The Cadence Conundrum: Why Blockchain's 'Picks and Shovels' Are Also Undervalued

Solitude is the only auditor that never sleeps. Over the past week, I’ve been digging into the valuation debate around Cadence Design Systems, a company the market treats as a semiconductor software vendor but whose CEO argues is deeply undervalued amid the AI boom. The parallels to blockchain infrastructure are impossible to ignore. Here is the context. Cadence is one of two dominant players in Electronic Design Automation (EDA), the software required to design every modern chip. Without EDA, no AI GPU, no custom ASIC, no quantum processor. The industry is a duopoly with Synopsys, and its tools are essential for any chip design—from 3nm nodes to 2nm GAA. The CEO’s recent comments in Crypto Briefing (a crypto-focused outlet, interestingly) claimed the company is being priced as a traditional software firm while its value is actually that of a critical infrastructure provider for AI. This is exactly the same narrative we see in blockchain: the “picks and shovels” protocols that power the entire ecosystem—oracles, data indexing, layer-0s—are often valued by the market as simple utility tokens, while their true economic leverage is far greater. Now, the core insight. Based on my audit experience in 2017, I learned that the most undervalued assets are those whose revenue model is invisible to the standard valuation lens. Cadence’s core moat is threefold: decades of design IP, deep integration with foundries (TSMC, Samsung, Intel), and a full RTL-to-GDSII flow that makes switching costs prohibitive. The market sees this as a software license business. But the hidden layer is that every AI chip design cycle—from NVIDIA’s H100 to a custom ASIC for a hyperscaler—pays a “tax” to Cadence that scales with chip complexity. A single 2nm chip design now costs $5-7 billion, and roughly 25-30% goes to EDA and IP. That means each new AI chip generation increases Cadence’s revenue per chip, not just the number of chips. This is a volume + value expansion, not just volume. In blockchain, the same logic applies to middleware like Chainlink or The Graph: every new dApp, every cross-chain bridge, every oracle call increases the “tax” they collect, but the market often prices them as if they are capped by a fixed number of transactions. However, the contrarian angle demands scrutiny. The loudest voice is rarely the most aligned. The CEO’s “undervalued” claim might be a self-serving narrative to justify a premium acquisition or to boost stock before a lockup expiry. The real test is whether the market is right to be skeptical. Cadence’s R&D intensity is ~30% of revenue, which is high for a software company. This spending is necessary to embed AI into its own tools (Cadence.AI platform) and to expand into system-level design (via acquisitions like Invecas and OpenFive). The market may be discounting this because the payoff is uncertain and long-cycle. Similarly, in blockchain, many infrastructure projects have high token emissions to fund development, but the market often treats that as dilution rather than investment. The contrarian truth is that the market is systematically undervaluing any asset whose revenue is tied to future productivity gains rather than current transaction volume. Let me ground this with a personal experience. In 2020, I founded The Silent Node, a community for women in Web3. I saw firsthand how the loudest projects—the ones with the most hype and the highest TVL—were often the most fragile. The quiet ones, the ones that focused on deep infrastructure, were the ones that survived the 2022 crash. Cadence is that quiet infrastructure. Its revenue is not dependent on which AI chip wins; it’s dependent on the fact that chips need to be designed. That is a structural, non-cyclical demand. The same applies to blockchain’s base layers: Ethereum, Solana, and Bitcoin. The market often trades them as speculative assets, but their real value is as settlement layers that collect fees from every transaction above them. The fee growth is a leveraged bet on ecosystem activity, and that leverage is often underappreciated. Code is law, but conscience is the interpreter. The takeaway here is not that Cadence is a buy or that every blockchain infrastructure token is undervalued. The takeaway is that the market consistently misprices levered infrastructure. In the AI boom, the EDA picks and shovels are undervalued because the market still sees them as software vendors. In the blockchain boom, the oracle, data, and settlement layers are undervalued because the market still sees them as token projects. The correct frame is to ask: what is the “tax” that this protocol collects, and how does that tax grow with the underlying industry? If the tax is a fixed percentage of a growing pie, the asset is a leveraged bet on that pie. The market often misses that leverage because it focuses on the numerator (token price) rather than the denominator (total addressable market). Solitude is the only auditor that never sleeps. The next time you hear a CEO claim their company is undervalued, ask yourself: is the market underestimating the leverage of the tax, or is the CEO underestimating the risk of commoditization? In Cadence’s case, the moat is real. In blockchain, the moats are often less visible but equally structural. The quiet conviction of infrastructure is what moves markets over the long term.