Why Buying Formal Tools Is Not the Same as Adopting Formal

The AI era is pushing semiconductor verification into a new reality. Designs are growing more complex, quality expectations are rising, and the cost of late bugs has never been higher. Across the industry, first-silicon success has fallen sharply and more projects are requiring multiple spins. In response, many companies have invested heavily in formal verification tools, expecting to improve confidence, reduce risk, and catch critical problems earlier. But for many teams, the results still fall short of that promise. The tools are there, yet the old pattern remains: heavy reliance on simulation, extensive debug, and too many bugs discovered too late.

So why are so many organizations struggling to extract value from their formal investment? We identified five major reasons:

1. Skills Gap. Formal verification is not just simulation with different syntax. It requires engineers to think in terms of properties, abstractions, assumptions, proof convergence, and counterexample analysis. That does not mean formal is only for specialists, and modern tools are clearly becoming easier to use. But it does mean companies cannot assume they can build a strong formal practice simply by hiring a few junior engineers and handing them licenses. Industry commentary continues to cite workforce capability as a primary adoption barrier, and broader semiconductor hiring pressure makes experienced formal talent even harder to find.

2. Methodology. Many organizations approach formal mainly as a tool rather than as a verification discipline. They run a pilot on one block, prove that it can find bugs, and then stall because they do not have a repeatable way to choose targets, create and maintain properties, interpret results, and feed those results back into the broader verification plan. Real adoption requires workflow design, not just tool access. Often usage splits into disconnected extremes, where a few specialists apply formal deeply on targeted problems, while a broader set of engineers use only the automated push-button formal apps. What is often missing is the disciplined middle ground: a systematic way to apply formal to mainstream verification challenges, especially the corner cases that simulation struggles to cover. That is where many teams fail to realize the full value of their formal investment. They have the tools, but they lack the methodology, training, and support needed to make formal part of the everyday flow.

3. Misconceptions. Engineers are often confused about what formal actually is and where it is useful. Some still associate formal mainly with equivalence checking, while others assume it can only be used on small blocks or a narrow class of problems. It is often believed that formal can only be used for functional correctness, and requires special expertise. Those misconceptions have been around for years, and they continue to hold adoption back.

4. Convergence. This is one of the least glamorous barriers, but one of the most important. Engineers can believe in formal and still fail to get value if proofs do not converge, if the abstraction is weak, or if the team does not know how to restructure the problem. When that happens, runs get terminated, confidence falls, and future investment becomes harder to justify. This is one reason methodology and experience matter so much. The challenge is not just finding a bug. It is making formal practical enough, predictable enough, and scalable enough that teams trust it in the flow. Industry writing on formal fundamentals still identifies state-space growth and proof tractability as central challenges, even as tools improve.

5. Institutional Inertia. Many organizations have been able to scrape by with simulation-heavy methods for years, so excessive regressions and late bug discovery start to feel like a fact of life rather than a sign that the flow needs to change. Formal then gets treated as a specialist side practice instead of a core capability. But current industry practice suggests the pressure is building in the opposite direction. Formal is starting to be used across NoCs, RISC-V, arithmetic logic, low-power behavior, security scenarios, coverage-gap analysis, and area optimization. The most successful teams are integrating it where exhaustive reasoning changes outcomes the most.

Axiomise is a company focused on helping engineering teams overcome exactly these barriers to practical formal adoption. It does not replace the underlying formal engine. Instead, it helps teams translate formal capability into practical adoption through training, consulting, reusable property packages, and focused applications. Axiomise lowers the barriers to success and unlocks the value of formal tool investments.

For engineering leaders, that may be the most important shift to recognize. The bottleneck is no longer proving that formal has value. The bottleneck is turning formal into a reliable part of how the organization works. As design complexity rises, the gap between owning formal tools and adopting formal effectively is becoming more visible and more costly.

The teams that move ahead will be the ones that treat formal as a first-class capability that must be taught, integrated, and deliberately scaled. In the AI era, that is not just a methodology decision. It is a matter of market survival.