The semiconductor industry should expect a renewed focus on alternative computing paradigms and advanced materials science, driven by figures like Gelsinger. While traditional silicon scaling will continue to push boundaries through methods like 3D packaging, the search for entirely new physics to enable exponential growth is likely to intensify. This will manifest in increased research and development into areas like silicon photonics and new forms of optical computing, potentially leading to breakthroughs that could redefine chip architecture in the next decade.
Image: courtesy of Wired
Pat Gelsinger's Next Act: Can 'Light' Re-Ignite Moore's Law Beyond Intel?
Former Intel CEO Pat Gelsinger, a long-time champion of Moore's Law, is now looking to new technologies, specifically mentioning 'light,' to overcome the physical and economic hurdles that have slowed chip advancement. His new push comes as the industry grapples with the diminishing returns of traditional silicon scaling, raising questions about what this next generation of innovation might entail and its potential impact on the semiconductor landscape.
Outlook
Background
For decades, Moore's Law served as the guiding principle for the semiconductor industry: the number of transistors on an integrated circuit would double approximately every two years. This consistent exponential growth fueled the digital revolution, making computing power cheaper and more accessible with each generation.
However, the law's pace has visibly slowed. Brian Krzanich, another former Intel CEO, noted in 2015 that the industry's cadence had stretched to closer to two and a half years. By 2023, Pat Gelsinger himself, while still at the helm of Intel, acknowledged a further deceleration to a three-year cycle. This slowdown is not due to a lack of effort but rather fundamental physical and economic constraints. Shrinking transistors further requires increasingly complex and expensive manufacturing processes, pushing against the atomic limits of silicon.
Despite this reality, Gelsinger has consistently maintained an optimistic stance, declaring Moore's Law 'alive and well' since becoming CEO in 2021. He championed a strategy dubbed 'Super Moore's Law' or 'Moore's Law 2.0' at Intel, which focused on advanced packaging technologies like 2.5D and 3D stacking (such as Foveros) to boost transistor counts without relying solely on traditional two-dimensional scaling. This approach aimed to extend the law's relevance until at least 2031.
Yesterday, July 21, 2026, news emerged that Gelsinger is now advocating for new technologies, specifically involving 'light,' to jumpstart Moore's Law. This suggests a potential shift or expansion beyond the 3D packaging strategies he promoted at Intel, hinting at a more radical departure from conventional electron-based computing.
Precedents
The history of computing is marked by a continuous push against physical limits, followed by innovative breakthroughs that redefine those boundaries. Moore's Law itself was an observation, not a physical law, but it became a self-fulfilling prophecy, driving engineers to meet its pace. When silicon-based scaling began to face challenges in the past, the industry responded with architectural innovations, multi-core processors, and eventually, advanced packaging techniques like those Gelsinger championed at Intel.
This isn't the first time 'light' has been considered a potential successor to electrons for data transmission and processing. The concept of silicon photonics, which integrates optical components with electronic circuits on a single silicon chip, has been explored for decades. Its promise lies in faster data transfer, lower power consumption, and reduced heat generation, overcoming some of the core limitations of electrical interconnects. The shift from copper wires to fiber optics for long-distance communication is a macro-scale precedent for this kind of transition.
Historically, major technological shifts in computing have often come from outside the dominant paradigm. IBM and Intel, while massive innovators, have also seen challenges emerge from new architectures or material science. Gelsinger's current advocacy, post-Intel CEO, might allow him greater freedom to explore more disruptive, higher-risk avenues that a large, publicly traded company like Intel might find difficult to prioritize in the short term, given existing capital allocation and operational constraints.
The continued advancement of computing power, often measured by Moore's Law, is not merely an academic pursuit; it underpins nearly every aspect of modern technology and economic growth. From artificial intelligence and big data analytics to medical research and climate modeling, faster and more efficient processors are fundamental. If Moore's Law truly stagnates, the pace of innovation across these critical fields could slow dramatically.
Gelsinger's new focus on 'light' represents a high-stakes search for the next foundational technology. If successful, it could unlock entirely new capabilities, allowing for processors that are orders of magnitude faster and more energy-efficient than current silicon-based designs. This would have profound implications for data centers struggling with power consumption, for AI models demanding ever more computational muscle, and for the development of truly pervasive, intelligent systems.
For Intel, Gelsinger's former company, this development presents a complex scenario. While Intel continues to innovate with its 'Moore's Law 2.0' strategy, a breakthrough driven by Gelsinger in an adjacent or entirely new field could either validate its long-term vision or signal a missed opportunity if it fails to adapt. The broader semiconductor industry, including rivals like Nvidia, TSMC, and Samsung, will be closely watching for any signs of viability in optical computing or related 'light-based' technologies, as it could dictate the next major wave of investment and strategic direction.
Scenarios
Analysis1. Optical Computing Gains Traction: Gelsinger's advocacy, combined with ongoing research, could significantly accelerate investment and development in silicon photonics or pure optical computing. This could lead to practical, commercially viable products within the next 5-10 years, initially in specialized applications like high-performance computing or data center interconnects, before potentially expanding into mainstream processors. This outcome would mark a genuine shift in how computing is done, moving beyond the electron as the primary information carrier.
2. Incremental Advancements Continue: While 'light' technologies may see some progress, the primary gains in chip performance could continue to come from the existing trajectory: advanced packaging (2.5D/3D), new materials for transistors (e.g., Gate-All-Around FETs), and architectural innovations. In this scenario, 'light' might remain a niche solution for specific bottlenecks rather than a wholesale replacement for electronic processing, serving as a complementary technology rather than a revolutionary one.
3. Long-Term Research, Limited Immediate Impact: The challenges of integrating optical components with electronic ones, or building entirely optical processors, are immense. Gelsinger's push might inspire further academic and early-stage industrial research, but without a major scientific or engineering breakthrough, widespread commercial application could remain decades away. The practical hurdles, from manufacturing costs to thermal management of hybrid systems, could prove too complex for rapid deployment.
Timeline
2015
Moore's Law Slowdown Acknowledged
Brian Krzanich, then CEO of Intel, announced that the pace of semiconductor advancement had slowed to a cadence closer to two and a half years, rather than the traditional two years predicted by Moore's Law.
2021
Gelsinger Becomes Intel CEO, Reaffirms Moore's Law
Pat Gelsinger takes the position of CEO at Intel, emphatically stating that Moore's Law is 'alive and well' and committing Intel to its continued advancement.
2023
Moore's Law Cadence Shifts to Three Years
Pat Gelsinger confirmed that Moore's Law was slowing to a three-year cadence, acknowledging the increasing difficulty of transistor scaling while still at Intel.
June 29, 2026
Intel Foundry Leadership Appointment
Intel announced a leadership appointment at Intel Foundry, signaling continued efforts to accelerate development and manufacturing within its core chip production business.
June 30, 2026
Intel's U.S. Innovation Push
Intel released news highlighting its role in advancing U.S. innovation, AI, and manufacturing, aligning with national strategic goals for semiconductor independence.
July 21, 2026
Gelsinger's New Focus on 'Light'
Former Intel CEO Pat Gelsinger signaled his intention to jumpstart Moore's Law using new technologies, specifically mentioning 'light,' indicating a potential new direction for chip advancement.
Frequently Asked Questions
Moore's Law is an observation made by Intel co-founder Gordon Moore in 1965, stating that the number of transistors on a microchip roughly doubles every two years. This exponential growth in computing power has driven technological progress for decades.
Discussion
Be the first to share your thoughts.