IP licensing wins
  • Five end markets: consumer IoT, industrial, automotive, infrastructure, PC
  • More than 2 billion devices incorporating Ceva technologies ship annually
  • More than 21 billion devices shipped cumulatively, trusted by 400+ customers worldwide

Two Billion Devices a Year, and the Hybrid Push

Scale is the second pillar of the announcement. Ceva’s press release states that more than 2 billion devices incorporating Ceva technologies ship annually across consumer electronics, automotive, industrial IoT and mobile markets. The “About Ceva” section puts the cumulative count even higher: more than 21 billion devices shipped, trusted by 400+ customers worldwide. Those numbers describe footprint, not edge AI NPUs specifically, and Ceva was careful in the release to keep the licensing wins and the device totals in separate sentences. The company is positioning NeuPro traction on top of an installed base, not in place of one.

The breadth of Ceva’s portfolio is what makes that installed base usable for the hybrid AI thesis. Connectivity IP gives a device a path to the cloud when it needs one. Sensing IP gives it context, and inference IP gives it the ability to act locally. The table below sketches the three pillars against the workloads they target.

Ceva Pillar IP Categories Edge Workload
Connect 5G, cellular IoT, Bluetooth, Wi-Fi, UWB Wireless link to cloud and peer devices
Sense Sensor fusion processors, AI DSPs Turn raw signals into model-ready context
Infer Scalable edge AI NPUs (NeuPro family) On-device model execution

The technology mix is also why Ceva was comfortable pushing back against the cloud-only default. Ceva’s framing of distributed inference across cloud and local devices is consistent with the edge device end markets the company names in the release, including automotive and PC.

Ceva describes its IP as providing “ultra-low-power performance in minimal silicon footprint, helping customers accelerate development, reduce risk, and bring innovative products to market faster.” That pitch is closer to an EDA vendor’s framing than a chipmaker’s, since Ceva is not in the chip business, it is in the building block licensing business. The 2-billion-device installed base is the customer list the NeuPro pitch rides on, and the architecture argument lines up with Ceva’s separate NeuPro-Nano NPU Embedded Award at Embedded World 2026.

Why an Awards Program Spent Nine Years Tracking Edge AI

AI Breakthrough has been running its annual awards program for nine years, and the 2026 cycle attracted thousands of nominations from more than 20 countries. The judges evaluate entries across AI hardware, agentic AI, generative AI, computer vision, MLOps, large language models and intelligent document processing, the same fields Ceva’s edge stack now touches from a different angle. A nine-year run is unusual in an awards market that tends to cycle programs with the news cycle, which is part of why the Ceva citation leans on architectural framing rather than single-product novelty.

Steve Johansson, Managing Director of AI Breakthrough, credited Panush in the same release with positioning Ceva at the forefront of the edge AI revolution and with anticipating where AI inference was heading. Johansson named the strategic call directly, a sharper signal than a generic innovation citation would be.

The 2026 program is the first one to organize winners under an Industry Leadership category that lists “Company CEO of the Year” alongside vertical AI solutions, and Panush’s name is the one printed there in the official release. The parent structure helps explain why the program can afford nine-year continuity, since Tech Breakthrough runs the recognition programs as part of a broader intelligence platform. For Ceva, the citation lands as a third-party read on a portfolio strategy that has taken years to assemble and is now visible in design wins.

This award reflects the dedication of the entire Ceva team and the trust our customers place in us as their IP partner. AI inference is increasingly a distributed challenge. The cloud will always play a role, but billions of connected devices need to sense, reason and act locally.

The quotation above is from Amir Panush, Chief Executive Officer of Ceva, in the company’s June 25, 2026 press release on the AI Breakthrough CEO of the Year award.

Frequently Asked Questions

What award did Amir Panush win?

Panush was named “Artificial Intelligence Company CEO of the Year” in the 2026 AI Breakthrough Awards program, the 9th annual cycle run by AI Breakthrough. Ceva announced the recognition on June 25, 2026, via PR Newswire.

Why did Ceva’s CEO get the recognition?

AI Breakthrough cited his leadership in identifying the shift to hybrid AI inference and aligning Ceva’s connectivity, sensing and inference IP for edge devices. The strategic call positions Ceva’s portfolio for the local half of a distributed inference architecture, with cloud kept in the loop where it makes sense.

What is the NeuPro NPU?

NeuPro is Ceva’s family of scalable edge AI NPU IP, designed to run AI inference on-device rather than in the cloud. Ceva has secured more than a dozen NeuPro NPU IP licensing wins across consumer IoT, industrial, automotive, infrastructure and PC applications, the company said in its June 25, 2026 release.

How many devices use Ceva IP?

More than 2 billion devices incorporating Ceva technologies ship annually across consumer electronics, automotive, industrial IoT and mobile markets, the company says. The same release puts the cumulative count at more than 21 billion devices shipped and 400+ customers worldwide.

When did Ceva announce the award?

Ceva announced the recognition on June 25, 2026, in a press release distributed via PR Newswire. The full AI Breakthrough 2026 winners list was published the same day via GlobeNewswire.

Logan Pierce is a writer and web publisher with over seven years of experience covering consumer technology. He has published work on independent tech blogs and freelance bylines covering Android devices, privacy focused software, and budget gadgets. Logan founded Oton Technology to publish clear, no nonsense tech news and reviews based on real hands on testing. He has personally tested and reviewed dozens of mid range and budget Android phones, written extensively about app privacy, and built and managed multiple WordPress publications over the past decade. Logan holds a bachelor's degree in English and studied digital marketing at a certificate level.

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending