The Lifecycle
Rewritten
For fifty years, deep tech ran a predictable course: a long climb through scientific risk, a single valley of death, and a slow ascent to scale. That curve is being redrawn. The science phase is collapsing, the hardest gap has moved downstream, and the capital and demand around it are being rebuilt from scratch.
SilverX Fund Research
Investment Intelligence · July 2026
Two ways to build the same company.
The classic lifecycle (gold) spent its energy surviving the science phase. The reshaped lifecycle (blue) compresses discovery with AI, then faces a steeper, deeper chasm at first commercial deployment - where state equity and offtake contracts now bridge the canyon private markets cannot cross alone.
Vertical axis = capital-at-stake and risk carried; horizontal axis = time across the four lifecycle stages.
The geometry of deep tech has changed. Not just the speed - the shape.
Five forces are converging simultaneously: AI is collapsing the science phase, the hardest financing gap has migrated to first commercial scale, capital stacks are being rebuilt around blended sovereign and private capital, demand is arriving years earlier from defence and hyperscalers, and talent premiums have shifted from PhDs to industrialists. Taken together, these don't simply speed the old lifecycle up. They redraw its geometry - and with it, the entire playbook for founders and investors alike.
The Science Phase Is Collapsing
AI cuts materials discovery from 10–20 years to 1–2. Drug preclinical timelines fell from 3–4 years to 18 months. But AI compresses invention, not approval - the bottleneck has migrated downstream.
The FOAK Valley Is the New Chasm
The $50M–$500M gap between a working pilot and full commercial deployment now sits in a financing no-man's-land - too heavy for venture, too unproven for infrastructure lenders. 51% of investors named it the hardest stage.
The Capital Stack Has Been Rebuilt
Equity alone cannot fund a first-of-a-kind facility. A rough rule of thumb: 30% equity, 40% grants, 30% debt. Governments are no longer grant givers - they are strategic co-investors with sovereign time horizons.
India Is Running the New Playbook
₹1 lakh crore in patient capital, fabs live, space funding doubling. India is building the demand, the capital, and the manufacturing base the reshaped lifecycle requires - faster than almost anywhere.
The Baseline
The classic march
The textbook deep-tech company was defined less by what it built than by how long it took to build it. Where software startups lived and died on market and customer risk - is the market big enough, does the product fit? - deep tech carried a different burden: technical and financing risk. If you could actually build a commercial quantum computer or a fusion reactor, no one doubted the demand. The question was whether you could survive the years of science required to get there.
9.2 yrs
Average age of a leading deep-tech company before commercialisation
$801M
Average capital raised by leading deep-tech firms before meaningful revenue
TRL 7
Typical readiness level - a working prototype, not yet a business
Those years were long. In Bessemer's 2023 study of the world's leading deep-tech companies, the average firm sat at a technology readiness level of about 7 - a working prototype, not yet a business. Capital went in for the better part of a decade before revenue appeared. That shape held for a generation. It is now breaking in several places at once.
Where the Curve Breaks
Five forces redrawing the shape
Taken together, these don't simply speed the old lifecycle up. They change its geometry.
Force 01 - The front collapses
The longest, most expensive phase of the classic lifecycle - discovery - is the one shrinking fastest, and AI is the reason. In materials science, computational prediction and autonomous self-driving labs are compressing the road from concept to candidate material from a traditional 10–20 years to as little as 1–2. In drug discovery, AI now cuts early-stage timelines by an estimated 30–40%; in 2025, the first drug with both its target and its molecule designed entirely by AI completed Phase IIa trials, having reached preclinical candidates in 18 months rather than the usual three to four years.
AI compresses invention, not approval. Clinical trials, regulatory review, and manufacturing scale-up have not moved. The science phase is collapsing - which means the bottleneck is migrating to everything that comes after it.
AI Compresses the Front of the Curve
The longest phase of the classic lifecycle is collapsing fastest. Materials discovery and drug preclinical timelines have been cut by 80–90% by AI-driven computation and autonomous labs.
Source: Bessemer Venture Partners Deep Tech Report 2023; MIT Technology Review AI Drug Discovery 2025.
Force 02 - The valley moves downstream
For a physical world company, invention was never the hardest part of scaling - building the first commercial plant was. That gap now has a name: the first-of-a-kind (FOAK) valley of death, the stretch between a working pilot and full-scale deployment. It typically demands $50–$500 million - occasionally approaching $1 billion - and it falls in a financing no-man's-land: too capital-heavy for venture funds sized for software, too unproven for infrastructure lenders who want de-risked, bankable assets.
51%
of investors in a 2025 CTVC survey named first commercial scale facilities the single hardest stage to finance
69%
expected that capital to get scarcer, not more abundant in the years ahead
"The technology is the entry fee. The supply chain is the product."
The moat has moved with the risk: from the patent to the production line.
Force 03 - The capital stack is rebuilt
A company that lives in the world of atoms can no longer be financed the way a SaaS company is - equity, round after round. The emerging model is blended. Capital is also concentrating later: in quantum, Series B and beyond now accounts for about 63% of venture dollars, a decisive shift toward funding commercialisation over experimentation. The deeper change is who sits at the table: governments are moving from grant givers to strategic investors, taking equity, extending patient debt, and treating frontier companies as national infrastructure with time horizons no venture fund can match.
The Emerging Rule of Thumb for a First Commercial Facility
A physical world company cannot be financed like SaaS - equity round after equity round. A blended model is emerging: roughly 30% equity, 40% grants or subsidies, 30% debt or project finance.
Venture & strategic investors
Government, sovereign programmes
Infrastructure lenders, DFIs
An illustrative rule of thumb circulating among hard-tech investors. Source: CTVC Hard-Tech Capital Stack Survey 2025.
Quantum capital: 63% now flows to Series B and beyond
A decisive shift toward funding commercialisation over experimentation - the single clearest signal that the deep-tech investor base is repricing risk from the science phase to the deployment phase.
Force 04 - Demand arrives early
In the classic lifecycle, a deep-tech company spent a decade pre-revenue. Increasing demand now shows up years sooner, pulled forward by three buyers with deep pockets and strategic urgency: defence, sovereign programmes, and hyperscalers hungry for compute and energy. Defence tech is the clearest case. A buyer willing to sign an offtake or procurement contract converts a science-risk company into a contracted revenue company, far earlier in the curve. This is why deep tech's share of venture capital has roughly doubled over the decade.
$510M
Chaos Industries, late 2025 - explicitly to stand up manufacturing and deployment, not fund basic research
Raised at a $4.5B valuation
The era of raising nine-figure rounds to build more science has shifted. Chaos Industries raised $510M to build factories and deploy product into paying customers - the procurement contract as the proof-of-market, not a product demo.
Deep Tech's Share of Global Venture Capital (~10% to ~20%)
The category stopped being a patience trade and became a demand trade. Early revenue from defence, sovereign programmes, and hyperscalers pulled forward the demand signal.
Source: Deep Tech Alliance Global Report 2025; Dealroom Deep Tech Tracker 2025.
Force 05 - The talent equation flips and inflections arrive faster
The classic deep-tech founder was a PhD. That still matters, but the scarce skill has shifted with the bottleneck. As the hard part moved from invention to production, the premium moved with it - toward industrialists who know how to move atoms at scale, where a background in high-volume manufacturing can outweigh a third postdoc.
And because the front of the curve is compressed while demand arrives early, the old pattern - years under the radar, then suddenly a household name - is steepening. SpaceX and OpenAI were the archetypes; the distance between obscurity and ubiquity keeps getting shorter.
The New Shape, in One Line
Bessemer summarised the deep-tech challenge as technology, talent, and tons of cash. The reshaped lifecycle can be summarised just as tightly - as three things that now gate every hard-tech company.
The science is the entry fee.
AI has made invention faster and cheaper. A breakthrough alone no longer buys you a defensible company; it only qualifies you to compete.
The supply chain is the product.
The moat, the risk, and the valuation step change now live in the ability to manufacture and deploy at scale - not in the lab result.
The state is a co-founder.
Sovereign capital and strategic demand are no longer background funders. They underwrite the stage private markets cannot cross alone.
The Case in Point
India: the sovereign playbook
This is the environment India is engineering on purpose. Where the last generation of Indian deep-tech founders flew to Silicon Valley for patient capital, the state is now building that capital at home - and doing so precisely where the new lifecycle is hardest.
The centrepiece is the ₹1 lakh crore (~$12 billion) Research, Development and Innovation (RDI) Scheme, approved in July 2025 and launched that November, with ₹20,000 crore allocated in its first year. Its design reads like a direct answer to the FOAK problem: long tenure, low or zero interest loans and direct equity. The state is a strategic investor, not grant giver - funding up to half of project cost for ventures at TRL 4 and above, with the intellectual property required to stay in India.
₹1L Cr
~$12B
RDI Scheme Patient capital, approved July 2025
₹76k Cr
Semiconductor Mission
First commercial fabs now live for tape-outs
$276M
33 deals in 2025
Space funding - more than the prior two years combined
~6,000
60% outside metros
Deep-tech ventures across India
Around it sits a full stack of missions. The ₹76,000 crore Semiconductor Mission is shifting from planning to execution as the first commercial fabs near production and open their lines to Indian startups for tape-outs. Space drew $276 million across 33 deals in 2025 - more than the previous two years combined - backed by dedicated public funds. And defence indigenisation has tripled its targets.
But running the new playbook exposes the new frictions.
Missing middle at Series B+
Domestic capital still shies from five- to seven-year hardware cycles. The gap between seed and scale remains the single largest structural risk for India's deep-tech companies.
Grant liquidity trap
A December 2025 NASSCOM report flagged approved government funding failing to arrive when pre-revenue startups need working capital most.
Productisation deficit
Record patent filings turn into shelved research. India disposes of a patent application in about 58 months - against ~20 in China and ~23 in the US.
Sovereignty paradox
A hardware ecosystem that still imports an estimated 90–95% of its semiconductor and electronic components remains one trade restriction away from grounding its own drones.
The opportunity and the risk are the same fact. India is building the demand, the patient capital, and the manufacturing base the new lifecycle requires faster than almost anywhere. Whether that translates into companies that own components rather than assemble them is the question the next five years will answer.
Implications
What it means
For Founders
The strategic centre of gravity has moved. A decade ago the hard question was can we make the science work. Today, increasingly, it is can we make it at scale, at margin, with the right capital stack and the right anchor customer.
The pitch that wins in 2026 models the full journey from lab bench to first-of-a-kind facility - and names the grants, the offtakes, and the loan programmes that bridge the canyon.
For Investors
Underwriting deep tech now means underwriting a supply chain and a capital stack - not just a breakthrough - and increasingly, partnering with the state rather than working around it.
The returns are real: deep tech has delivered on par with top-tier ventures while taking a fifth of all venture dollars. But the shape of the win has changed. The old curve rewarded patience. The new one rewards orchestration.
The companies that will define the next chapter - in compute, in space, in energy, in the materials and machines a sovereign economy runs on - are being built right now, in the compressed front and the deep canyon of a lifecycle that no longer looks the way the textbooks describe.
The map has been redrawn.