As artificial intelligence continues to keep markets hot, industry leaders are increasingly convinced that semiconductor innovation now depends as much on collaboration as it does on transistor counts. The complexity of AI-era systems has outgrown the ability of any single company to deliver progress alone.

That shift was front and center at this year’s GSA Awards, an annual gathering honoring the companies shaping the chip industry’s next chapter. Enterprises now operate within a value chain so interconnected — spanning fabrication, test equipment, software and talent — that coordinated ecosystems, not standalone breakthroughs, are becoming the defining advantage, according to Charlie Kawwas (pictured), president of Broadcom Inc.

“My view is collaboration is at the heart of what we do in semiconductors,” Kawwas told theCUBE. “You can’t get everything from a single company. You need [fabrication plants], you need test equipment, you need chip manufacturers, you need software and you need people who pull all of this together.”

Kawwas and other industry leaders spoke with John Furrier following the GSA Awards Celebration, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. TheCUBE’s backstage program and related coverage served as a real-time scorecard for global semiconductor innovation, turning the industry’s biggest celebration into a lens on the next phase of AI and chip growth.

Here are 16 takeaways that show how theCUBE’s GSA Awards Celebration coverage captured innovation in motion, from clean-energy power to meet rising infrastructure demands, to advanced packaging and capital markets.

1. Semiconductor innovation demands open, scalable and power-efficient standards

For Broadcom in particular, OSP — “open, scalable, power-efficient” — has become a mantra for how its silicon and networking portfolio must be built for the AI era, Kawwas noted. The idea is simple: Lean on open standards, design for massive scale and performance per watt and rely on a tightly knit ecosystem with partners who can deliver on promises.

Don’t miss the exclusive insights on theCUBE.

2. Talent defines the next wave of semiconductor innovation

Advanced Micro Devices Inc. is channeling semiconductor innovation into the AI era by helping lead the Ultra Ethernet Consortium, which this year published its first specification for extending Ethernet to large-scale AI workloads, according to Rong Pan, corporate vice president at AMD. Innovations such as these, as well as Pan receiving the Women of Influence Award, highlight how AI’s promise is turning semiconductors into a magnet for engineering talent while keeping AMD at the center of the AI-scale systems conversation.

Catch the full story on theCUBE.

3. Hyperconverged edge is pushing cloud-scale expectations into every device

AI has driven explosive growth across cloud and semiconductors, pushing hyperscalers to rapidly scale system configurations that combine AI accelerators with general-purpose compute — even as they contend with real supply constraints, according to Danielle Shown, silicon cloud hardware and infrastructure at Microsoft Corp. Looking ahead, the next phase of semiconductor innovation is likely to center on a “hyperconverged edge,” where phones and small-footprint systems begin to operate like miniature AI factories, shifting the real challenge from building individual agents to orchestrating them effectively at scale.

Hear the full conversation with theCUBE.

4. Markets brace for an AI-fueled wave of new listings in semiconductors and beyond

Semiconductor innovation is back under the Wall Street spotlight, with AI-linked chip and infrastructure players helping to drive a surge of billion-dollar tech listings at the New York Stock Exchange this year, according to Bonnie Hyun, head of United States capital markets at the NYSE. The hype on semiconductors now stretches from graphics processing units to edge systems on a chip — such as Ambiq Micro Inc.’s healthcare devices — making this new hardware architecture a proving ground for fresh public companies, she added.

Watch the full sit-down with theCUBE.

5. Rewriting Moore’s Law economics with a new era of computing

Moore’s Law — the steady doubling of transistor counts and falling cost per transistor — has always depended on pushing lithography to print ever-smaller features at viable economics. xLight Inc.’s new light source is designed to do exactly that by driving shorter wavelengths and higher-power light to speed wafer output, a move Pat Gelsinger, chairman of the board of xLight Inc., framed as part of a broader reset in semiconductor innovation that leans on new materials, 3D packaging and a shift from copper to optical interconnects across the stack.

Hear the full story from theCUBE.

6. Engineering leadership is paramount for an ethical AI-powered future

Semiconductor innovation’s long arc, from million-fold transistor gains to multibillion-dollar fabs, has always been a collective achievement rather than the work of lone engineers, according to Tsu-Jae Liu, president of the National Academy of Engineering. Looking ahead to AI’s next wave, that same collaborative ecosystem is poised to train engineers who rigorously consider global impact, so the industry’s power is consistently leveraged for humanity’s long-term good.

Check out the full conversation with theCUBE.

7. AI infrastructure wave to reshape automotive, robotics and real-time edge compute

Tenstorrent Holdings Inc. is targeting its semiconductor innovation at real-time, safety-critical AI workloads by building central processing units optimized for edge and data center inference across robotics and autonomous-vehicle systems, said Erik Goodman, chief financial officer of Tenstorrent. This example of industry-spanning AI buildout constitutes a macro shift that will turn cars and edge devices into rolling data centers, with success hinging on breaking into entrenched automotive supply chains and delivering hardware in the right form factors.

Dig into the full conversation with theCUBE.

8. Storage steps into the spotlight as data pipelines strain legacy infrastructure

ScaleFlux Inc. is riding a breakout moment for AI storage as hyperscalers and enterprises finally recognize data pipelines as a third pillar of AI infrastructure alongside compute and interconnect, said Hao Zhong, co-founder and chief executive officer of ScaleFlux Inc. Years of underinvestment in solid state drives and data-path innovation have left a gap just as AI drives explosive data growth and context-heavy workloads, turning storage into one of the hottest levers for sustaining semiconductor innovation at scale.

Get the full story with theCUBE.

9. AI’s runaway power demands are turning energy use into the defining bottleneck

Auradine Inc. is addressing headline-making energy constraints with ultra-low-power AI extended processing unit chips, aiming to cut data-center consumption by an order of magnitude. That targeted mix of aggressive power reduction and novel cooling techniques is why the company was a finalist for Most Respected Private Semiconductor Company this year, Sanjay Gupta, chief strategy officer of Auradine, told theCUBE.

Watch the full interview with theCUBE.

10. Neural processing units move AI compute from the cloud to the edge

Kneron Inc. is betting big on NPUs succeeding CPUs and GPUs as the most efficient engines for on-device AI, shrinking energy use while pushing semiconductor innovation into everything from robots to everyday electronics, according to Albert Liu, chief executive officer of Kneron. That move treats NPUs as the “MP3” of compute, a compact accelerator that can match the work of many CPUs or GPUs as AI and wireless networking converge at the edge.

Watch the whole interview on theCUBE.

11. Security silicon remains in the AI infrastructure spotlight

Axiado Corp. is positioning its secure control modules as a new layer of AI infrastructure, embedding AI-driven security and management directly alongside traditional processing units, said Gopi Sirineni, founder, president and chief executive officer of Axiado. Fresh $100 million Series C+ funding will fuel a rapid ramp as data center builders race to harden systems at the silicon level and keep pace with a new boom in semiconductor innovation.

Hear the full conversation with theCUBE.

12. AI-era interconnects push silicon into a new Cambrian phase

As demand for AI-driven hardware grows, Marvell Technology Inc. is embracing what it calls a “Cambrian evolution of interconnect,” making high-speed connectivity central to next-generation data centers and AI factories. The industry is approaching a large-scale shift in computing, with quantum research and new hardware paradigms entering the mainstream and creating new opportunities for engineers, Manisha Gambhir, senior director of engineering at Marvell Technology, told theCUBE’s Furrier.

Catch the whole segment with theCUBE.

13. Low-power FPGAs quietly become the glue of data center and physical AI systems

AI factories and “physical AI” devices alike need frugal, flexible silicon to sit alongside big CPUs and GPUs and stitch together sensors, networking and storage. Lattice Semiconductor Corp. is positioning its small, low-power Field-Programmable Gate Arrays as that “companionship” layer for data centers and edge systems from image sensors and LiDAR to industrial equipment, according to Ford Tamer, chief executive officer of Lattice Semiconductor Corp. He pointed to robotaxis, new medical devices and future humanoid helpers as early examples of how this quiet layer of programmable logic will shape everyday life as today’s semiconductor innovation moves into tomorrow’s daily reality.

Hear the full conversation with theCUBE.

14. AI-era chip winners graduate from private darlings to public market leaders

On the topic of AI moving from novelty to everyday utility, the companies wiring up infrastructure are being pulled into the public market spotlight. Astera Labs Inc. is leaning into that moment after winning the GSA’s Most Respected Public Semiconductor Company award, doubling down on ever more complex AI-focused products while treating customer trust as its main currency, according to Jitendra Mohan, chief executive officer of Astera Labs.

Watch the full discussion with theCUBE’s John Furrier.

15. Private chip innovators find room to run in a land of giants

Opportunities in custom silicon are expanding as AI-driven demand creates space for agile specialists, even among industry giants. SiFive Inc. is capitalizing on this trend as a respected private semiconductor company with hundreds of major high-tech customers and a potential public listing on the horizon, in a market where today’s frantic pace will be the slowest it ever moves, Patrick Little, chief executive officer and chairman of SiFive, told theCUBE.

Catch theCUBE’s exclusive full segment.

16. Semiconductors move to the frontline of the clean-energy transition

Semiconductors now sit at the center of efforts to decarbonize the planet, turning energy efficiency into a critical climate imperative — and a major business opportunity. Infineon Technologies AG is leaning into that role with microcontrollers, sensors and power semiconductors that help customers use less electricity to get more output, a contribution Andreas Urschitz, chief marketing officer of Infineon Technologies AG, said is essential to making a digital, low-carbon future viable.

Watch the full segment with theCUBE.

Here’s the complete video playlist from SiliconANGLE’s and theCUBE’s GSA Awards Celebration coverage:

Photo: SiliconANGLE