Exploring What the SPIE Debate Means for the Next Decade with Erik Hosler

In February, leading voices from across the semiconductor ecosystem gathered at the SPIE Advanced Lithography symposium to explore one of the most consequential questions in modern technology: Is Moore’s Law nearing its end, or can it be saved? The panel brought together representatives from all seven SPIE conferences, offering a panoramic view of both the threats to continued scaling and the innovations that could extend it. Erik Hosler, a strategic consultant with a focus on bridging emerging technologies with user impact, captured the tension that permeated the conversation. Progress is possible, but it is no longer inevitable.

The discussion was far from conclusive. Some believed clever engineering workarounds had already rescued Moore’s Law. Others argued that new materials, optical methods, or 3D strategies could keep it going. Yet beneath the optimism was a common thread: moving from theory to scalable solutions will require unprecedented coordination across sectors, disciplines, and even companies. The next decade will not just be about new tools. It will be about translating ideas into infrastructure.

A Decade of Diverging Paths

For much of the last 50 years, Moore’s Law served as both a benchmark and a unifying force. It offered a shared roadmap that designers, manufacturers, and investors could follow. But as feature size reductions slow and the physics becomes harder to control, that roadmap has begun to fracture.

The SPIE panel highlighted this divergence. EUV experts spoke of higher numerical aperture systems. Materials scientists pointed to photoresist chemistry breakthroughs. Others emphasized advanced metrology, AI-driven process control, or design-side workarounds. None were wrong, but none alone were enough.

A single scaling breakthrough may not define the next ten years. Instead, they will be shaped by the ability to integrate multiple approaches across domains and to do so in a way that scales in cost, manufacturability, and time.

From Breakthrough to Deployment

One of the panel’s most striking themes was the gap between technological possibility and production reality. Many innovations already exist in lab environments or pilot lines. High-NA EUV, multicolor lithography, 3D stacking, and machine learning-based process control have all shown promise. But turning these into high-volume manufacturing tools is a different challenge. It requires new infrastructure, new partnerships, and new models for risk-sharing between suppliers and foundries.

Rich Wise of Lam Research bluntly stated during the panel that progress will depend on co-development between companies that once worked in silos. Lithography and etch suppliers, for instance, will need to co-optimize tools in tandem. That kind of collaboration requires more than a handshake. It demands shared goals, synchronized roadmaps, and trust.

Institutionalizing Innovation

In the past, major leaps often came from a single invention or tool. Think immersion lithography or chemical-mechanical planarization. Today, the gains are smaller, more distributed, and more dependent on systemic change.

That is why the SPIE panel spent so much time discussing process ecosystems rather than standalone technologies. As scaling gets harder, innovation must be institutionalized. That means tighter collaboration between industry and academia, stronger feedback loops between R&D and fab floor implementation, and more investment in talent that can navigate both physics and systems.

The panel made it clear that scaling is no longer a question of technical ingenuity alone. It is just as much a challenge of coordination across supply chains, institutions, and design ecosystems. Erik Hosler notes,
“But avoiding the death of Moore’s Law won’t be easy.” This observation captures the industry’s core tension. The latest ideas are not in short supply, but moving them from conference presentations to high-volume fabs requires structural change. That means integrating R&D pipelines, redefining success metrics, and investing in a new generation of cross-disciplinary talent.

Rethinking Metrics of Success

The panel also surfaced an important philosophical shift. In the past, progress was easy to measure. Every two years, chips got smaller, faster, and cheaper. But with shrinking feature sizes delivering diminishing returns, the metrics must develop.

The next decade will demand a more nuanced view of advancement. Metrics like performance-per-watt, integration density, and system-level efficiency may replace raw transistor counts. These are harder to explain, harder to track, and harder to standardize, but they are better aligned with how modern technology creates value.

Moore’s Law, in its original form, may fade. But the spirit behind it, the drive for consistent, measurable improvement, can live on if new metrics gain credibility and adoption.

New Frontiers Beyond Lithography

While lithography will remain a critical lever for scaling, the panel made clear that future gains may come from outside the cleanroom. System-level design, advanced packaging, co-optimization with software, and the use of AI in chip design and verification are all becoming equally important.

Photonics, MEMS, and even quantum components are entering the conversation. These are not replacements for CMOS scaling but extensions of it. The future will include hybrid systems where different computing modalities are integrated, not simply different feature sizes.

It means that engineering expertise must also be developed. The fab engineer of the 2030s may need to be fluent in software, optics, and materials science simultaneously. The companies that thrive will be those that can build teams capable of that kind of interdisciplinary thinking.

The Role of Strategic Partnerships

A decade ago, it was feasible for a leading-edge company to develop most of its innovation in-house. That is no longer the case. The capital intensity, complexity, and scale of modern semiconductor development now require deep, strategic partnerships.

The SPIE discussion pointed to an emerging trend: shared platforms between tool vendors, co-investment in process development, and longer-term joint ventures between foundries and suppliers. These are not mere vendor relationships. They are strategic alignments that allow for the pooling of expertise, risk, and reward. In this new world, competitive advantages will come not just from owning the best tools, but from building the best coalitions.

A Decade That Will Redefine the Industry

The panel did not answer whether Moore’s Law is dead or alive. But it made one thing clear: the 2020s will redefine what semiconductor progress looks like. It will be less about transistor dimensions and more about architectural creativity, process control, and collaboration.

The next generation of innovation will emerge not just from patents and prototypes but also from policies, partnerships, and people. From how research is funded to how fabs are designed to how talent is trained, the scope of what matters is expanding. It is no longer just an engineering challenge. It is a coordination challenge. How the industry meets that challenge will determine whether Moore’s Law becomes a legacy or a launching point.

Conclusion: Turning Insight into Action

The SPIE panel made one thing clear. The ideas to extend Moore’s Law are already here. What is missing is the system to turn them into scalable, manufacturable, economically viable products.

That system will require new thinking across the board. From redefined success metrics to deeper cross-sector partnerships, from wider skillsets to institutionalized feedback loops between innovation and implementation, the next decade will be shaped not by any single breakthrough, but by how well we integrate them. Avoiding the death of Moore’s Law is possible. But it will demand more than new tools. It will demand new habits.